diff --git "a/checkpoint-10452/trainer_state.json" "b/checkpoint-10452/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-10452/trainer_state.json"
@@ -0,0 +1,73225 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999521645539344,
+  "eval_steps": 2613,
+  "global_step": 10452,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.240769189720181,
+      "learning_rate": 1.9138755980861247e-07,
+      "loss": 1.9865,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.280837297439575,
+      "eval_runtime": 4185.9784,
+      "eval_samples_per_second": 19.976,
+      "eval_steps_per_second": 2.497,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.223090418187222,
+      "learning_rate": 3.8277511961722493e-07,
+      "loss": 1.9703,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6145046678806407,
+      "learning_rate": 5.741626794258373e-07,
+      "loss": 2.1381,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.6296311605868885,
+      "learning_rate": 7.655502392344499e-07,
+      "loss": 2.456,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8401650197500428,
+      "learning_rate": 9.569377990430622e-07,
+      "loss": 2.1077,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.683890691252437,
+      "learning_rate": 1.1483253588516746e-06,
+      "loss": 2.1413,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6432534368591782,
+      "learning_rate": 1.339712918660287e-06,
+      "loss": 2.1568,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.663365277527622,
+      "learning_rate": 1.5311004784688997e-06,
+      "loss": 2.3463,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6372978147069697,
+      "learning_rate": 1.722488038277512e-06,
+      "loss": 2.1712,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.852872878211269,
+      "learning_rate": 1.9138755980861244e-06,
+      "loss": 2.2722,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6549109607459043,
+      "learning_rate": 2.105263157894737e-06,
+      "loss": 1.9729,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2318002906734535,
+      "learning_rate": 2.2966507177033493e-06,
+      "loss": 2.0595,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6991822490019413,
+      "learning_rate": 2.488038277511962e-06,
+      "loss": 2.0928,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.159355481440864,
+      "learning_rate": 2.679425837320574e-06,
+      "loss": 2.1411,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.061437005322893,
+      "learning_rate": 2.870813397129187e-06,
+      "loss": 2.3334,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7145200953671766,
+      "learning_rate": 3.0622009569377995e-06,
+      "loss": 1.9779,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.3951391847417565,
+      "learning_rate": 3.2535885167464113e-06,
+      "loss": 2.24,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.225335307764433,
+      "learning_rate": 3.444976076555024e-06,
+      "loss": 2.2625,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.08014522806263,
+      "learning_rate": 3.636363636363636e-06,
+      "loss": 2.1072,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.750695114221436,
+      "learning_rate": 3.827751196172249e-06,
+      "loss": 2.2998,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.576295112434383,
+      "learning_rate": 4.019138755980861e-06,
+      "loss": 2.1414,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.5573739807898805,
+      "learning_rate": 4.210526315789474e-06,
+      "loss": 2.3373,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.369397245889477,
+      "learning_rate": 4.401913875598086e-06,
+      "loss": 2.1658,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.28624090773153,
+      "learning_rate": 4.5933014354066986e-06,
+      "loss": 2.2938,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.9534751396409655,
+      "learning_rate": 4.784688995215311e-06,
+      "loss": 2.2733,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.381089984082533,
+      "learning_rate": 4.976076555023924e-06,
+      "loss": 2.2244,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.226111694547378,
+      "learning_rate": 5.167464114832536e-06,
+      "loss": 2.2223,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5666909272147245,
+      "learning_rate": 5.358851674641148e-06,
+      "loss": 1.7762,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.398756537107184,
+      "learning_rate": 5.550239234449761e-06,
+      "loss": 2.4073,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.315805192926013,
+      "learning_rate": 5.741626794258374e-06,
+      "loss": 2.1847,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.532277066830244,
+      "learning_rate": 5.933014354066986e-06,
+      "loss": 2.2245,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.413171467566141,
+      "learning_rate": 6.124401913875599e-06,
+      "loss": 2.0755,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2310871639043746,
+      "learning_rate": 6.315789473684211e-06,
+      "loss": 2.2345,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3848849049261207,
+      "learning_rate": 6.5071770334928226e-06,
+      "loss": 2.0588,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0082260641673706,
+      "learning_rate": 6.698564593301436e-06,
+      "loss": 2.1575,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.45419363377176,
+      "learning_rate": 6.889952153110048e-06,
+      "loss": 2.2532,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.316572782291838,
+      "learning_rate": 7.081339712918661e-06,
+      "loss": 1.9725,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2302956707592998,
+      "learning_rate": 7.272727272727272e-06,
+      "loss": 2.2166,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2417025568909201,
+      "learning_rate": 7.464114832535886e-06,
+      "loss": 2.2785,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0623452341947435,
+      "learning_rate": 7.655502392344498e-06,
+      "loss": 2.1188,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9459508227302174,
+      "learning_rate": 7.846889952153112e-06,
+      "loss": 2.1225,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9821381236389229,
+      "learning_rate": 8.038277511961722e-06,
+      "loss": 1.9644,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8227720071849114,
+      "learning_rate": 8.229665071770336e-06,
+      "loss": 1.8818,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9772589694714745,
+      "learning_rate": 8.421052631578948e-06,
+      "loss": 2.1738,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8288391194584817,
+      "learning_rate": 8.61244019138756e-06,
+      "loss": 2.0598,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7365091983688079,
+      "learning_rate": 8.803827751196173e-06,
+      "loss": 2.0051,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8202389292532828,
+      "learning_rate": 8.995215311004785e-06,
+      "loss": 2.0005,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7517874679840845,
+      "learning_rate": 9.186602870813397e-06,
+      "loss": 1.9756,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8041110653388241,
+      "learning_rate": 9.377990430622011e-06,
+      "loss": 2.1019,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7364867668939584,
+      "learning_rate": 9.569377990430622e-06,
+      "loss": 1.8784,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.5933286532299948,
+      "learning_rate": 9.760765550239234e-06,
+      "loss": 1.743,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.792868413879173,
+      "learning_rate": 9.952153110047848e-06,
+      "loss": 1.8373,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8312221240251918,
+      "learning_rate": 1.014354066985646e-05,
+      "loss": 1.9098,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8539046550616419,
+      "learning_rate": 1.0334928229665072e-05,
+      "loss": 2.0752,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7680076367144099,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 2.0076,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6386760911858004,
+      "learning_rate": 1.0717703349282297e-05,
+      "loss": 1.8095,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7287884197298926,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 1.9806,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7722026786294885,
+      "learning_rate": 1.1100478468899523e-05,
+      "loss": 1.9635,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7231409481435274,
+      "learning_rate": 1.1291866028708133e-05,
+      "loss": 1.8613,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7824727930965788,
+      "learning_rate": 1.1483253588516747e-05,
+      "loss": 2.1725,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6677563917530093,
+      "learning_rate": 1.167464114832536e-05,
+      "loss": 1.9352,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7009328416411565,
+      "learning_rate": 1.1866028708133972e-05,
+      "loss": 1.8884,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9109348341862854,
+      "learning_rate": 1.2057416267942584e-05,
+      "loss": 2.3183,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8411347289624691,
+      "learning_rate": 1.2248803827751198e-05,
+      "loss": 1.8075,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7674394625955453,
+      "learning_rate": 1.2440191387559808e-05,
+      "loss": 1.9963,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7848289715014719,
+      "learning_rate": 1.2631578947368422e-05,
+      "loss": 1.9276,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7238409990110771,
+      "learning_rate": 1.2822966507177035e-05,
+      "loss": 1.9353,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8235100929493389,
+      "learning_rate": 1.3014354066985645e-05,
+      "loss": 1.8938,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6283052176051154,
+      "learning_rate": 1.320574162679426e-05,
+      "loss": 1.8028,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.654778519712797,
+      "learning_rate": 1.3397129186602871e-05,
+      "loss": 1.6469,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7728600382414659,
+      "learning_rate": 1.3588516746411483e-05,
+      "loss": 1.7839,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9160985506203067,
+      "learning_rate": 1.3779904306220096e-05,
+      "loss": 1.8154,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8592513182940328,
+      "learning_rate": 1.397129186602871e-05,
+      "loss": 1.8178,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0716389160468884,
+      "learning_rate": 1.4162679425837322e-05,
+      "loss": 1.9752,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0526260733114412,
+      "learning_rate": 1.4354066985645934e-05,
+      "loss": 1.8998,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9676822741503439,
+      "learning_rate": 1.4545454545454545e-05,
+      "loss": 1.7514,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2545561036432802,
+      "learning_rate": 1.4736842105263157e-05,
+      "loss": 1.8165,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.834675739891747,
+      "learning_rate": 1.4928229665071772e-05,
+      "loss": 1.8987,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.25568745378183,
+      "learning_rate": 1.5119617224880383e-05,
+      "loss": 1.8836,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8547248072580093,
+      "learning_rate": 1.5311004784688995e-05,
+      "loss": 1.625,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9130624521617219,
+      "learning_rate": 1.5502392344497607e-05,
+      "loss": 1.65,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4571618225829908,
+      "learning_rate": 1.5693779904306223e-05,
+      "loss": 1.7592,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5681812827811266,
+      "learning_rate": 1.5885167464114832e-05,
+      "loss": 1.7274,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7707101692645586,
+      "learning_rate": 1.6076555023923444e-05,
+      "loss": 1.7496,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4920166991759014,
+      "learning_rate": 1.6267942583732056e-05,
+      "loss": 1.815,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.765151570274179,
+      "learning_rate": 1.6459330143540672e-05,
+      "loss": 1.6131,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7780322419019302,
+      "learning_rate": 1.6650717703349284e-05,
+      "loss": 1.6553,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6971849026842456,
+      "learning_rate": 1.6842105263157896e-05,
+      "loss": 1.6212,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6737627863204966,
+      "learning_rate": 1.7033492822966505e-05,
+      "loss": 1.6319,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7581487338796276,
+      "learning_rate": 1.722488038277512e-05,
+      "loss": 1.5457,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6884712501693219,
+      "learning_rate": 1.7416267942583733e-05,
+      "loss": 1.682,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6472187205118166,
+      "learning_rate": 1.7607655502392345e-05,
+      "loss": 1.7893,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5392317136449472,
+      "learning_rate": 1.7799043062200958e-05,
+      "loss": 1.5724,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.613083864173068,
+      "learning_rate": 1.799043062200957e-05,
+      "loss": 1.5933,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.554854239010525,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.4444,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5895819501317506,
+      "learning_rate": 1.8373205741626794e-05,
+      "loss": 1.463,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6834369586439761,
+      "learning_rate": 1.8564593301435407e-05,
+      "loss": 1.6334,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5941595870489239,
+      "learning_rate": 1.8755980861244022e-05,
+      "loss": 1.6881,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6765509171345243,
+      "learning_rate": 1.8947368421052634e-05,
+      "loss": 1.5896,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5523045136213458,
+      "learning_rate": 1.9138755980861243e-05,
+      "loss": 1.4551,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6537194645181346,
+      "learning_rate": 1.9330143540669855e-05,
+      "loss": 1.5513,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5087498837023968,
+      "learning_rate": 1.9521531100478468e-05,
+      "loss": 1.4715,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6335719742406299,
+      "learning_rate": 1.9712918660287083e-05,
+      "loss": 1.5313,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.647032541641365,
+      "learning_rate": 1.9904306220095696e-05,
+      "loss": 1.4794,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5641054209541718,
+      "learning_rate": 2.0095693779904308e-05,
+      "loss": 1.5301,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.623411201252464,
+      "learning_rate": 2.028708133971292e-05,
+      "loss": 1.4618,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6469566283028377,
+      "learning_rate": 2.0478468899521532e-05,
+      "loss": 1.5438,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5328026134772941,
+      "learning_rate": 2.0669856459330144e-05,
+      "loss": 1.4587,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5404945621124031,
+      "learning_rate": 2.0861244019138757e-05,
+      "loss": 1.3749,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6198671572468233,
+      "learning_rate": 2.105263157894737e-05,
+      "loss": 1.5662,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5344071404450242,
+      "learning_rate": 2.1244019138755985e-05,
+      "loss": 1.6258,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5077736180550365,
+      "learning_rate": 2.1435406698564593e-05,
+      "loss": 1.4085,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5916690974950262,
+      "learning_rate": 2.1626794258373206e-05,
+      "loss": 1.4315,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4799015981708095,
+      "learning_rate": 2.1818181818181818e-05,
+      "loss": 1.4183,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4571232168458159,
+      "learning_rate": 2.2009569377990433e-05,
+      "loss": 1.3658,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4917591648756044,
+      "learning_rate": 2.2200956937799046e-05,
+      "loss": 1.3724,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4170683829952185,
+      "learning_rate": 2.2392344497607658e-05,
+      "loss": 1.4197,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.49893891430948417,
+      "learning_rate": 2.2583732057416267e-05,
+      "loss": 1.4005,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4056789281529712,
+      "learning_rate": 2.2775119617224882e-05,
+      "loss": 1.4419,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47274513283719816,
+      "learning_rate": 2.2966507177033495e-05,
+      "loss": 1.3163,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4495315355981231,
+      "learning_rate": 2.3157894736842107e-05,
+      "loss": 1.3599,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4142057365520808,
+      "learning_rate": 2.334928229665072e-05,
+      "loss": 1.4484,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4421864791497608,
+      "learning_rate": 2.354066985645933e-05,
+      "loss": 1.3629,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.44850267009277406,
+      "learning_rate": 2.3732057416267943e-05,
+      "loss": 1.4036,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.49021107734394687,
+      "learning_rate": 2.3923444976076556e-05,
+      "loss": 1.3724,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.47738353199872885,
+      "learning_rate": 2.4114832535885168e-05,
+      "loss": 1.3304,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4704616937001858,
+      "learning_rate": 2.4306220095693784e-05,
+      "loss": 1.376,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4351806446488345,
+      "learning_rate": 2.4497607655502396e-05,
+      "loss": 1.3438,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5275547600152736,
+      "learning_rate": 2.4688995215311005e-05,
+      "loss": 1.2731,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3787982803600526,
+      "learning_rate": 2.4880382775119617e-05,
+      "loss": 1.3392,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3850735556535156,
+      "learning_rate": 2.507177033492823e-05,
+      "loss": 1.2673,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4339128757111585,
+      "learning_rate": 2.5263157894736845e-05,
+      "loss": 1.3284,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.39005187564392557,
+      "learning_rate": 2.5454545454545454e-05,
+      "loss": 1.3347,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.35602922226403455,
+      "learning_rate": 2.564593301435407e-05,
+      "loss": 1.2362,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.40601160369590406,
+      "learning_rate": 2.583732057416268e-05,
+      "loss": 1.3591,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4408392546545607,
+      "learning_rate": 2.602870813397129e-05,
+      "loss": 1.2837,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.46362552165830334,
+      "learning_rate": 2.6220095693779906e-05,
+      "loss": 1.316,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3504892333833388,
+      "learning_rate": 2.641148325358852e-05,
+      "loss": 1.1857,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4362884443543931,
+      "learning_rate": 2.660287081339713e-05,
+      "loss": 1.2435,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.50850093231113,
+      "learning_rate": 2.6794258373205743e-05,
+      "loss": 1.3235,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4187742784108479,
+      "learning_rate": 2.698564593301435e-05,
+      "loss": 1.3693,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.44107080428520234,
+      "learning_rate": 2.7177033492822967e-05,
+      "loss": 1.202,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.42167608964495673,
+      "learning_rate": 2.7368421052631583e-05,
+      "loss": 1.2693,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4269486411389392,
+      "learning_rate": 2.755980861244019e-05,
+      "loss": 1.1825,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4220464029961541,
+      "learning_rate": 2.7751196172248807e-05,
+      "loss": 1.3205,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4620642882432513,
+      "learning_rate": 2.794258373205742e-05,
+      "loss": 1.2891,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.43743294942437483,
+      "learning_rate": 2.8133971291866028e-05,
+      "loss": 1.1969,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.4414019243397515,
+      "learning_rate": 2.8325358851674644e-05,
+      "loss": 1.1672,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1295244665123245,
+      "learning_rate": 2.8516746411483253e-05,
+      "loss": 1.3175,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5055922001428373,
+      "learning_rate": 2.8708133971291868e-05,
+      "loss": 1.2696,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.48319661000499564,
+      "learning_rate": 2.889952153110048e-05,
+      "loss": 1.3168,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5720466946520338,
+      "learning_rate": 2.909090909090909e-05,
+      "loss": 1.2982,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.68753321360197,
+      "learning_rate": 2.9282296650717705e-05,
+      "loss": 1.2791,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6221880802372888,
+      "learning_rate": 2.9473684210526314e-05,
+      "loss": 1.2647,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.5504261765445813,
+      "learning_rate": 2.966507177033493e-05,
+      "loss": 1.3102,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6618113787746298,
+      "learning_rate": 2.9856459330143545e-05,
+      "loss": 1.292,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6002633450221915,
+      "learning_rate": 3.0047846889952154e-05,
+      "loss": 1.33,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7118508153621136,
+      "learning_rate": 3.0239234449760766e-05,
+      "loss": 1.2062,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6591754857765905,
+      "learning_rate": 3.043062200956938e-05,
+      "loss": 1.2539,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4714338006090391,
+      "learning_rate": 3.062200956937799e-05,
+      "loss": 1.2384,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0300987484894157,
+      "learning_rate": 3.08133971291866e-05,
+      "loss": 1.2171,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6469298394859133,
+      "learning_rate": 3.1004784688995215e-05,
+      "loss": 1.0836,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5601908591127592,
+      "learning_rate": 3.119617224880383e-05,
+      "loss": 1.1867,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7363177930575833,
+      "learning_rate": 3.1387559808612446e-05,
+      "loss": 1.3128,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7107903409156414,
+      "learning_rate": 3.157894736842105e-05,
+      "loss": 1.2501,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.464636577316954,
+      "learning_rate": 3.1770334928229664e-05,
+      "loss": 1.1671,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8259444575108333,
+      "learning_rate": 3.196172248803828e-05,
+      "loss": 1.2928,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7945572957938246,
+      "learning_rate": 3.215311004784689e-05,
+      "loss": 1.2516,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5895739273419747,
+      "learning_rate": 3.234449760765551e-05,
+      "loss": 1.162,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6927568627921579,
+      "learning_rate": 3.253588516746411e-05,
+      "loss": 1.2802,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.639241808296525,
+      "learning_rate": 3.272727272727273e-05,
+      "loss": 1.1999,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8979192020036151,
+      "learning_rate": 3.2918660287081344e-05,
+      "loss": 1.1854,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5314730657868201,
+      "learning_rate": 3.311004784688995e-05,
+      "loss": 1.2365,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3835011610660225,
+      "learning_rate": 3.330143540669857e-05,
+      "loss": 1.3014,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46600481942740146,
+      "learning_rate": 3.349282296650718e-05,
+      "loss": 1.2222,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.568331869349057,
+      "learning_rate": 3.368421052631579e-05,
+      "loss": 1.2219,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5161782745938108,
+      "learning_rate": 3.3875598086124405e-05,
+      "loss": 1.0828,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5678907934059946,
+      "learning_rate": 3.406698564593301e-05,
+      "loss": 1.241,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.40729089487435055,
+      "learning_rate": 3.425837320574163e-05,
+      "loss": 1.2201,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5273474366443621,
+      "learning_rate": 3.444976076555024e-05,
+      "loss": 1.1262,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4807276904108668,
+      "learning_rate": 3.4641148325358854e-05,
+      "loss": 1.1957,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5372452514446321,
+      "learning_rate": 3.4832535885167466e-05,
+      "loss": 1.1621,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6605408366116067,
+      "learning_rate": 3.502392344497608e-05,
+      "loss": 1.2327,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5555779108688886,
+      "learning_rate": 3.521531100478469e-05,
+      "loss": 1.1938,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5517608446497938,
+      "learning_rate": 3.54066985645933e-05,
+      "loss": 1.2651,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4524359827250868,
+      "learning_rate": 3.5598086124401915e-05,
+      "loss": 1.2182,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4495056117139968,
+      "learning_rate": 3.578947368421053e-05,
+      "loss": 1.2133,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.43420633627744476,
+      "learning_rate": 3.598086124401914e-05,
+      "loss": 1.078,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4991578407930919,
+      "learning_rate": 3.617224880382775e-05,
+      "loss": 1.0683,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46736895171034254,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 1.1533,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48830760264155376,
+      "learning_rate": 3.6555023923444976e-05,
+      "loss": 1.2539,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5148588474362115,
+      "learning_rate": 3.674641148325359e-05,
+      "loss": 1.0486,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4669961130515612,
+      "learning_rate": 3.693779904306221e-05,
+      "loss": 1.1949,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.32879504352969363,
+      "learning_rate": 3.712918660287081e-05,
+      "loss": 1.189,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.35633398240141656,
+      "learning_rate": 3.7320574162679425e-05,
+      "loss": 1.2112,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.42378382756263683,
+      "learning_rate": 3.7511961722488044e-05,
+      "loss": 1.1999,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.41483450868577093,
+      "learning_rate": 3.770334928229665e-05,
+      "loss": 1.1283,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4603179290827383,
+      "learning_rate": 3.789473684210527e-05,
+      "loss": 1.2353,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5574543274654291,
+      "learning_rate": 3.8086124401913874e-05,
+      "loss": 1.1948,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6846239185739854,
+      "learning_rate": 3.8277511961722486e-05,
+      "loss": 1.3231,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.520683870817622,
+      "learning_rate": 3.8468899521531105e-05,
+      "loss": 1.2201,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.42683935885681074,
+      "learning_rate": 3.866028708133971e-05,
+      "loss": 1.165,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39470003294526096,
+      "learning_rate": 3.885167464114833e-05,
+      "loss": 1.1994,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4391753396869876,
+      "learning_rate": 3.9043062200956935e-05,
+      "loss": 1.1813,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4449276816002856,
+      "learning_rate": 3.9234449760765554e-05,
+      "loss": 1.1764,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45472581470602813,
+      "learning_rate": 3.9425837320574167e-05,
+      "loss": 1.1766,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37367376899076454,
+      "learning_rate": 3.961722488038277e-05,
+      "loss": 1.0581,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5183447282360463,
+      "learning_rate": 3.980861244019139e-05,
+      "loss": 1.1351,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5895054254055697,
+      "learning_rate": 4e-05,
+      "loss": 1.2017,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39968348371397444,
+      "learning_rate": 4.0191387559808616e-05,
+      "loss": 1.2353,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4089518582688394,
+      "learning_rate": 4.038277511961723e-05,
+      "loss": 1.2072,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5490805672572786,
+      "learning_rate": 4.057416267942584e-05,
+      "loss": 1.2593,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.42397521041394676,
+      "learning_rate": 4.076555023923445e-05,
+      "loss": 1.155,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4981026360180907,
+      "learning_rate": 4.0956937799043064e-05,
+      "loss": 1.2279,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.503200310794026,
+      "learning_rate": 4.114832535885168e-05,
+      "loss": 1.1134,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5127306365983669,
+      "learning_rate": 4.133971291866029e-05,
+      "loss": 1.1993,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4935452094233486,
+      "learning_rate": 4.15311004784689e-05,
+      "loss": 1.1962,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.38834813884933866,
+      "learning_rate": 4.172248803827751e-05,
+      "loss": 1.2529,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4742710182667002,
+      "learning_rate": 4.1913875598086126e-05,
+      "loss": 1.1861,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4762816225973413,
+      "learning_rate": 4.210526315789474e-05,
+      "loss": 1.2221,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5405607329270613,
+      "learning_rate": 4.229665071770335e-05,
+      "loss": 1.3099,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.38474621695351074,
+      "learning_rate": 4.248803827751197e-05,
+      "loss": 1.1669,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.47522706488556143,
+      "learning_rate": 4.2679425837320574e-05,
+      "loss": 1.1813,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.43816325603257117,
+      "learning_rate": 4.287081339712919e-05,
+      "loss": 1.1912,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.47579537819240086,
+      "learning_rate": 4.3062200956937806e-05,
+      "loss": 1.165,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5148567080870619,
+      "learning_rate": 4.325358851674641e-05,
+      "loss": 1.1997,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4019413785534228,
+      "learning_rate": 4.344497607655503e-05,
+      "loss": 1.2448,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4766231281692962,
+      "learning_rate": 4.3636363636363636e-05,
+      "loss": 1.1374,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4375975135050194,
+      "learning_rate": 4.382775119617225e-05,
+      "loss": 1.2484,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4312959866778293,
+      "learning_rate": 4.401913875598087e-05,
+      "loss": 1.1143,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4329834158025715,
+      "learning_rate": 4.421052631578947e-05,
+      "loss": 1.2335,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37836691609878936,
+      "learning_rate": 4.440191387559809e-05,
+      "loss": 1.118,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37526619539689987,
+      "learning_rate": 4.45933014354067e-05,
+      "loss": 1.2968,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4175158591003783,
+      "learning_rate": 4.4784688995215316e-05,
+      "loss": 1.1954,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37094222885704037,
+      "learning_rate": 4.497607655502393e-05,
+      "loss": 1.0976,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4586856505320375,
+      "learning_rate": 4.5167464114832533e-05,
+      "loss": 1.1628,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.44587593738328374,
+      "learning_rate": 4.535885167464115e-05,
+      "loss": 1.1917,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3594860823761438,
+      "learning_rate": 4.5550239234449765e-05,
+      "loss": 1.192,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.38283214874386773,
+      "learning_rate": 4.574162679425838e-05,
+      "loss": 1.2205,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.33492904991363276,
+      "learning_rate": 4.593301435406699e-05,
+      "loss": 1.2155,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4113511012996176,
+      "learning_rate": 4.6124401913875595e-05,
+      "loss": 1.2222,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4337572841707065,
+      "learning_rate": 4.6315789473684214e-05,
+      "loss": 1.0193,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.34650969754183997,
+      "learning_rate": 4.6507177033492826e-05,
+      "loss": 1.2571,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4758935405313615,
+      "learning_rate": 4.669856459330144e-05,
+      "loss": 1.1967,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.39684085109310996,
+      "learning_rate": 4.688995215311005e-05,
+      "loss": 1.1169,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4823470965603195,
+      "learning_rate": 4.708133971291866e-05,
+      "loss": 1.2347,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.37717028125791024,
+      "learning_rate": 4.7272727272727275e-05,
+      "loss": 1.094,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3667379872819918,
+      "learning_rate": 4.746411483253589e-05,
+      "loss": 1.2391,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3407975972338634,
+      "learning_rate": 4.76555023923445e-05,
+      "loss": 1.0632,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.49238987132507245,
+      "learning_rate": 4.784688995215311e-05,
+      "loss": 1.2121,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46706882037804415,
+      "learning_rate": 4.8038277511961724e-05,
+      "loss": 1.2437,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4088473587226235,
+      "learning_rate": 4.8229665071770336e-05,
+      "loss": 1.1572,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3821462444529073,
+      "learning_rate": 4.842105263157895e-05,
+      "loss": 1.1603,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4782990514183433,
+      "learning_rate": 4.861244019138757e-05,
+      "loss": 1.2409,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.45087555655452305,
+      "learning_rate": 4.880382775119617e-05,
+      "loss": 1.1556,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3865676678179632,
+      "learning_rate": 4.899521531100479e-05,
+      "loss": 1.1945,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4099982787948705,
+      "learning_rate": 4.91866028708134e-05,
+      "loss": 1.17,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.46016977600480446,
+      "learning_rate": 4.937799043062201e-05,
+      "loss": 1.2024,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4712335677622079,
+      "learning_rate": 4.956937799043063e-05,
+      "loss": 1.1252,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3955435220531026,
+      "learning_rate": 4.9760765550239234e-05,
+      "loss": 1.201,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.48382588647342273,
+      "learning_rate": 4.995215311004785e-05,
+      "loss": 1.3135,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39028060459262043,
+      "learning_rate": 5.014354066985646e-05,
+      "loss": 1.2573,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4094204023390644,
+      "learning_rate": 5.033492822966508e-05,
+      "loss": 1.1411,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37405702073279246,
+      "learning_rate": 5.052631578947369e-05,
+      "loss": 1.1924,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38656220760519894,
+      "learning_rate": 5.0717703349282295e-05,
+      "loss": 1.2062,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4152109433386491,
+      "learning_rate": 5.090909090909091e-05,
+      "loss": 1.2412,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3485732917182752,
+      "learning_rate": 5.1100478468899526e-05,
+      "loss": 1.2917,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3551920352066169,
+      "learning_rate": 5.129186602870814e-05,
+      "loss": 1.1494,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3817527119708618,
+      "learning_rate": 5.1483253588516744e-05,
+      "loss": 1.2689,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.35043446712727827,
+      "learning_rate": 5.167464114832536e-05,
+      "loss": 1.1263,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4560107489672836,
+      "learning_rate": 5.1866028708133975e-05,
+      "loss": 1.2355,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.411519693827521,
+      "learning_rate": 5.205741626794258e-05,
+      "loss": 1.2213,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4017892181431639,
+      "learning_rate": 5.22488038277512e-05,
+      "loss": 1.2103,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.46457066156222404,
+      "learning_rate": 5.244019138755981e-05,
+      "loss": 1.1638,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41073989648253845,
+      "learning_rate": 5.2631578947368424e-05,
+      "loss": 1.2773,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3730027098705776,
+      "learning_rate": 5.282296650717704e-05,
+      "loss": 1.1376,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3860157617102252,
+      "learning_rate": 5.301435406698565e-05,
+      "loss": 1.2109,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4349386298334119,
+      "learning_rate": 5.320574162679426e-05,
+      "loss": 1.1575,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4640773330272392,
+      "learning_rate": 5.339712918660288e-05,
+      "loss": 1.1985,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.45865729558497004,
+      "learning_rate": 5.3588516746411485e-05,
+      "loss": 1.2145,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37111837421497373,
+      "learning_rate": 5.37799043062201e-05,
+      "loss": 1.1538,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40099943789483994,
+      "learning_rate": 5.39712918660287e-05,
+      "loss": 1.0804,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.42817755702563415,
+      "learning_rate": 5.416267942583733e-05,
+      "loss": 1.202,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.34712198940947375,
+      "learning_rate": 5.4354066985645934e-05,
+      "loss": 1.2341,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39457404932238177,
+      "learning_rate": 5.4545454545454546e-05,
+      "loss": 1.1571,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36930568124767427,
+      "learning_rate": 5.4736842105263165e-05,
+      "loss": 1.1275,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4662030178725213,
+      "learning_rate": 5.492822966507177e-05,
+      "loss": 1.1811,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40587674329213624,
+      "learning_rate": 5.511961722488038e-05,
+      "loss": 1.102,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.35779077893034156,
+      "learning_rate": 5.5311004784689e-05,
+      "loss": 1.2788,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.31159530938785,
+      "learning_rate": 5.5502392344497614e-05,
+      "loss": 1.1778,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4160898781250681,
+      "learning_rate": 5.569377990430622e-05,
+      "loss": 1.2064,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43660026046564104,
+      "learning_rate": 5.588516746411484e-05,
+      "loss": 1.1494,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40119987919562433,
+      "learning_rate": 5.607655502392345e-05,
+      "loss": 1.1623,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37717288362355916,
+      "learning_rate": 5.6267942583732056e-05,
+      "loss": 1.1194,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4057229096001704,
+      "learning_rate": 5.645933014354067e-05,
+      "loss": 1.0581,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4670818825975868,
+      "learning_rate": 5.665071770334929e-05,
+      "loss": 1.171,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5482039795115164,
+      "learning_rate": 5.68421052631579e-05,
+      "loss": 1.1082,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4468277071138632,
+      "learning_rate": 5.7033492822966505e-05,
+      "loss": 1.2785,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3500497863626851,
+      "learning_rate": 5.7224880382775124e-05,
+      "loss": 1.1713,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5457539009822924,
+      "learning_rate": 5.7416267942583736e-05,
+      "loss": 1.1259,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3521279155346515,
+      "learning_rate": 5.760765550239234e-05,
+      "loss": 1.2143,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38014559184795155,
+      "learning_rate": 5.779904306220096e-05,
+      "loss": 1.2169,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4368587205355613,
+      "learning_rate": 5.799043062200957e-05,
+      "loss": 1.1354,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3773750084543655,
+      "learning_rate": 5.818181818181818e-05,
+      "loss": 1.204,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4661908984222087,
+      "learning_rate": 5.8373205741626804e-05,
+      "loss": 1.2339,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4233656313317972,
+      "learning_rate": 5.856459330143541e-05,
+      "loss": 1.3069,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4452233614591402,
+      "learning_rate": 5.875598086124402e-05,
+      "loss": 1.1966,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.379869871739064,
+      "learning_rate": 5.894736842105263e-05,
+      "loss": 1.1233,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3634900550395493,
+      "learning_rate": 5.9138755980861246e-05,
+      "loss": 1.2106,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38506818193045056,
+      "learning_rate": 5.933014354066986e-05,
+      "loss": 1.2411,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41159846103302217,
+      "learning_rate": 5.9521531100478464e-05,
+      "loss": 1.3279,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.28921210577002654,
+      "learning_rate": 5.971291866028709e-05,
+      "loss": 1.0785,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38616302251379914,
+      "learning_rate": 5.9904306220095695e-05,
+      "loss": 1.2536,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4101883225822147,
+      "learning_rate": 6.009569377990431e-05,
+      "loss": 1.1666,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3423514370886581,
+      "learning_rate": 6.028708133971293e-05,
+      "loss": 1.2331,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.41207456262790537,
+      "learning_rate": 6.047846889952153e-05,
+      "loss": 1.1403,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40588734564935647,
+      "learning_rate": 6.0669856459330144e-05,
+      "loss": 1.1986,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37104840888562607,
+      "learning_rate": 6.086124401913876e-05,
+      "loss": 1.169,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39655898223668296,
+      "learning_rate": 6.105263157894737e-05,
+      "loss": 1.1296,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5221245550161207,
+      "learning_rate": 6.124401913875598e-05,
+      "loss": 1.0809,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4021445879644258,
+      "learning_rate": 6.143540669856461e-05,
+      "loss": 1.1897,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38180254930918234,
+      "learning_rate": 6.16267942583732e-05,
+      "loss": 1.1859,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3828401683172079,
+      "learning_rate": 6.181818181818182e-05,
+      "loss": 1.0985,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.43356536766643855,
+      "learning_rate": 6.200956937799043e-05,
+      "loss": 1.1524,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.39197554253988764,
+      "learning_rate": 6.220095693779904e-05,
+      "loss": 1.1607,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3768192836960789,
+      "learning_rate": 6.239234449760765e-05,
+      "loss": 1.2115,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4123016217273888,
+      "learning_rate": 6.258373205741627e-05,
+      "loss": 1.2046,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38690332283125584,
+      "learning_rate": 6.277511961722489e-05,
+      "loss": 1.2207,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3958359496615601,
+      "learning_rate": 6.296650717703349e-05,
+      "loss": 1.1334,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36561214007046366,
+      "learning_rate": 6.31578947368421e-05,
+      "loss": 1.1196,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36953124111301633,
+      "learning_rate": 6.334928229665073e-05,
+      "loss": 1.1203,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.391180586728009,
+      "learning_rate": 6.354066985645933e-05,
+      "loss": 1.2479,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.483739471817649,
+      "learning_rate": 6.373205741626794e-05,
+      "loss": 1.1225,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4233252985553476,
+      "learning_rate": 6.392344497607657e-05,
+      "loss": 1.1772,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4261190391654762,
+      "learning_rate": 6.411483253588518e-05,
+      "loss": 1.1752,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4217550233368759,
+      "learning_rate": 6.430622009569378e-05,
+      "loss": 1.2335,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4126104400962645,
+      "learning_rate": 6.449760765550239e-05,
+      "loss": 1.1095,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.48994576863601885,
+      "learning_rate": 6.468899521531101e-05,
+      "loss": 1.2933,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3928982582790676,
+      "learning_rate": 6.488038277511961e-05,
+      "loss": 1.1598,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38476512783934874,
+      "learning_rate": 6.507177033492823e-05,
+      "loss": 1.3208,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3810331248367921,
+      "learning_rate": 6.526315789473685e-05,
+      "loss": 1.1183,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.38985962649469763,
+      "learning_rate": 6.545454545454546e-05,
+      "loss": 1.2319,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.44280071804731985,
+      "learning_rate": 6.564593301435406e-05,
+      "loss": 1.2374,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3221949054184306,
+      "learning_rate": 6.583732057416269e-05,
+      "loss": 1.182,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3765883193637935,
+      "learning_rate": 6.60287081339713e-05,
+      "loss": 1.085,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36077725709386577,
+      "learning_rate": 6.62200956937799e-05,
+      "loss": 1.1741,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3844614909912839,
+      "learning_rate": 6.641148325358852e-05,
+      "loss": 1.1308,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.356160190962461,
+      "learning_rate": 6.660287081339714e-05,
+      "loss": 1.2303,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40199466397728084,
+      "learning_rate": 6.679425837320575e-05,
+      "loss": 1.1404,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40760457805411343,
+      "learning_rate": 6.698564593301436e-05,
+      "loss": 1.1441,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36664650507067836,
+      "learning_rate": 6.717703349282297e-05,
+      "loss": 1.0822,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3774008230280487,
+      "learning_rate": 6.736842105263159e-05,
+      "loss": 1.1882,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3743284992535933,
+      "learning_rate": 6.755980861244018e-05,
+      "loss": 1.1174,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3803668006858151,
+      "learning_rate": 6.775119617224881e-05,
+      "loss": 1.2074,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3777040131335034,
+      "learning_rate": 6.794258373205742e-05,
+      "loss": 1.0946,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3794465077922223,
+      "learning_rate": 6.813397129186602e-05,
+      "loss": 1.0683,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3794400803152343,
+      "learning_rate": 6.832535885167465e-05,
+      "loss": 1.1377,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3246160370634747,
+      "learning_rate": 6.851674641148326e-05,
+      "loss": 1.2539,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.3663994418682161,
+      "learning_rate": 6.870813397129187e-05,
+      "loss": 1.179,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4165644814006045,
+      "learning_rate": 6.889952153110048e-05,
+      "loss": 1.2068,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.40002794816857074,
+      "learning_rate": 6.90909090909091e-05,
+      "loss": 1.236,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.36752801689828113,
+      "learning_rate": 6.928229665071771e-05,
+      "loss": 1.1806,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4118641376720516,
+      "learning_rate": 6.947368421052632e-05,
+      "loss": 1.2099,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.342905855911731,
+      "learning_rate": 6.966507177033493e-05,
+      "loss": 1.1753,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.37601728725049927,
+      "learning_rate": 6.985645933014354e-05,
+      "loss": 1.2123,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3996876187572624,
+      "learning_rate": 7.004784688995216e-05,
+      "loss": 1.2063,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4995513430917594,
+      "learning_rate": 7.023923444976077e-05,
+      "loss": 1.1249,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38011883982869116,
+      "learning_rate": 7.043062200956938e-05,
+      "loss": 1.091,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40185997614954744,
+      "learning_rate": 7.0622009569378e-05,
+      "loss": 1.2879,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37530412149005404,
+      "learning_rate": 7.08133971291866e-05,
+      "loss": 1.2832,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4067556360842582,
+      "learning_rate": 7.100478468899522e-05,
+      "loss": 1.0513,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4450195271251879,
+      "learning_rate": 7.119617224880383e-05,
+      "loss": 1.1781,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3878038365086055,
+      "learning_rate": 7.138755980861244e-05,
+      "loss": 1.2654,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34902250016212555,
+      "learning_rate": 7.157894736842105e-05,
+      "loss": 1.0807,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4789371923266489,
+      "learning_rate": 7.177033492822967e-05,
+      "loss": 1.0782,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3443621112474482,
+      "learning_rate": 7.196172248803828e-05,
+      "loss": 1.0874,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4354953964559702,
+      "learning_rate": 7.215311004784689e-05,
+      "loss": 1.1893,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3806981757319183,
+      "learning_rate": 7.23444976076555e-05,
+      "loss": 1.1033,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3640708591532273,
+      "learning_rate": 7.253588516746413e-05,
+      "loss": 1.1008,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40487824027906655,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 1.2676,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3974099352341619,
+      "learning_rate": 7.291866028708134e-05,
+      "loss": 1.1297,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4412922873426402,
+      "learning_rate": 7.311004784688995e-05,
+      "loss": 1.2022,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4177424000999219,
+      "learning_rate": 7.330143540669856e-05,
+      "loss": 1.0769,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37843676467423115,
+      "learning_rate": 7.349282296650718e-05,
+      "loss": 1.1895,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37904784029757954,
+      "learning_rate": 7.368421052631579e-05,
+      "loss": 1.0346,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3678229762318383,
+      "learning_rate": 7.387559808612442e-05,
+      "loss": 1.1415,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.44631509974989175,
+      "learning_rate": 7.406698564593301e-05,
+      "loss": 1.1593,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4904633395013365,
+      "learning_rate": 7.425837320574163e-05,
+      "loss": 1.1944,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37152475181241584,
+      "learning_rate": 7.444976076555025e-05,
+      "loss": 1.1698,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4011204212131284,
+      "learning_rate": 7.464114832535885e-05,
+      "loss": 1.205,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.36761682195572815,
+      "learning_rate": 7.483253588516746e-05,
+      "loss": 1.2973,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40374546939981903,
+      "learning_rate": 7.502392344497609e-05,
+      "loss": 1.1507,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4255006771988514,
+      "learning_rate": 7.52153110047847e-05,
+      "loss": 1.1891,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3865518214834786,
+      "learning_rate": 7.54066985645933e-05,
+      "loss": 1.185,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41827717268535636,
+      "learning_rate": 7.559808612440191e-05,
+      "loss": 1.0622,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37110879057589674,
+      "learning_rate": 7.578947368421054e-05,
+      "loss": 1.1381,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.46618966480421226,
+      "learning_rate": 7.598086124401914e-05,
+      "loss": 1.2105,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4266096767917638,
+      "learning_rate": 7.617224880382775e-05,
+      "loss": 1.2304,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3951434711007925,
+      "learning_rate": 7.636363636363637e-05,
+      "loss": 1.1194,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5249468851294161,
+      "learning_rate": 7.655502392344497e-05,
+      "loss": 1.1557,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35849242393679126,
+      "learning_rate": 7.674641148325359e-05,
+      "loss": 1.1586,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4163074752217252,
+      "learning_rate": 7.693779904306221e-05,
+      "loss": 1.1884,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39359654231287206,
+      "learning_rate": 7.712918660287082e-05,
+      "loss": 1.2162,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40661594470681456,
+      "learning_rate": 7.732057416267942e-05,
+      "loss": 1.1578,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4060870336417891,
+      "learning_rate": 7.751196172248805e-05,
+      "loss": 1.2462,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34776732343372285,
+      "learning_rate": 7.770334928229666e-05,
+      "loss": 1.1907,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37012929061225097,
+      "learning_rate": 7.789473684210526e-05,
+      "loss": 1.084,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3191687282557576,
+      "learning_rate": 7.808612440191387e-05,
+      "loss": 1.1337,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.428711696888013,
+      "learning_rate": 7.82775119617225e-05,
+      "loss": 1.2142,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5197761387762115,
+      "learning_rate": 7.846889952153111e-05,
+      "loss": 1.2915,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.41273170978967216,
+      "learning_rate": 7.866028708133971e-05,
+      "loss": 1.1028,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3568549043871766,
+      "learning_rate": 7.885167464114833e-05,
+      "loss": 1.1575,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43257689825178397,
+      "learning_rate": 7.904306220095695e-05,
+      "loss": 1.1115,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4355392516104201,
+      "learning_rate": 7.923444976076554e-05,
+      "loss": 1.208,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.48359013518775035,
+      "learning_rate": 7.942583732057417e-05,
+      "loss": 1.1864,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3730295702560985,
+      "learning_rate": 7.961722488038278e-05,
+      "loss": 1.1261,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3966685187279031,
+      "learning_rate": 7.98086124401914e-05,
+      "loss": 1.2786,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3786523559458125,
+      "learning_rate": 8e-05,
+      "loss": 1.1127,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4039420784784596,
+      "learning_rate": 8.019138755980862e-05,
+      "loss": 1.1653,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3836563304661635,
+      "learning_rate": 8.038277511961723e-05,
+      "loss": 1.1504,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3915882211031514,
+      "learning_rate": 8.057416267942584e-05,
+      "loss": 1.1881,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3585833381861096,
+      "learning_rate": 8.076555023923446e-05,
+      "loss": 1.1215,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34093033582588217,
+      "learning_rate": 8.095693779904307e-05,
+      "loss": 1.1502,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35015224800454925,
+      "learning_rate": 8.114832535885168e-05,
+      "loss": 1.1377,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.36225302300603957,
+      "learning_rate": 8.133971291866029e-05,
+      "loss": 1.146,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.401290151391233,
+      "learning_rate": 8.15311004784689e-05,
+      "loss": 1.1257,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42354295033157774,
+      "learning_rate": 8.172248803827752e-05,
+      "loss": 1.1861,
+      "step": 427
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.47531433461725525,
+      "learning_rate": 8.191387559808613e-05,
+      "loss": 1.2387,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3956478442168383,
+      "learning_rate": 8.210526315789474e-05,
+      "loss": 1.2046,
+      "step": 429
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40300702532566746,
+      "learning_rate": 8.229665071770335e-05,
+      "loss": 1.2882,
+      "step": 430
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4247965626832039,
+      "learning_rate": 8.248803827751197e-05,
+      "loss": 1.2151,
+      "step": 431
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3667653514802216,
+      "learning_rate": 8.267942583732058e-05,
+      "loss": 1.2397,
+      "step": 432
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.368289021497025,
+      "learning_rate": 8.287081339712919e-05,
+      "loss": 1.1336,
+      "step": 433
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.35548517154282255,
+      "learning_rate": 8.30622009569378e-05,
+      "loss": 1.0854,
+      "step": 434
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.43281064565162874,
+      "learning_rate": 8.325358851674641e-05,
+      "loss": 1.1418,
+      "step": 435
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.36804478901968624,
+      "learning_rate": 8.344497607655503e-05,
+      "loss": 1.1754,
+      "step": 436
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3680855445618144,
+      "learning_rate": 8.363636363636364e-05,
+      "loss": 1.1909,
+      "step": 437
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38659927520583764,
+      "learning_rate": 8.382775119617225e-05,
+      "loss": 1.2076,
+      "step": 438
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.341767677446713,
+      "learning_rate": 8.401913875598086e-05,
+      "loss": 1.0057,
+      "step": 439
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4256342728148254,
+      "learning_rate": 8.421052631578948e-05,
+      "loss": 1.2104,
+      "step": 440
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37216536384041526,
+      "learning_rate": 8.440191387559809e-05,
+      "loss": 1.1144,
+      "step": 441
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3791754076001631,
+      "learning_rate": 8.45933014354067e-05,
+      "loss": 1.2319,
+      "step": 442
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.42421700255160355,
+      "learning_rate": 8.478468899521531e-05,
+      "loss": 1.1934,
+      "step": 443
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3678445305043515,
+      "learning_rate": 8.497607655502394e-05,
+      "loss": 1.1886,
+      "step": 444
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4219847133147515,
+      "learning_rate": 8.516746411483254e-05,
+      "loss": 1.1748,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.33966219534041503,
+      "learning_rate": 8.535885167464115e-05,
+      "loss": 1.1416,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3426950270667347,
+      "learning_rate": 8.555023923444977e-05,
+      "loss": 1.2117,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31268235060757094,
+      "learning_rate": 8.574162679425837e-05,
+      "loss": 1.1099,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4283403832969395,
+      "learning_rate": 8.593301435406699e-05,
+      "loss": 1.1204,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.360981174023238,
+      "learning_rate": 8.612440191387561e-05,
+      "loss": 1.1203,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31441146897954253,
+      "learning_rate": 8.631578947368421e-05,
+      "loss": 1.1703,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6665116249955306,
+      "learning_rate": 8.650717703349282e-05,
+      "loss": 1.1965,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.34282354864583103,
+      "learning_rate": 8.669856459330143e-05,
+      "loss": 1.1108,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3353048313959663,
+      "learning_rate": 8.688995215311006e-05,
+      "loss": 1.0893,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3851891541705526,
+      "learning_rate": 8.708133971291866e-05,
+      "loss": 1.065,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40582359225585646,
+      "learning_rate": 8.727272727272727e-05,
+      "loss": 1.1551,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3310055383140587,
+      "learning_rate": 8.74641148325359e-05,
+      "loss": 1.1374,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3697503368052755,
+      "learning_rate": 8.76555023923445e-05,
+      "loss": 1.0852,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3494933513359935,
+      "learning_rate": 8.784688995215311e-05,
+      "loss": 1.1092,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.359799849220072,
+      "learning_rate": 8.803827751196173e-05,
+      "loss": 1.0291,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3692901471865079,
+      "learning_rate": 8.822966507177035e-05,
+      "loss": 1.1516,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.31771285202360866,
+      "learning_rate": 8.842105263157894e-05,
+      "loss": 1.0414,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.39382351019262535,
+      "learning_rate": 8.861244019138757e-05,
+      "loss": 1.1987,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37948154502600623,
+      "learning_rate": 8.880382775119618e-05,
+      "loss": 1.0162,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4454086610740419,
+      "learning_rate": 8.899521531100478e-05,
+      "loss": 1.1686,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3992597027171503,
+      "learning_rate": 8.91866028708134e-05,
+      "loss": 1.1067,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.402787251335542,
+      "learning_rate": 8.937799043062202e-05,
+      "loss": 1.1733,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.3772514735030477,
+      "learning_rate": 8.956937799043063e-05,
+      "loss": 1.1926,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38379331799775773,
+      "learning_rate": 8.976076555023923e-05,
+      "loss": 1.2207,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.37849416076396786,
+      "learning_rate": 8.995215311004786e-05,
+      "loss": 1.158,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40932072449271345,
+      "learning_rate": 9.014354066985647e-05,
+      "loss": 1.124,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34313554427354404,
+      "learning_rate": 9.033492822966507e-05,
+      "loss": 1.2141,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.2960457574671995,
+      "learning_rate": 9.052631578947369e-05,
+      "loss": 1.1269,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3364724543703671,
+      "learning_rate": 9.07177033492823e-05,
+      "loss": 1.0963,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4102740455894671,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 0.9836,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3576390479541009,
+      "learning_rate": 9.110047846889953e-05,
+      "loss": 1.088,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4063740081724684,
+      "learning_rate": 9.129186602870814e-05,
+      "loss": 1.0916,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3866688811135483,
+      "learning_rate": 9.148325358851675e-05,
+      "loss": 1.1582,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35233155736688976,
+      "learning_rate": 9.167464114832537e-05,
+      "loss": 1.2166,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3523742613417453,
+      "learning_rate": 9.186602870813398e-05,
+      "loss": 1.1492,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4091175614367992,
+      "learning_rate": 9.205741626794259e-05,
+      "loss": 1.2015,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.40461705008021104,
+      "learning_rate": 9.224880382775119e-05,
+      "loss": 1.1021,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3555229450892527,
+      "learning_rate": 9.244019138755981e-05,
+      "loss": 1.1677,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39219146946985217,
+      "learning_rate": 9.263157894736843e-05,
+      "loss": 1.1448,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3959881230046531,
+      "learning_rate": 9.282296650717704e-05,
+      "loss": 1.1031,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37437224905160343,
+      "learning_rate": 9.301435406698565e-05,
+      "loss": 1.2172,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34627056702655357,
+      "learning_rate": 9.320574162679426e-05,
+      "loss": 1.0524,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34958866976468206,
+      "learning_rate": 9.339712918660288e-05,
+      "loss": 1.1962,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3592293846050586,
+      "learning_rate": 9.358851674641149e-05,
+      "loss": 1.1292,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4512710733697596,
+      "learning_rate": 9.37799043062201e-05,
+      "loss": 1.1657,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3630522232972779,
+      "learning_rate": 9.397129186602871e-05,
+      "loss": 1.1852,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37647580929450847,
+      "learning_rate": 9.416267942583733e-05,
+      "loss": 1.1968,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4947723206359184,
+      "learning_rate": 9.435406698564594e-05,
+      "loss": 1.1535,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.384481476371926,
+      "learning_rate": 9.454545454545455e-05,
+      "loss": 1.0762,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.30759408641071373,
+      "learning_rate": 9.473684210526316e-05,
+      "loss": 1.2025,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39262382714536653,
+      "learning_rate": 9.492822966507177e-05,
+      "loss": 1.1072,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3707577579487458,
+      "learning_rate": 9.511961722488039e-05,
+      "loss": 1.201,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37259701318014815,
+      "learning_rate": 9.5311004784689e-05,
+      "loss": 1.1808,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3899497483331902,
+      "learning_rate": 9.550239234449761e-05,
+      "loss": 1.155,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3056736695030619,
+      "learning_rate": 9.569377990430622e-05,
+      "loss": 1.1485,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3771828447592661,
+      "learning_rate": 9.588516746411484e-05,
+      "loss": 1.1475,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3512506645855738,
+      "learning_rate": 9.607655502392345e-05,
+      "loss": 1.137,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44787972025939427,
+      "learning_rate": 9.626794258373206e-05,
+      "loss": 1.1137,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.43899468578920847,
+      "learning_rate": 9.645933014354067e-05,
+      "loss": 1.198,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37520028378057024,
+      "learning_rate": 9.66507177033493e-05,
+      "loss": 1.12,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.44437839309325783,
+      "learning_rate": 9.68421052631579e-05,
+      "loss": 1.1266,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3713684887370264,
+      "learning_rate": 9.703349282296651e-05,
+      "loss": 1.0703,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3785759236620265,
+      "learning_rate": 9.722488038277513e-05,
+      "loss": 1.1825,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3949986518868616,
+      "learning_rate": 9.741626794258373e-05,
+      "loss": 1.205,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32820866590924014,
+      "learning_rate": 9.760765550239235e-05,
+      "loss": 1.0473,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3438055422000543,
+      "learning_rate": 9.779904306220096e-05,
+      "loss": 1.2331,
+      "step": 511
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4782013192332416,
+      "learning_rate": 9.799043062200958e-05,
+      "loss": 1.0641,
+      "step": 512
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37127997208806374,
+      "learning_rate": 9.818181818181818e-05,
+      "loss": 1.1557,
+      "step": 513
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37810255782547136,
+      "learning_rate": 9.83732057416268e-05,
+      "loss": 1.1545,
+      "step": 514
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4105466030008024,
+      "learning_rate": 9.856459330143542e-05,
+      "loss": 1.2272,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4517743884986355,
+      "learning_rate": 9.875598086124402e-05,
+      "loss": 1.1229,
+      "step": 516
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.399218104135688,
+      "learning_rate": 9.894736842105263e-05,
+      "loss": 1.1467,
+      "step": 517
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3693460754043359,
+      "learning_rate": 9.913875598086126e-05,
+      "loss": 1.169,
+      "step": 518
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3446333300820591,
+      "learning_rate": 9.933014354066987e-05,
+      "loss": 1.0783,
+      "step": 519
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41719266096581403,
+      "learning_rate": 9.952153110047847e-05,
+      "loss": 1.211,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4139824733210239,
+      "learning_rate": 9.97129186602871e-05,
+      "loss": 1.2271,
+      "step": 521
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3284583647624778,
+      "learning_rate": 9.99043062200957e-05,
+      "loss": 1.1363,
+      "step": 522
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3743628377274405,
+      "learning_rate": 0.0001000956937799043,
+      "loss": 1.1254,
+      "step": 523
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4187480747840408,
+      "learning_rate": 0.00010028708133971292,
+      "loss": 1.1542,
+      "step": 524
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35701214246846424,
+      "learning_rate": 0.00010047846889952153,
+      "loss": 1.1318,
+      "step": 525
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39966360139311247,
+      "learning_rate": 0.00010066985645933015,
+      "loss": 1.0679,
+      "step": 526
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.42408986083219885,
+      "learning_rate": 0.00010086124401913877,
+      "loss": 1.086,
+      "step": 527
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33931985700987544,
+      "learning_rate": 0.00010105263157894738,
+      "loss": 1.1684,
+      "step": 528
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33040989829144524,
+      "learning_rate": 0.00010124401913875599,
+      "loss": 1.1301,
+      "step": 529
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3404311052002135,
+      "learning_rate": 0.00010143540669856459,
+      "loss": 1.0848,
+      "step": 530
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34611637641008364,
+      "learning_rate": 0.0001016267942583732,
+      "loss": 1.1635,
+      "step": 531
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3672822694004909,
+      "learning_rate": 0.00010181818181818181,
+      "loss": 1.0811,
+      "step": 532
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34761252015363225,
+      "learning_rate": 0.00010200956937799044,
+      "loss": 1.1502,
+      "step": 533
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.27503858553543464,
+      "learning_rate": 0.00010220095693779905,
+      "loss": 1.1257,
+      "step": 534
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3368407495501332,
+      "learning_rate": 0.00010239234449760766,
+      "loss": 1.1195,
+      "step": 535
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3448997570516004,
+      "learning_rate": 0.00010258373205741628,
+      "loss": 1.1955,
+      "step": 536
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.4366845539188124,
+      "learning_rate": 0.00010277511961722488,
+      "loss": 1.1175,
+      "step": 537
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35681716286224935,
+      "learning_rate": 0.00010296650717703349,
+      "loss": 1.1577,
+      "step": 538
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3359937686441125,
+      "learning_rate": 0.00010315789473684211,
+      "loss": 1.1318,
+      "step": 539
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3398927803425864,
+      "learning_rate": 0.00010334928229665073,
+      "loss": 1.1278,
+      "step": 540
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38532827109393014,
+      "learning_rate": 0.00010354066985645934,
+      "loss": 1.1273,
+      "step": 541
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3740363062511281,
+      "learning_rate": 0.00010373205741626795,
+      "loss": 1.0253,
+      "step": 542
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34342518179762227,
+      "learning_rate": 0.00010392344497607656,
+      "loss": 1.1462,
+      "step": 543
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3590449087862375,
+      "learning_rate": 0.00010411483253588516,
+      "loss": 1.1392,
+      "step": 544
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37655451333728496,
+      "learning_rate": 0.00010430622009569377,
+      "loss": 1.1218,
+      "step": 545
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3336387102623628,
+      "learning_rate": 0.0001044976076555024,
+      "loss": 1.2268,
+      "step": 546
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38550698369605846,
+      "learning_rate": 0.00010468899521531101,
+      "loss": 1.1752,
+      "step": 547
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.39228288833022,
+      "learning_rate": 0.00010488038277511962,
+      "loss": 1.2562,
+      "step": 548
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.34746795623865406,
+      "learning_rate": 0.00010507177033492824,
+      "loss": 1.2464,
+      "step": 549
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.350081344824654,
+      "learning_rate": 0.00010526315789473685,
+      "loss": 1.0909,
+      "step": 550
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.35258812822080665,
+      "learning_rate": 0.00010545454545454545,
+      "loss": 1.1647,
+      "step": 551
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3894512796451852,
+      "learning_rate": 0.00010564593301435409,
+      "loss": 1.0903,
+      "step": 552
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3560681821041678,
+      "learning_rate": 0.00010583732057416268,
+      "loss": 1.0612,
+      "step": 553
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37166266505359585,
+      "learning_rate": 0.0001060287081339713,
+      "loss": 1.2101,
+      "step": 554
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36530860942069354,
+      "learning_rate": 0.00010622009569377991,
+      "loss": 1.044,
+      "step": 555
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33955376662887854,
+      "learning_rate": 0.00010641148325358852,
+      "loss": 1.093,
+      "step": 556
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31114780875878933,
+      "learning_rate": 0.00010660287081339712,
+      "loss": 1.146,
+      "step": 557
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3676757652032558,
+      "learning_rate": 0.00010679425837320576,
+      "loss": 1.0386,
+      "step": 558
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32774291097340136,
+      "learning_rate": 0.00010698564593301437,
+      "loss": 1.0935,
+      "step": 559
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.33490784632156506,
+      "learning_rate": 0.00010717703349282297,
+      "loss": 1.1692,
+      "step": 560
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3502773499002359,
+      "learning_rate": 0.00010736842105263158,
+      "loss": 1.1153,
+      "step": 561
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36351558175076165,
+      "learning_rate": 0.0001075598086124402,
+      "loss": 1.0359,
+      "step": 562
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.32317295912626365,
+      "learning_rate": 0.00010775119617224881,
+      "loss": 1.0597,
+      "step": 563
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37457465032706805,
+      "learning_rate": 0.0001079425837320574,
+      "loss": 1.2346,
+      "step": 564
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3230784156429463,
+      "learning_rate": 0.00010813397129186604,
+      "loss": 1.0292,
+      "step": 565
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.29938140614694264,
+      "learning_rate": 0.00010832535885167466,
+      "loss": 1.068,
+      "step": 566
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37830783204331137,
+      "learning_rate": 0.00010851674641148326,
+      "loss": 1.1449,
+      "step": 567
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.36419764707025626,
+      "learning_rate": 0.00010870813397129187,
+      "loss": 1.0981,
+      "step": 568
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3748154030309419,
+      "learning_rate": 0.00010889952153110048,
+      "loss": 1.2252,
+      "step": 569
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3368617151990764,
+      "learning_rate": 0.00010909090909090909,
+      "loss": 1.2124,
+      "step": 570
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.37988770907164504,
+      "learning_rate": 0.00010928229665071772,
+      "loss": 1.2043,
+      "step": 571
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.31449121687746323,
+      "learning_rate": 0.00010947368421052633,
+      "loss": 1.1775,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3568124415520435,
+      "learning_rate": 0.00010966507177033494,
+      "loss": 1.1014,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3673211031508777,
+      "learning_rate": 0.00010985645933014354,
+      "loss": 1.2182,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34202187757429514,
+      "learning_rate": 0.00011004784688995215,
+      "loss": 1.0369,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3905415136176411,
+      "learning_rate": 0.00011023923444976077,
+      "loss": 1.1507,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3468401555240628,
+      "learning_rate": 0.00011043062200956938,
+      "loss": 1.1642,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4192072995004005,
+      "learning_rate": 0.000110622009569378,
+      "loss": 1.164,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.379758035778582,
+      "learning_rate": 0.00011081339712918662,
+      "loss": 1.1536,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.41419578240684846,
+      "learning_rate": 0.00011100478468899523,
+      "loss": 1.0945,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3250523681546178,
+      "learning_rate": 0.00011119617224880383,
+      "loss": 1.099,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38315025894656307,
+      "learning_rate": 0.00011138755980861244,
+      "loss": 1.1642,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30382004794249146,
+      "learning_rate": 0.00011157894736842105,
+      "loss": 1.252,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36873995920664776,
+      "learning_rate": 0.00011177033492822968,
+      "loss": 1.1613,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40209606862367175,
+      "learning_rate": 0.00011196172248803829,
+      "loss": 1.2187,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39972092223451644,
+      "learning_rate": 0.0001121531100478469,
+      "loss": 1.1339,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3581792841100828,
+      "learning_rate": 0.00011234449760765551,
+      "loss": 1.1467,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33476559121109767,
+      "learning_rate": 0.00011253588516746411,
+      "loss": 1.1809,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3450568464002908,
+      "learning_rate": 0.00011272727272727272,
+      "loss": 1.1955,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30713056981495374,
+      "learning_rate": 0.00011291866028708134,
+      "loss": 1.153,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40056590609091713,
+      "learning_rate": 0.00011311004784688996,
+      "loss": 1.0824,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32202083458180564,
+      "learning_rate": 0.00011330143540669858,
+      "loss": 1.1039,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.28281909786135145,
+      "learning_rate": 0.00011349282296650719,
+      "loss": 1.1722,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2930441788185507,
+      "learning_rate": 0.0001136842105263158,
+      "loss": 1.1902,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3491373061731604,
+      "learning_rate": 0.0001138755980861244,
+      "loss": 1.1247,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3110622779886572,
+      "learning_rate": 0.00011406698564593301,
+      "loss": 1.2289,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34519065720613423,
+      "learning_rate": 0.00011425837320574164,
+      "loss": 1.1169,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3066625621843041,
+      "learning_rate": 0.00011444976076555025,
+      "loss": 1.1645,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32116731229953854,
+      "learning_rate": 0.00011464114832535886,
+      "loss": 1.0933,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3511568531959789,
+      "learning_rate": 0.00011483253588516747,
+      "loss": 1.1087,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32112239871920967,
+      "learning_rate": 0.00011502392344497607,
+      "loss": 1.1406,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39367325401303266,
+      "learning_rate": 0.00011521531100478468,
+      "loss": 1.1545,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3392107735520774,
+      "learning_rate": 0.0001154066985645933,
+      "loss": 1.1566,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35936783606471423,
+      "learning_rate": 0.00011559808612440192,
+      "loss": 1.1135,
+      "step": 604
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3453223570806925,
+      "learning_rate": 0.00011578947368421053,
+      "loss": 1.1679,
+      "step": 605
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3988207022091826,
+      "learning_rate": 0.00011598086124401915,
+      "loss": 1.1266,
+      "step": 606
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35616581701014133,
+      "learning_rate": 0.00011617224880382776,
+      "loss": 1.0747,
+      "step": 607
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34856430848542924,
+      "learning_rate": 0.00011636363636363636,
+      "loss": 1.1737,
+      "step": 608
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39749502570874873,
+      "learning_rate": 0.00011655502392344497,
+      "loss": 1.1367,
+      "step": 609
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3817892480214725,
+      "learning_rate": 0.00011674641148325361,
+      "loss": 1.1423,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37169774084550616,
+      "learning_rate": 0.00011693779904306221,
+      "loss": 1.2363,
+      "step": 611
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36680842275104286,
+      "learning_rate": 0.00011712918660287082,
+      "loss": 1.1137,
+      "step": 612
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30862259202802894,
+      "learning_rate": 0.00011732057416267943,
+      "loss": 1.1156,
+      "step": 613
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3381253043590406,
+      "learning_rate": 0.00011751196172248804,
+      "loss": 1.1913,
+      "step": 614
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35640646013161875,
+      "learning_rate": 0.00011770334928229664,
+      "loss": 1.2953,
+      "step": 615
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3180351478003401,
+      "learning_rate": 0.00011789473684210525,
+      "loss": 1.1151,
+      "step": 616
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2838039245590443,
+      "learning_rate": 0.0001180861244019139,
+      "loss": 1.1178,
+      "step": 617
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32734113884885613,
+      "learning_rate": 0.00011827751196172249,
+      "loss": 1.1095,
+      "step": 618
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4976044214884747,
+      "learning_rate": 0.0001184688995215311,
+      "loss": 1.1425,
+      "step": 619
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32042584160221055,
+      "learning_rate": 0.00011866028708133972,
+      "loss": 1.0799,
+      "step": 620
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3258958466495425,
+      "learning_rate": 0.00011885167464114833,
+      "loss": 1.221,
+      "step": 621
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3112060174372619,
+      "learning_rate": 0.00011904306220095693,
+      "loss": 1.2115,
+      "step": 622
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30118176636144206,
+      "learning_rate": 0.00011923444976076557,
+      "loss": 1.1361,
+      "step": 623
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3361051300196263,
+      "learning_rate": 0.00011942583732057418,
+      "loss": 1.191,
+      "step": 624
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2931267938868614,
+      "learning_rate": 0.00011961722488038278,
+      "loss": 1.0851,
+      "step": 625
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32113885088697364,
+      "learning_rate": 0.00011980861244019139,
+      "loss": 1.1867,
+      "step": 626
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34101527118946584,
+      "learning_rate": 0.00012,
+      "loss": 1.1836,
+      "step": 627
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.27834089475655605,
+      "learning_rate": 0.00012019138755980862,
+      "loss": 1.2618,
+      "step": 628
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3689617429853991,
+      "learning_rate": 0.00012038277511961724,
+      "loss": 1.1289,
+      "step": 629
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32141982153995574,
+      "learning_rate": 0.00012057416267942585,
+      "loss": 1.245,
+      "step": 630
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3577846092529067,
+      "learning_rate": 0.00012076555023923447,
+      "loss": 1.1214,
+      "step": 631
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3547093306453314,
+      "learning_rate": 0.00012095693779904306,
+      "loss": 1.1151,
+      "step": 632
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3235863986240357,
+      "learning_rate": 0.00012114832535885168,
+      "loss": 1.1431,
+      "step": 633
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3315145079366769,
+      "learning_rate": 0.00012133971291866029,
+      "loss": 1.1589,
+      "step": 634
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.44616643435072917,
+      "learning_rate": 0.0001215311004784689,
+      "loss": 1.1899,
+      "step": 635
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34695172958733533,
+      "learning_rate": 0.00012172248803827753,
+      "loss": 1.2423,
+      "step": 636
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32832699890038897,
+      "learning_rate": 0.00012191387559808614,
+      "loss": 1.0669,
+      "step": 637
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.34250522043183074,
+      "learning_rate": 0.00012210526315789474,
+      "loss": 1.1031,
+      "step": 638
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3536036436534466,
+      "learning_rate": 0.00012229665071770336,
+      "loss": 1.1833,
+      "step": 639
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38952860754869895,
+      "learning_rate": 0.00012248803827751196,
+      "loss": 1.2269,
+      "step": 640
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38372067281414196,
+      "learning_rate": 0.00012267942583732056,
+      "loss": 1.1696,
+      "step": 641
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36604610363956575,
+      "learning_rate": 0.00012287081339712921,
+      "loss": 1.143,
+      "step": 642
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.29457638902628325,
+      "learning_rate": 0.0001230622009569378,
+      "loss": 1.1118,
+      "step": 643
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.31971053516113995,
+      "learning_rate": 0.0001232535885167464,
+      "loss": 1.2257,
+      "step": 644
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3479548829091419,
+      "learning_rate": 0.00012344497607655504,
+      "loss": 1.2003,
+      "step": 645
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39895729607686864,
+      "learning_rate": 0.00012363636363636364,
+      "loss": 1.2362,
+      "step": 646
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32542242184693576,
+      "learning_rate": 0.00012382775119617226,
+      "loss": 1.1924,
+      "step": 647
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37946269414290873,
+      "learning_rate": 0.00012401913875598086,
+      "loss": 1.2259,
+      "step": 648
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37785595963877666,
+      "learning_rate": 0.00012421052631578949,
+      "loss": 1.252,
+      "step": 649
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39908004119966145,
+      "learning_rate": 0.00012440191387559808,
+      "loss": 1.1444,
+      "step": 650
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38865614676807153,
+      "learning_rate": 0.0001245933014354067,
+      "loss": 1.2314,
+      "step": 651
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3204842742106689,
+      "learning_rate": 0.0001247846889952153,
+      "loss": 1.0824,
+      "step": 652
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35369352498295387,
+      "learning_rate": 0.00012497607655502393,
+      "loss": 1.0264,
+      "step": 653
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3305618992525529,
+      "learning_rate": 0.00012516746411483253,
+      "loss": 1.1012,
+      "step": 654
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3757616845139893,
+      "learning_rate": 0.00012535885167464116,
+      "loss": 1.2916,
+      "step": 655
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32567276622705355,
+      "learning_rate": 0.00012555023923444978,
+      "loss": 1.218,
+      "step": 656
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.30320222866051544,
+      "learning_rate": 0.00012574162679425838,
+      "loss": 1.0551,
+      "step": 657
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47041450898052456,
+      "learning_rate": 0.00012593301435406698,
+      "loss": 1.2101,
+      "step": 658
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3265512828583142,
+      "learning_rate": 0.0001261244019138756,
+      "loss": 1.0902,
+      "step": 659
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3095200661644063,
+      "learning_rate": 0.0001263157894736842,
+      "loss": 1.2483,
+      "step": 660
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.39865694866961127,
+      "learning_rate": 0.0001265071770334928,
+      "loss": 1.0507,
+      "step": 661
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.35606148522081404,
+      "learning_rate": 0.00012669856459330146,
+      "loss": 1.0753,
+      "step": 662
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3445720816931114,
+      "learning_rate": 0.00012688995215311006,
+      "loss": 1.143,
+      "step": 663
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3677160012348687,
+      "learning_rate": 0.00012708133971291866,
+      "loss": 1.0608,
+      "step": 664
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2980120877326159,
+      "learning_rate": 0.00012727272727272728,
+      "loss": 1.0872,
+      "step": 665
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.2896118505469009,
+      "learning_rate": 0.00012746411483253588,
+      "loss": 1.1485,
+      "step": 666
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3350125319603418,
+      "learning_rate": 0.0001276555023923445,
+      "loss": 1.1236,
+      "step": 667
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3517268797460554,
+      "learning_rate": 0.00012784688995215313,
+      "loss": 1.1378,
+      "step": 668
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4220707921759215,
+      "learning_rate": 0.00012803827751196173,
+      "loss": 1.1656,
+      "step": 669
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3098050517214006,
+      "learning_rate": 0.00012822966507177036,
+      "loss": 1.0732,
+      "step": 670
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38036416406983276,
+      "learning_rate": 0.00012842105263157895,
+      "loss": 1.1597,
+      "step": 671
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.32201151129472433,
+      "learning_rate": 0.00012861244019138755,
+      "loss": 1.1557,
+      "step": 672
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3477368553208273,
+      "learning_rate": 0.00012880382775119618,
+      "loss": 1.093,
+      "step": 673
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.33206153473346633,
+      "learning_rate": 0.00012899521531100478,
+      "loss": 1.0872,
+      "step": 674
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3797973671348287,
+      "learning_rate": 0.0001291866028708134,
+      "loss": 1.1932,
+      "step": 675
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.38021465107794655,
+      "learning_rate": 0.00012937799043062203,
+      "loss": 1.2037,
+      "step": 676
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40680529142131094,
+      "learning_rate": 0.00012956937799043063,
+      "loss": 1.13,
+      "step": 677
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3662653154346482,
+      "learning_rate": 0.00012976076555023923,
+      "loss": 1.137,
+      "step": 678
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.389523749301837,
+      "learning_rate": 0.00012995215311004785,
+      "loss": 1.1714,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33672031522727297,
+      "learning_rate": 0.00013014354066985645,
+      "loss": 1.08,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3259489924855725,
+      "learning_rate": 0.00013033492822966508,
+      "loss": 1.1592,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4310205006695421,
+      "learning_rate": 0.0001305263157894737,
+      "loss": 1.1125,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3354208256542673,
+      "learning_rate": 0.0001307177033492823,
+      "loss": 1.1612,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33633246177327786,
+      "learning_rate": 0.00013090909090909093,
+      "loss": 1.1075,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31028161222188255,
+      "learning_rate": 0.00013110047846889953,
+      "loss": 1.1152,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4524180110599837,
+      "learning_rate": 0.00013129186602870812,
+      "loss": 1.1259,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3742481866033862,
+      "learning_rate": 0.00013148325358851675,
+      "loss": 1.2494,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3619012526518613,
+      "learning_rate": 0.00013167464114832538,
+      "loss": 1.101,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32394020898287806,
+      "learning_rate": 0.00013186602870813397,
+      "loss": 1.1392,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33391715304609637,
+      "learning_rate": 0.0001320574162679426,
+      "loss": 1.1594,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28785180948649514,
+      "learning_rate": 0.0001322488038277512,
+      "loss": 1.0496,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.38088138786466363,
+      "learning_rate": 0.0001324401913875598,
+      "loss": 1.065,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3208927040153503,
+      "learning_rate": 0.00013263157894736842,
+      "loss": 1.1636,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3033641718971694,
+      "learning_rate": 0.00013282296650717705,
+      "loss": 1.1452,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2948469058966827,
+      "learning_rate": 0.00013301435406698565,
+      "loss": 1.1589,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30477569159510964,
+      "learning_rate": 0.00013320574162679427,
+      "loss": 1.181,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3410300113129024,
+      "learning_rate": 0.00013339712918660287,
+      "loss": 1.1628,
+      "step": 697
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3058561315902832,
+      "learning_rate": 0.0001335885167464115,
+      "loss": 1.084,
+      "step": 698
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34051994364928995,
+      "learning_rate": 0.0001337799043062201,
+      "loss": 1.0576,
+      "step": 699
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3096222099295184,
+      "learning_rate": 0.00013397129186602872,
+      "loss": 1.0741,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28113955153682396,
+      "learning_rate": 0.00013416267942583732,
+      "loss": 1.1019,
+      "step": 701
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2956254577619277,
+      "learning_rate": 0.00013435406698564595,
+      "loss": 1.1044,
+      "step": 702
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31157316700478505,
+      "learning_rate": 0.00013454545454545455,
+      "loss": 1.1694,
+      "step": 703
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3072975606896904,
+      "learning_rate": 0.00013473684210526317,
+      "loss": 1.1256,
+      "step": 704
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35422315692846823,
+      "learning_rate": 0.00013492822966507177,
+      "loss": 1.202,
+      "step": 705
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3039926886966394,
+      "learning_rate": 0.00013511961722488037,
+      "loss": 1.1784,
+      "step": 706
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.272337523028655,
+      "learning_rate": 0.00013531100478468902,
+      "loss": 1.126,
+      "step": 707
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3396634306821353,
+      "learning_rate": 0.00013550239234449762,
+      "loss": 1.1249,
+      "step": 708
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32221861659032364,
+      "learning_rate": 0.00013569377990430622,
+      "loss": 1.1312,
+      "step": 709
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30678629631733856,
+      "learning_rate": 0.00013588516746411485,
+      "loss": 1.1462,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.331787225449244,
+      "learning_rate": 0.00013607655502392344,
+      "loss": 1.1713,
+      "step": 711
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2617883116352453,
+      "learning_rate": 0.00013626794258373204,
+      "loss": 1.1893,
+      "step": 712
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34346153319515627,
+      "learning_rate": 0.0001364593301435407,
+      "loss": 1.1891,
+      "step": 713
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3606280112508664,
+      "learning_rate": 0.0001366507177033493,
+      "loss": 1.3581,
+      "step": 714
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30602332471532506,
+      "learning_rate": 0.0001368421052631579,
+      "loss": 1.2075,
+      "step": 715
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2966643308304905,
+      "learning_rate": 0.00013703349282296652,
+      "loss": 1.1248,
+      "step": 716
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3692677324708085,
+      "learning_rate": 0.00013722488038277512,
+      "loss": 1.1325,
+      "step": 717
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3153436398786279,
+      "learning_rate": 0.00013741626794258374,
+      "loss": 1.141,
+      "step": 718
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2993620796785782,
+      "learning_rate": 0.00013760765550239234,
+      "loss": 1.151,
+      "step": 719
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.360199053324579,
+      "learning_rate": 0.00013779904306220097,
+      "loss": 1.1671,
+      "step": 720
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34616040756962774,
+      "learning_rate": 0.0001379904306220096,
+      "loss": 1.1314,
+      "step": 721
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.32093543405134595,
+      "learning_rate": 0.0001381818181818182,
+      "loss": 1.1017,
+      "step": 722
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3072115942434,
+      "learning_rate": 0.0001383732057416268,
+      "loss": 1.2032,
+      "step": 723
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33680085828062,
+      "learning_rate": 0.00013856459330143542,
+      "loss": 1.191,
+      "step": 724
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.27852291513017413,
+      "learning_rate": 0.00013875598086124402,
+      "loss": 1.2035,
+      "step": 725
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3269080941652961,
+      "learning_rate": 0.00013894736842105264,
+      "loss": 1.1417,
+      "step": 726
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2911551586198448,
+      "learning_rate": 0.00013913875598086127,
+      "loss": 1.2055,
+      "step": 727
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3478754660709439,
+      "learning_rate": 0.00013933014354066987,
+      "loss": 1.1967,
+      "step": 728
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31136552748186935,
+      "learning_rate": 0.00013952153110047846,
+      "loss": 1.1666,
+      "step": 729
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29853571203421375,
+      "learning_rate": 0.0001397129186602871,
+      "loss": 1.1213,
+      "step": 730
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29830710212889877,
+      "learning_rate": 0.0001399043062200957,
+      "loss": 1.2283,
+      "step": 731
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2988658722078936,
+      "learning_rate": 0.00014009569377990431,
+      "loss": 1.2391,
+      "step": 732
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3453871024753651,
+      "learning_rate": 0.00014028708133971294,
+      "loss": 1.2111,
+      "step": 733
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3495768486847103,
+      "learning_rate": 0.00014047846889952154,
+      "loss": 1.1027,
+      "step": 734
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3164024002188871,
+      "learning_rate": 0.00014066985645933016,
+      "loss": 1.1015,
+      "step": 735
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3044139633248399,
+      "learning_rate": 0.00014086124401913876,
+      "loss": 1.0509,
+      "step": 736
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.281079329494108,
+      "learning_rate": 0.00014105263157894736,
+      "loss": 1.0826,
+      "step": 737
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.30636375000054217,
+      "learning_rate": 0.000141244019138756,
+      "loss": 1.2202,
+      "step": 738
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29149180784544115,
+      "learning_rate": 0.0001414354066985646,
+      "loss": 1.1551,
+      "step": 739
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3073819307679817,
+      "learning_rate": 0.0001416267942583732,
+      "loss": 1.2248,
+      "step": 740
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3217985704338287,
+      "learning_rate": 0.00014181818181818184,
+      "loss": 1.2045,
+      "step": 741
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3369269229369114,
+      "learning_rate": 0.00014200956937799044,
+      "loss": 1.1671,
+      "step": 742
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2981155510935532,
+      "learning_rate": 0.00014220095693779904,
+      "loss": 1.1354,
+      "step": 743
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3002935893022973,
+      "learning_rate": 0.00014239234449760766,
+      "loss": 1.0369,
+      "step": 744
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3061234355072447,
+      "learning_rate": 0.00014258373205741626,
+      "loss": 1.0122,
+      "step": 745
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.28759317860073835,
+      "learning_rate": 0.00014277511961722489,
+      "loss": 1.0997,
+      "step": 746
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3064828735905134,
+      "learning_rate": 0.0001429665071770335,
+      "loss": 1.1225,
+      "step": 747
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3604086247045263,
+      "learning_rate": 0.0001431578947368421,
+      "loss": 1.1635,
+      "step": 748
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2914789050629064,
+      "learning_rate": 0.0001433492822966507,
+      "loss": 1.1704,
+      "step": 749
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3105462532363453,
+      "learning_rate": 0.00014354066985645933,
+      "loss": 1.144,
+      "step": 750
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.29242484393022483,
+      "learning_rate": 0.00014373205741626793,
+      "loss": 1.0991,
+      "step": 751
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3009843941043775,
+      "learning_rate": 0.00014392344497607656,
+      "loss": 1.1409,
+      "step": 752
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.35368948528183997,
+      "learning_rate": 0.00014411483253588518,
+      "loss": 1.0839,
+      "step": 753
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2908104621427735,
+      "learning_rate": 0.00014430622009569378,
+      "loss": 1.0997,
+      "step": 754
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2761031765983028,
+      "learning_rate": 0.0001444976076555024,
+      "loss": 1.0389,
+      "step": 755
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34458640320872364,
+      "learning_rate": 0.000144688995215311,
+      "loss": 1.0666,
+      "step": 756
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3426791854461418,
+      "learning_rate": 0.0001448803827751196,
+      "loss": 1.0227,
+      "step": 757
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33484757936373594,
+      "learning_rate": 0.00014507177033492826,
+      "loss": 1.1276,
+      "step": 758
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31113235116750904,
+      "learning_rate": 0.00014526315789473686,
+      "loss": 1.1435,
+      "step": 759
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3187397115174627,
+      "learning_rate": 0.00014545454545454546,
+      "loss": 1.2091,
+      "step": 760
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.322859636752832,
+      "learning_rate": 0.00014564593301435408,
+      "loss": 1.1463,
+      "step": 761
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3054079698873811,
+      "learning_rate": 0.00014583732057416268,
+      "loss": 1.1532,
+      "step": 762
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.37827200804472255,
+      "learning_rate": 0.00014602870813397128,
+      "loss": 1.1536,
+      "step": 763
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33688004627148077,
+      "learning_rate": 0.0001462200956937799,
+      "loss": 1.1633,
+      "step": 764
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3057781806456222,
+      "learning_rate": 0.00014641148325358853,
+      "loss": 1.1336,
+      "step": 765
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3214472678202446,
+      "learning_rate": 0.00014660287081339713,
+      "loss": 1.178,
+      "step": 766
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3615283182183831,
+      "learning_rate": 0.00014679425837320576,
+      "loss": 1.1158,
+      "step": 767
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3147571028922824,
+      "learning_rate": 0.00014698564593301435,
+      "loss": 1.131,
+      "step": 768
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.269304950091198,
+      "learning_rate": 0.00014717703349282298,
+      "loss": 1.1885,
+      "step": 769
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2817004499058875,
+      "learning_rate": 0.00014736842105263158,
+      "loss": 1.0634,
+      "step": 770
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.34677018154047495,
+      "learning_rate": 0.0001475598086124402,
+      "loss": 1.2329,
+      "step": 771
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33187657643162116,
+      "learning_rate": 0.00014775119617224883,
+      "loss": 1.0673,
+      "step": 772
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33397275501257906,
+      "learning_rate": 0.00014794258373205743,
+      "loss": 1.0975,
+      "step": 773
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2977218730080119,
+      "learning_rate": 0.00014813397129186603,
+      "loss": 1.0581,
+      "step": 774
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3480266756123412,
+      "learning_rate": 0.00014832535885167465,
+      "loss": 1.2395,
+      "step": 775
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33681513429762355,
+      "learning_rate": 0.00014851674641148325,
+      "loss": 1.1306,
+      "step": 776
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.31749818370425387,
+      "learning_rate": 0.00014870813397129185,
+      "loss": 1.0901,
+      "step": 777
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.33455138564966774,
+      "learning_rate": 0.0001488995215311005,
+      "loss": 1.2032,
+      "step": 778
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3504419380990198,
+      "learning_rate": 0.0001490909090909091,
+      "loss": 1.0602,
+      "step": 779
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3023880222584541,
+      "learning_rate": 0.0001492822966507177,
+      "loss": 1.1374,
+      "step": 780
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5469036927255182,
+      "learning_rate": 0.00014947368421052633,
+      "loss": 1.1802,
+      "step": 781
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.308842167335779,
+      "learning_rate": 0.00014966507177033493,
+      "loss": 1.0936,
+      "step": 782
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.2986359721179498,
+      "learning_rate": 0.00014985645933014355,
+      "loss": 1.1307,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31664348432490785,
+      "learning_rate": 0.00015004784688995218,
+      "loss": 1.2007,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.265541663614485,
+      "learning_rate": 0.00015023923444976078,
+      "loss": 1.1276,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3204030653032886,
+      "learning_rate": 0.0001504306220095694,
+      "loss": 1.1439,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2783784466928858,
+      "learning_rate": 0.000150622009569378,
+      "loss": 1.2329,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30216230746794037,
+      "learning_rate": 0.0001508133971291866,
+      "loss": 1.1853,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3085281413718923,
+      "learning_rate": 0.00015100478468899522,
+      "loss": 1.1631,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3221875710439296,
+      "learning_rate": 0.00015119617224880382,
+      "loss": 1.0776,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2846073440656778,
+      "learning_rate": 0.00015138755980861245,
+      "loss": 1.1563,
+      "step": 791
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26550401235533877,
+      "learning_rate": 0.00015157894736842108,
+      "loss": 1.0467,
+      "step": 792
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.37881120245858113,
+      "learning_rate": 0.00015177033492822967,
+      "loss": 1.1699,
+      "step": 793
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33594596707038277,
+      "learning_rate": 0.00015196172248803827,
+      "loss": 1.1653,
+      "step": 794
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3233311908095246,
+      "learning_rate": 0.0001521531100478469,
+      "loss": 1.1507,
+      "step": 795
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.267677768320064,
+      "learning_rate": 0.0001523444976076555,
+      "loss": 1.1356,
+      "step": 796
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29484155965355746,
+      "learning_rate": 0.00015253588516746412,
+      "loss": 1.0681,
+      "step": 797
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3284519128368135,
+      "learning_rate": 0.00015272727272727275,
+      "loss": 1.0024,
+      "step": 798
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28701234783478413,
+      "learning_rate": 0.00015291866028708135,
+      "loss": 1.0892,
+      "step": 799
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31864235047065265,
+      "learning_rate": 0.00015311004784688995,
+      "loss": 1.1727,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2745618500591329,
+      "learning_rate": 0.00015330143540669857,
+      "loss": 1.0224,
+      "step": 801
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.30047657670046785,
+      "learning_rate": 0.00015349282296650717,
+      "loss": 1.1017,
+      "step": 802
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2914367942025512,
+      "learning_rate": 0.0001536842105263158,
+      "loss": 1.0163,
+      "step": 803
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2816221100141218,
+      "learning_rate": 0.00015387559808612442,
+      "loss": 1.1372,
+      "step": 804
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34419073166680986,
+      "learning_rate": 0.00015406698564593302,
+      "loss": 1.1991,
+      "step": 805
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2888432316245811,
+      "learning_rate": 0.00015425837320574165,
+      "loss": 1.1627,
+      "step": 806
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3191302154072048,
+      "learning_rate": 0.00015444976076555024,
+      "loss": 1.2458,
+      "step": 807
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2727293598902053,
+      "learning_rate": 0.00015464114832535884,
+      "loss": 1.1085,
+      "step": 808
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3029996217533104,
+      "learning_rate": 0.00015483253588516747,
+      "loss": 1.1319,
+      "step": 809
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2874429714766323,
+      "learning_rate": 0.0001550239234449761,
+      "loss": 1.1123,
+      "step": 810
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2665739546686572,
+      "learning_rate": 0.0001552153110047847,
+      "loss": 1.143,
+      "step": 811
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2958054625397739,
+      "learning_rate": 0.00015540669856459332,
+      "loss": 1.135,
+      "step": 812
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.255412895550101,
+      "learning_rate": 0.00015559808612440192,
+      "loss": 1.1859,
+      "step": 813
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31547097343732156,
+      "learning_rate": 0.00015578947368421052,
+      "loss": 1.1657,
+      "step": 814
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.34867394487181774,
+      "learning_rate": 0.00015598086124401914,
+      "loss": 1.0777,
+      "step": 815
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2921678038171233,
+      "learning_rate": 0.00015617224880382774,
+      "loss": 1.0522,
+      "step": 816
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28919534963089716,
+      "learning_rate": 0.00015636363636363637,
+      "loss": 1.0779,
+      "step": 817
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3114505303709412,
+      "learning_rate": 0.000156555023923445,
+      "loss": 1.1119,
+      "step": 818
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3065682691442617,
+      "learning_rate": 0.0001567464114832536,
+      "loss": 1.1352,
+      "step": 819
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3614199195441891,
+      "learning_rate": 0.00015693779904306222,
+      "loss": 1.1612,
+      "step": 820
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3101608207788147,
+      "learning_rate": 0.00015712918660287082,
+      "loss": 1.252,
+      "step": 821
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2975075722366304,
+      "learning_rate": 0.00015732057416267941,
+      "loss": 1.0687,
+      "step": 822
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2664048730695144,
+      "learning_rate": 0.00015751196172248807,
+      "loss": 1.0832,
+      "step": 823
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2952527620974602,
+      "learning_rate": 0.00015770334928229667,
+      "loss": 1.1082,
+      "step": 824
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3295245234429144,
+      "learning_rate": 0.00015789473684210527,
+      "loss": 1.162,
+      "step": 825
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3102397113238992,
+      "learning_rate": 0.0001580861244019139,
+      "loss": 1.0738,
+      "step": 826
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2693269386909286,
+      "learning_rate": 0.0001582775119617225,
+      "loss": 1.028,
+      "step": 827
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3669225930993825,
+      "learning_rate": 0.0001584688995215311,
+      "loss": 1.15,
+      "step": 828
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29318593683220057,
+      "learning_rate": 0.00015866028708133974,
+      "loss": 1.2408,
+      "step": 829
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2894248048442511,
+      "learning_rate": 0.00015885167464114834,
+      "loss": 1.1752,
+      "step": 830
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29125115224083087,
+      "learning_rate": 0.00015904306220095694,
+      "loss": 0.9574,
+      "step": 831
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.40878890760263803,
+      "learning_rate": 0.00015923444976076556,
+      "loss": 1.2104,
+      "step": 832
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27861032872082103,
+      "learning_rate": 0.00015942583732057416,
+      "loss": 1.1413,
+      "step": 833
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2982611167661862,
+      "learning_rate": 0.0001596172248803828,
+      "loss": 1.2593,
+      "step": 834
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29326214489772795,
+      "learning_rate": 0.0001598086124401914,
+      "loss": 1.18,
+      "step": 835
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27131394008530785,
+      "learning_rate": 0.00016,
+      "loss": 1.1674,
+      "step": 836
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26712034034407034,
+      "learning_rate": 0.00016019138755980864,
+      "loss": 1.0161,
+      "step": 837
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.36369725154573823,
+      "learning_rate": 0.00016038277511961724,
+      "loss": 1.175,
+      "step": 838
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3085671724318983,
+      "learning_rate": 0.00016057416267942584,
+      "loss": 1.1461,
+      "step": 839
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28077141855727894,
+      "learning_rate": 0.00016076555023923446,
+      "loss": 1.0922,
+      "step": 840
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3270351461507469,
+      "learning_rate": 0.00016095693779904306,
+      "loss": 1.0463,
+      "step": 841
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.23981764247780088,
+      "learning_rate": 0.0001611483253588517,
+      "loss": 0.9635,
+      "step": 842
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28201419160149344,
+      "learning_rate": 0.0001613397129186603,
+      "loss": 1.1173,
+      "step": 843
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26889491956006867,
+      "learning_rate": 0.0001615311004784689,
+      "loss": 1.1132,
+      "step": 844
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27688897066555573,
+      "learning_rate": 0.0001617224880382775,
+      "loss": 1.0963,
+      "step": 845
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.24565660227717426,
+      "learning_rate": 0.00016191387559808614,
+      "loss": 1.0694,
+      "step": 846
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28311675225629357,
+      "learning_rate": 0.00016210526315789473,
+      "loss": 1.0727,
+      "step": 847
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25275425247450756,
+      "learning_rate": 0.00016229665071770336,
+      "loss": 1.1726,
+      "step": 848
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3073349091629191,
+      "learning_rate": 0.00016248803827751199,
+      "loss": 1.237,
+      "step": 849
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3698105893782691,
+      "learning_rate": 0.00016267942583732058,
+      "loss": 1.1529,
+      "step": 850
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3066504421764291,
+      "learning_rate": 0.00016287081339712918,
+      "loss": 1.1875,
+      "step": 851
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2853734077261547,
+      "learning_rate": 0.0001630622009569378,
+      "loss": 1.1232,
+      "step": 852
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27498683022213083,
+      "learning_rate": 0.0001632535885167464,
+      "loss": 1.2345,
+      "step": 853
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26436373680139863,
+      "learning_rate": 0.00016344497607655503,
+      "loss": 1.1443,
+      "step": 854
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29039546604591765,
+      "learning_rate": 0.00016363636363636366,
+      "loss": 1.1451,
+      "step": 855
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2845332734411919,
+      "learning_rate": 0.00016382775119617226,
+      "loss": 1.1658,
+      "step": 856
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3118984941168386,
+      "learning_rate": 0.00016401913875598088,
+      "loss": 1.133,
+      "step": 857
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2910324342007811,
+      "learning_rate": 0.00016421052631578948,
+      "loss": 1.138,
+      "step": 858
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3067211385198509,
+      "learning_rate": 0.00016440191387559808,
+      "loss": 1.1517,
+      "step": 859
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.290740982507053,
+      "learning_rate": 0.0001645933014354067,
+      "loss": 1.0561,
+      "step": 860
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3144516777697552,
+      "learning_rate": 0.0001647846889952153,
+      "loss": 1.0661,
+      "step": 861
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2970636821654555,
+      "learning_rate": 0.00016497607655502393,
+      "loss": 1.1634,
+      "step": 862
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3146333025319219,
+      "learning_rate": 0.00016516746411483256,
+      "loss": 1.0652,
+      "step": 863
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2644767264588937,
+      "learning_rate": 0.00016535885167464116,
+      "loss": 1.1516,
+      "step": 864
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3006840203451009,
+      "learning_rate": 0.00016555023923444975,
+      "loss": 1.1175,
+      "step": 865
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2809420339644184,
+      "learning_rate": 0.00016574162679425838,
+      "loss": 1.1057,
+      "step": 866
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3769059520574524,
+      "learning_rate": 0.00016593301435406698,
+      "loss": 1.1713,
+      "step": 867
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.33622542833176833,
+      "learning_rate": 0.0001661244019138756,
+      "loss": 1.2223,
+      "step": 868
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2715119578007493,
+      "learning_rate": 0.00016631578947368423,
+      "loss": 1.0926,
+      "step": 869
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2788006611781337,
+      "learning_rate": 0.00016650717703349283,
+      "loss": 1.0598,
+      "step": 870
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29918887354582546,
+      "learning_rate": 0.00016669856459330145,
+      "loss": 1.1363,
+      "step": 871
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.27116956033088324,
+      "learning_rate": 0.00016688995215311005,
+      "loss": 1.1357,
+      "step": 872
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3651972053932287,
+      "learning_rate": 0.00016708133971291865,
+      "loss": 1.1862,
+      "step": 873
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2941314020229377,
+      "learning_rate": 0.00016727272727272728,
+      "loss": 1.229,
+      "step": 874
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29667386462622886,
+      "learning_rate": 0.0001674641148325359,
+      "loss": 1.105,
+      "step": 875
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2888327808151174,
+      "learning_rate": 0.0001676555023923445,
+      "loss": 1.1328,
+      "step": 876
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31564538771648376,
+      "learning_rate": 0.00016784688995215313,
+      "loss": 1.2407,
+      "step": 877
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.31419100780147885,
+      "learning_rate": 0.00016803827751196173,
+      "loss": 1.0472,
+      "step": 878
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.26956520480679047,
+      "learning_rate": 0.00016822966507177033,
+      "loss": 1.1524,
+      "step": 879
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2726391902939466,
+      "learning_rate": 0.00016842105263157895,
+      "loss": 1.11,
+      "step": 880
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2993169367221595,
+      "learning_rate": 0.00016861244019138758,
+      "loss": 1.2059,
+      "step": 881
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2936290798225595,
+      "learning_rate": 0.00016880382775119618,
+      "loss": 1.1026,
+      "step": 882
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.25728140759420537,
+      "learning_rate": 0.0001689952153110048,
+      "loss": 1.136,
+      "step": 883
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2659884049250215,
+      "learning_rate": 0.0001691866028708134,
+      "loss": 1.0311,
+      "step": 884
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.29849696827544475,
+      "learning_rate": 0.00016937799043062203,
+      "loss": 1.0995,
+      "step": 885
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.28395796526200556,
+      "learning_rate": 0.00016956937799043062,
+      "loss": 1.0948,
+      "step": 886
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3445355283030851,
+      "learning_rate": 0.00016976076555023925,
+      "loss": 1.1103,
+      "step": 887
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.2538735838657434,
+      "learning_rate": 0.00016995215311004788,
+      "loss": 1.0752,
+      "step": 888
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3026322031952384,
+      "learning_rate": 0.00017014354066985647,
+      "loss": 1.1359,
+      "step": 889
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3200296529545863,
+      "learning_rate": 0.00017033492822966507,
+      "loss": 1.1851,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29333134950174405,
+      "learning_rate": 0.0001705263157894737,
+      "loss": 1.0547,
+      "step": 891
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3189167751421413,
+      "learning_rate": 0.0001707177033492823,
+      "loss": 1.0605,
+      "step": 892
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28038685627245685,
+      "learning_rate": 0.0001709090909090909,
+      "loss": 1.1087,
+      "step": 893
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27600976689482803,
+      "learning_rate": 0.00017110047846889955,
+      "loss": 1.1681,
+      "step": 894
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.30930194872855393,
+      "learning_rate": 0.00017129186602870815,
+      "loss": 1.1325,
+      "step": 895
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31300532276472537,
+      "learning_rate": 0.00017148325358851675,
+      "loss": 1.1002,
+      "step": 896
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2991016495421357,
+      "learning_rate": 0.00017167464114832537,
+      "loss": 1.076,
+      "step": 897
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31594108017658423,
+      "learning_rate": 0.00017186602870813397,
+      "loss": 1.136,
+      "step": 898
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27497095491638146,
+      "learning_rate": 0.0001720574162679426,
+      "loss": 1.2323,
+      "step": 899
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.310962024257193,
+      "learning_rate": 0.00017224880382775122,
+      "loss": 1.1398,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29697512814616595,
+      "learning_rate": 0.00017244019138755982,
+      "loss": 1.1342,
+      "step": 901
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.239946301235031,
+      "learning_rate": 0.00017263157894736842,
+      "loss": 1.2081,
+      "step": 902
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2874130680609666,
+      "learning_rate": 0.00017282296650717705,
+      "loss": 1.125,
+      "step": 903
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3210023421862061,
+      "learning_rate": 0.00017301435406698564,
+      "loss": 1.2268,
+      "step": 904
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25985019789372976,
+      "learning_rate": 0.00017320574162679427,
+      "loss": 1.0704,
+      "step": 905
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28932579761775323,
+      "learning_rate": 0.00017339712918660287,
+      "loss": 1.1939,
+      "step": 906
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2984436701321717,
+      "learning_rate": 0.0001735885167464115,
+      "loss": 1.1441,
+      "step": 907
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33279429925895665,
+      "learning_rate": 0.00017377990430622012,
+      "loss": 1.2299,
+      "step": 908
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28785481315035893,
+      "learning_rate": 0.00017397129186602872,
+      "loss": 1.118,
+      "step": 909
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31655385538112546,
+      "learning_rate": 0.00017416267942583732,
+      "loss": 1.189,
+      "step": 910
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3038855880357351,
+      "learning_rate": 0.00017435406698564594,
+      "loss": 1.0654,
+      "step": 911
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3042729440177485,
+      "learning_rate": 0.00017454545454545454,
+      "loss": 1.1553,
+      "step": 912
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.266680706575244,
+      "learning_rate": 0.00017473684210526317,
+      "loss": 1.1006,
+      "step": 913
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2741425104907168,
+      "learning_rate": 0.0001749282296650718,
+      "loss": 1.2319,
+      "step": 914
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2520129822636353,
+      "learning_rate": 0.0001751196172248804,
+      "loss": 1.1394,
+      "step": 915
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28607836069753895,
+      "learning_rate": 0.000175311004784689,
+      "loss": 1.2108,
+      "step": 916
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28836296745411716,
+      "learning_rate": 0.00017550239234449762,
+      "loss": 1.0772,
+      "step": 917
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26291170956841414,
+      "learning_rate": 0.00017569377990430622,
+      "loss": 1.1532,
+      "step": 918
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27624746568071396,
+      "learning_rate": 0.00017588516746411484,
+      "loss": 1.1178,
+      "step": 919
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28200386776822395,
+      "learning_rate": 0.00017607655502392347,
+      "loss": 1.1105,
+      "step": 920
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25462518315632554,
+      "learning_rate": 0.00017626794258373207,
+      "loss": 1.0717,
+      "step": 921
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27932944411599797,
+      "learning_rate": 0.0001764593301435407,
+      "loss": 1.2486,
+      "step": 922
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29462379215808215,
+      "learning_rate": 0.0001766507177033493,
+      "loss": 1.226,
+      "step": 923
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2741976731865599,
+      "learning_rate": 0.0001768421052631579,
+      "loss": 1.1797,
+      "step": 924
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2532434659032646,
+      "learning_rate": 0.00017703349282296652,
+      "loss": 1.0828,
+      "step": 925
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.329346060797211,
+      "learning_rate": 0.00017722488038277514,
+      "loss": 1.1125,
+      "step": 926
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2644644824352827,
+      "learning_rate": 0.00017741626794258374,
+      "loss": 1.1048,
+      "step": 927
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2617940651450908,
+      "learning_rate": 0.00017760765550239237,
+      "loss": 1.2178,
+      "step": 928
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29432756373678265,
+      "learning_rate": 0.00017779904306220096,
+      "loss": 1.1336,
+      "step": 929
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28911304731696175,
+      "learning_rate": 0.00017799043062200956,
+      "loss": 1.1578,
+      "step": 930
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3006870934673598,
+      "learning_rate": 0.0001781818181818182,
+      "loss": 1.0588,
+      "step": 931
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31210608325092193,
+      "learning_rate": 0.0001783732057416268,
+      "loss": 1.2426,
+      "step": 932
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27626145357478726,
+      "learning_rate": 0.0001785645933014354,
+      "loss": 1.1609,
+      "step": 933
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2683905399507039,
+      "learning_rate": 0.00017875598086124404,
+      "loss": 1.1457,
+      "step": 934
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2661353870666551,
+      "learning_rate": 0.00017894736842105264,
+      "loss": 1.2095,
+      "step": 935
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.33062559297582395,
+      "learning_rate": 0.00017913875598086126,
+      "loss": 1.2396,
+      "step": 936
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26950737804952357,
+      "learning_rate": 0.00017933014354066986,
+      "loss": 1.1983,
+      "step": 937
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29499843784362234,
+      "learning_rate": 0.00017952153110047846,
+      "loss": 1.2442,
+      "step": 938
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31186904072609634,
+      "learning_rate": 0.00017971291866028709,
+      "loss": 1.1494,
+      "step": 939
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2514154775647367,
+      "learning_rate": 0.0001799043062200957,
+      "loss": 1.1061,
+      "step": 940
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28595401946483395,
+      "learning_rate": 0.0001800956937799043,
+      "loss": 1.0322,
+      "step": 941
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32459069445525873,
+      "learning_rate": 0.00018028708133971294,
+      "loss": 1.2007,
+      "step": 942
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31789984803696647,
+      "learning_rate": 0.00018047846889952154,
+      "loss": 1.1057,
+      "step": 943
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2893543986536651,
+      "learning_rate": 0.00018066985645933013,
+      "loss": 1.1108,
+      "step": 944
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.225754938363265,
+      "learning_rate": 0.00018086124401913876,
+      "loss": 1.0842,
+      "step": 945
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3473860341063463,
+      "learning_rate": 0.00018105263157894739,
+      "loss": 1.0824,
+      "step": 946
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2922950981615233,
+      "learning_rate": 0.00018124401913875598,
+      "loss": 1.1143,
+      "step": 947
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29161352434420446,
+      "learning_rate": 0.0001814354066985646,
+      "loss": 1.2325,
+      "step": 948
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29481985803408,
+      "learning_rate": 0.0001816267942583732,
+      "loss": 1.1527,
+      "step": 949
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2874729386092549,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 1.1476,
+      "step": 950
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.24921692256090058,
+      "learning_rate": 0.00018200956937799043,
+      "loss": 1.1567,
+      "step": 951
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.31742487463024494,
+      "learning_rate": 0.00018220095693779906,
+      "loss": 1.1013,
+      "step": 952
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27456172247325683,
+      "learning_rate": 0.00018239234449760766,
+      "loss": 1.2069,
+      "step": 953
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28859064993450634,
+      "learning_rate": 0.00018258373205741628,
+      "loss": 1.2123,
+      "step": 954
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2750397123362856,
+      "learning_rate": 0.00018277511961722488,
+      "loss": 1.1231,
+      "step": 955
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.23915748062608722,
+      "learning_rate": 0.0001829665071770335,
+      "loss": 1.1612,
+      "step": 956
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2722479832588104,
+      "learning_rate": 0.0001831578947368421,
+      "loss": 1.1897,
+      "step": 957
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2863352675014705,
+      "learning_rate": 0.00018334928229665073,
+      "loss": 1.1062,
+      "step": 958
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25258807482282203,
+      "learning_rate": 0.00018354066985645936,
+      "loss": 1.1574,
+      "step": 959
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27455220282180104,
+      "learning_rate": 0.00018373205741626796,
+      "loss": 1.0886,
+      "step": 960
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2693559937731345,
+      "learning_rate": 0.00018392344497607656,
+      "loss": 1.1873,
+      "step": 961
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2487980816971801,
+      "learning_rate": 0.00018411483253588518,
+      "loss": 1.2245,
+      "step": 962
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6732577941816555,
+      "learning_rate": 0.00018430622009569378,
+      "loss": 1.1813,
+      "step": 963
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2912759304052633,
+      "learning_rate": 0.00018449760765550238,
+      "loss": 1.168,
+      "step": 964
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.26989856763778836,
+      "learning_rate": 0.00018468899521531103,
+      "loss": 1.0137,
+      "step": 965
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25602835842131616,
+      "learning_rate": 0.00018488038277511963,
+      "loss": 1.1879,
+      "step": 966
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.25725078226468107,
+      "learning_rate": 0.00018507177033492823,
+      "loss": 1.174,
+      "step": 967
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27889203556658276,
+      "learning_rate": 0.00018526315789473685,
+      "loss": 1.2013,
+      "step": 968
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28462011286220296,
+      "learning_rate": 0.00018545454545454545,
+      "loss": 1.0878,
+      "step": 969
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27682759364760257,
+      "learning_rate": 0.00018564593301435408,
+      "loss": 1.085,
+      "step": 970
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28136179421463786,
+      "learning_rate": 0.0001858373205741627,
+      "loss": 1.1807,
+      "step": 971
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2506017584700625,
+      "learning_rate": 0.0001860287081339713,
+      "loss": 1.1538,
+      "step": 972
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2681898339952538,
+      "learning_rate": 0.00018622009569377993,
+      "loss": 1.0615,
+      "step": 973
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2838246571307257,
+      "learning_rate": 0.00018641148325358853,
+      "loss": 1.1778,
+      "step": 974
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2758038504041395,
+      "learning_rate": 0.00018660287081339713,
+      "loss": 1.1038,
+      "step": 975
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28266068816982276,
+      "learning_rate": 0.00018679425837320575,
+      "loss": 1.1487,
+      "step": 976
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2655825547541941,
+      "learning_rate": 0.00018698564593301435,
+      "loss": 1.0846,
+      "step": 977
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2750864417199089,
+      "learning_rate": 0.00018717703349282298,
+      "loss": 1.0925,
+      "step": 978
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.28328763891237363,
+      "learning_rate": 0.0001873684210526316,
+      "loss": 1.1602,
+      "step": 979
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.274427495879147,
+      "learning_rate": 0.0001875598086124402,
+      "loss": 1.0184,
+      "step": 980
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.29677822769592865,
+      "learning_rate": 0.0001877511961722488,
+      "loss": 1.1316,
+      "step": 981
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.35675044865453487,
+      "learning_rate": 0.00018794258373205743,
+      "loss": 1.1299,
+      "step": 982
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27471990871455726,
+      "learning_rate": 0.00018813397129186602,
+      "loss": 1.0977,
+      "step": 983
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.32677576558264015,
+      "learning_rate": 0.00018832535885167465,
+      "loss": 1.0416,
+      "step": 984
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3449420887466517,
+      "learning_rate": 0.00018851674641148328,
+      "loss": 1.218,
+      "step": 985
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3187055721961639,
+      "learning_rate": 0.00018870813397129187,
+      "loss": 1.1091,
+      "step": 986
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3143792697319127,
+      "learning_rate": 0.0001888995215311005,
+      "loss": 1.067,
+      "step": 987
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2742909947428014,
+      "learning_rate": 0.0001890909090909091,
+      "loss": 1.225,
+      "step": 988
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.27319677319302543,
+      "learning_rate": 0.0001892822966507177,
+      "loss": 1.1487,
+      "step": 989
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2758157497549949,
+      "learning_rate": 0.00018947368421052632,
+      "loss": 1.1109,
+      "step": 990
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.2574079506381213,
+      "learning_rate": 0.00018966507177033495,
+      "loss": 1.1476,
+      "step": 991
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.332702187603211,
+      "learning_rate": 0.00018985645933014355,
+      "loss": 1.0896,
+      "step": 992
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2861721583962,
+      "learning_rate": 0.00019004784688995217,
+      "loss": 1.2234,
+      "step": 993
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2467936267051518,
+      "learning_rate": 0.00019023923444976077,
+      "loss": 1.1656,
+      "step": 994
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28781558903434595,
+      "learning_rate": 0.00019043062200956937,
+      "loss": 1.1853,
+      "step": 995
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2916455357271407,
+      "learning_rate": 0.000190622009569378,
+      "loss": 1.0269,
+      "step": 996
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3201870144576391,
+      "learning_rate": 0.00019081339712918662,
+      "loss": 1.1852,
+      "step": 997
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27545254213477577,
+      "learning_rate": 0.00019100478468899522,
+      "loss": 1.0957,
+      "step": 998
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2826496819385951,
+      "learning_rate": 0.00019119617224880385,
+      "loss": 1.2255,
+      "step": 999
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2967102485192698,
+      "learning_rate": 0.00019138755980861245,
+      "loss": 1.1536,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29117608778714893,
+      "learning_rate": 0.00019157894736842104,
+      "loss": 1.0878,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28851304804169287,
+      "learning_rate": 0.00019177033492822967,
+      "loss": 1.0898,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27111717804566754,
+      "learning_rate": 0.00019196172248803827,
+      "loss": 1.2214,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29632228590140464,
+      "learning_rate": 0.0001921531100478469,
+      "loss": 1.1534,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30166486227944156,
+      "learning_rate": 0.00019234449760765552,
+      "loss": 1.1784,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.261168294050402,
+      "learning_rate": 0.00019253588516746412,
+      "loss": 1.2274,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2696524388115216,
+      "learning_rate": 0.00019272727272727274,
+      "loss": 1.1256,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30883168940001077,
+      "learning_rate": 0.00019291866028708134,
+      "loss": 1.1804,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2901725454324794,
+      "learning_rate": 0.00019311004784688994,
+      "loss": 1.0656,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.30050679633218647,
+      "learning_rate": 0.0001933014354066986,
+      "loss": 1.1217,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2763711001518656,
+      "learning_rate": 0.0001934928229665072,
+      "loss": 1.2114,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2676109407157463,
+      "learning_rate": 0.0001936842105263158,
+      "loss": 1.0474,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2747480845011328,
+      "learning_rate": 0.00019387559808612442,
+      "loss": 1.038,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24960295337688276,
+      "learning_rate": 0.00019406698564593302,
+      "loss": 1.0625,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2721591800223072,
+      "learning_rate": 0.00019425837320574162,
+      "loss": 1.1327,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2877329511310855,
+      "learning_rate": 0.00019444976076555027,
+      "loss": 1.228,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2568028077694964,
+      "learning_rate": 0.00019464114832535887,
+      "loss": 1.0683,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2678405294971607,
+      "learning_rate": 0.00019483253588516747,
+      "loss": 1.1125,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2963652522200652,
+      "learning_rate": 0.0001950239234449761,
+      "loss": 1.0905,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26009393679319537,
+      "learning_rate": 0.0001952153110047847,
+      "loss": 1.1036,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3049720818580699,
+      "learning_rate": 0.00019540669856459332,
+      "loss": 1.1964,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3050130613963167,
+      "learning_rate": 0.00019559808612440191,
+      "loss": 1.1293,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24297369971258104,
+      "learning_rate": 0.00019578947368421054,
+      "loss": 1.1143,
+      "step": 1023
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24077286684290172,
+      "learning_rate": 0.00019598086124401917,
+      "loss": 1.0764,
+      "step": 1024
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3113100418888948,
+      "learning_rate": 0.00019617224880382777,
+      "loss": 1.1246,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2784731985247703,
+      "learning_rate": 0.00019636363636363636,
+      "loss": 1.0998,
+      "step": 1026
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2542533680624268,
+      "learning_rate": 0.000196555023923445,
+      "loss": 1.114,
+      "step": 1027
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28332309977048276,
+      "learning_rate": 0.0001967464114832536,
+      "loss": 1.1719,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25261282572279636,
+      "learning_rate": 0.00019693779904306221,
+      "loss": 1.1069,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28908512950153364,
+      "learning_rate": 0.00019712918660287084,
+      "loss": 1.0939,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2624681443069945,
+      "learning_rate": 0.00019732057416267944,
+      "loss": 1.1426,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26954809036931093,
+      "learning_rate": 0.00019751196172248804,
+      "loss": 1.0777,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2921435580998635,
+      "learning_rate": 0.00019770334928229666,
+      "loss": 1.0915,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.21257005701595452,
+      "learning_rate": 0.00019789473684210526,
+      "loss": 1.1055,
+      "step": 1034
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27523674720420943,
+      "learning_rate": 0.0001980861244019139,
+      "loss": 1.0515,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26415508019617007,
+      "learning_rate": 0.0001982775119617225,
+      "loss": 1.0515,
+      "step": 1036
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25592610307218705,
+      "learning_rate": 0.0001984688995215311,
+      "loss": 1.1751,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25430310175648296,
+      "learning_rate": 0.00019866028708133974,
+      "loss": 1.123,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2861528947212422,
+      "learning_rate": 0.00019885167464114834,
+      "loss": 1.0859,
+      "step": 1039
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2738046774076065,
+      "learning_rate": 0.00019904306220095693,
+      "loss": 1.1405,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2726143048105954,
+      "learning_rate": 0.00019923444976076556,
+      "loss": 1.264,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27872152562297303,
+      "learning_rate": 0.0001994258373205742,
+      "loss": 1.3155,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2852218650666301,
+      "learning_rate": 0.00019961722488038279,
+      "loss": 1.1655,
+      "step": 1043
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23588395214808744,
+      "learning_rate": 0.0001998086124401914,
+      "loss": 1.0397,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2666368869674148,
+      "learning_rate": 0.0002,
+      "loss": 1.1416,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2930022628207633,
+      "learning_rate": 0.00019999999874871857,
+      "loss": 1.0405,
+      "step": 1046
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2710774549293637,
+      "learning_rate": 0.00019999999499487433,
+      "loss": 1.1506,
+      "step": 1047
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25385995300238745,
+      "learning_rate": 0.00019999998873846737,
+      "loss": 1.267,
+      "step": 1048
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27774997441775784,
+      "learning_rate": 0.00019999997997949785,
+      "loss": 1.1939,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26076832391011084,
+      "learning_rate": 0.00019999996871796597,
+      "loss": 1.1117,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.235868498103916,
+      "learning_rate": 0.00019999995495387202,
+      "loss": 1.1701,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24077986955464514,
+      "learning_rate": 0.00019999993868721638,
+      "loss": 1.137,
+      "step": 1052
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2512166776788439,
+      "learning_rate": 0.0001999999199179994,
+      "loss": 1.1414,
+      "step": 1053
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2654407191870313,
+      "learning_rate": 0.00019999989864622159,
+      "loss": 1.0333,
+      "step": 1054
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2642901952276395,
+      "learning_rate": 0.00019999987487188348,
+      "loss": 1.1385,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23723222653673273,
+      "learning_rate": 0.00019999984859498562,
+      "loss": 1.1103,
+      "step": 1056
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2576508658140534,
+      "learning_rate": 0.00019999981981552872,
+      "loss": 1.0641,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2512702002371694,
+      "learning_rate": 0.00019999978853351346,
+      "loss": 1.1742,
+      "step": 1058
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2544113883838849,
+      "learning_rate": 0.0001999997547489407,
+      "loss": 1.249,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29453181610522905,
+      "learning_rate": 0.00019999971846181117,
+      "loss": 1.0817,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3046660451518799,
+      "learning_rate": 0.00019999967967212587,
+      "loss": 1.1202,
+      "step": 1061
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2711240525076058,
+      "learning_rate": 0.0001999996383798857,
+      "loss": 1.1488,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26069781523913904,
+      "learning_rate": 0.0001999995945850918,
+      "loss": 1.1204,
+      "step": 1063
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2759936162347398,
+      "learning_rate": 0.00019999954828774514,
+      "loss": 1.2418,
+      "step": 1064
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.32559445650786323,
+      "learning_rate": 0.00019999949948784696,
+      "loss": 1.0396,
+      "step": 1065
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29446236701732303,
+      "learning_rate": 0.00019999944818539843,
+      "loss": 0.9828,
+      "step": 1066
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2512414839667447,
+      "learning_rate": 0.00019999939438040092,
+      "loss": 1.0964,
+      "step": 1067
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2522166834487399,
+      "learning_rate": 0.00019999933807285567,
+      "loss": 1.1991,
+      "step": 1068
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2776488352071124,
+      "learning_rate": 0.00019999927926276417,
+      "loss": 1.2357,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3107756422517976,
+      "learning_rate": 0.00019999921795012783,
+      "loss": 1.1875,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23140904026513692,
+      "learning_rate": 0.00019999915413494823,
+      "loss": 1.0612,
+      "step": 1071
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28830876945790945,
+      "learning_rate": 0.00019999908781722693,
+      "loss": 1.0622,
+      "step": 1072
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24641652710238304,
+      "learning_rate": 0.00019999901899696564,
+      "loss": 1.1553,
+      "step": 1073
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3285726477920543,
+      "learning_rate": 0.00019999894767416603,
+      "loss": 1.1287,
+      "step": 1074
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2868313621923491,
+      "learning_rate": 0.00019999887384882992,
+      "loss": 1.1679,
+      "step": 1075
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2888935086026084,
+      "learning_rate": 0.00019999879752095914,
+      "loss": 1.078,
+      "step": 1076
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.25581751198117825,
+      "learning_rate": 0.0001999987186905556,
+      "loss": 1.1637,
+      "step": 1077
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28668728467855203,
+      "learning_rate": 0.0001999986373576213,
+      "loss": 1.1557,
+      "step": 1078
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2858855815822476,
+      "learning_rate": 0.00019999855352215824,
+      "loss": 1.1543,
+      "step": 1079
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.24866679351848656,
+      "learning_rate": 0.0001999984671841685,
+      "loss": 1.088,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2802988000108613,
+      "learning_rate": 0.00019999837834365432,
+      "loss": 1.1045,
+      "step": 1081
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28290564790646,
+      "learning_rate": 0.00019999828700061786,
+      "loss": 1.1013,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2670174309559056,
+      "learning_rate": 0.0001999981931550614,
+      "loss": 1.0202,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.29742334914408336,
+      "learning_rate": 0.00019999809680698734,
+      "loss": 1.1634,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2613603430268145,
+      "learning_rate": 0.00019999799795639804,
+      "loss": 1.1906,
+      "step": 1085
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2368844788947555,
+      "learning_rate": 0.000199997896603296,
+      "loss": 1.1789,
+      "step": 1086
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28495588713062425,
+      "learning_rate": 0.00019999779274768376,
+      "loss": 1.1759,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2822715528001476,
+      "learning_rate": 0.0001999976863895639,
+      "loss": 1.0508,
+      "step": 1088
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28574997489486803,
+      "learning_rate": 0.0001999975775289391,
+      "loss": 1.1224,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.261898017632014,
+      "learning_rate": 0.00019999746616581208,
+      "loss": 1.1035,
+      "step": 1090
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28859851706983464,
+      "learning_rate": 0.00019999735230018562,
+      "loss": 1.1726,
+      "step": 1091
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26017489671317706,
+      "learning_rate": 0.00019999723593206256,
+      "loss": 1.0777,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.2808346316808804,
+      "learning_rate": 0.00019999711706144584,
+      "loss": 1.1169,
+      "step": 1093
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.26961618081539596,
+      "learning_rate": 0.0001999969956883384,
+      "loss": 1.1525,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.23998423085029977,
+      "learning_rate": 0.0001999968718127433,
+      "loss": 1.1088,
+      "step": 1095
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.27133623294565795,
+      "learning_rate": 0.00019999674543466368,
+      "loss": 1.0187,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.28888071684005906,
+      "learning_rate": 0.00019999661655410261,
+      "loss": 1.0476,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26874358976360063,
+      "learning_rate": 0.0001999964851710634,
+      "loss": 1.1429,
+      "step": 1098
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2803133560421145,
+      "learning_rate": 0.0001999963512855493,
+      "loss": 1.2504,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2913063287806824,
+      "learning_rate": 0.00019999621489756364,
+      "loss": 1.1604,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27440490495841235,
+      "learning_rate": 0.00019999607600710984,
+      "loss": 1.1517,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.277572309820751,
+      "learning_rate": 0.00019999593461419144,
+      "loss": 1.0957,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3075924298382781,
+      "learning_rate": 0.0001999957907188119,
+      "loss": 1.142,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23985413219751897,
+      "learning_rate": 0.00019999564432097487,
+      "loss": 1.1932,
+      "step": 1104
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2408338302884486,
+      "learning_rate": 0.00019999549542068395,
+      "loss": 1.0735,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26874856387294116,
+      "learning_rate": 0.00019999534401794297,
+      "loss": 1.1553,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3014584853984502,
+      "learning_rate": 0.00019999519011275566,
+      "loss": 1.1655,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2843833242046219,
+      "learning_rate": 0.00019999503370512583,
+      "loss": 1.1877,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2512315616335756,
+      "learning_rate": 0.00019999487479505746,
+      "loss": 1.2,
+      "step": 1109
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3854687733857706,
+      "learning_rate": 0.00019999471338255452,
+      "loss": 1.1755,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2957638364283729,
+      "learning_rate": 0.00019999454946762103,
+      "loss": 1.1496,
+      "step": 1111
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2866505879252708,
+      "learning_rate": 0.00019999438305026108,
+      "loss": 0.9671,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31100005319009444,
+      "learning_rate": 0.00019999421413047886,
+      "loss": 1.1924,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2736868129625665,
+      "learning_rate": 0.00019999404270827856,
+      "loss": 1.0565,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3082559508155182,
+      "learning_rate": 0.00019999386878366454,
+      "loss": 1.1636,
+      "step": 1115
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2709734888315765,
+      "learning_rate": 0.0001999936923566411,
+      "loss": 1.1289,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32185710854614685,
+      "learning_rate": 0.00019999351342721262,
+      "loss": 1.1404,
+      "step": 1117
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31162451372291133,
+      "learning_rate": 0.0001999933319953837,
+      "loss": 1.112,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2752825720487004,
+      "learning_rate": 0.00019999314806115872,
+      "loss": 1.143,
+      "step": 1119
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2917340741765025,
+      "learning_rate": 0.0001999929616245424,
+      "loss": 1.0736,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.302518080441679,
+      "learning_rate": 0.0001999927726855394,
+      "loss": 1.0372,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25312327730893897,
+      "learning_rate": 0.00019999258124415442,
+      "loss": 1.1355,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2656439197184839,
+      "learning_rate": 0.00019999238730039222,
+      "loss": 1.0496,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24862847164472834,
+      "learning_rate": 0.00019999219085425768,
+      "loss": 1.0786,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28410932219305585,
+      "learning_rate": 0.00019999199190575575,
+      "loss": 1.0904,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2720824714159536,
+      "learning_rate": 0.00019999179045489135,
+      "loss": 1.0153,
+      "step": 1126
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2759096778009793,
+      "learning_rate": 0.00019999158650166958,
+      "loss": 1.1001,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6106436469666682,
+      "learning_rate": 0.0001999913800460955,
+      "loss": 1.1342,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23702252854532238,
+      "learning_rate": 0.00019999117108817428,
+      "loss": 1.0916,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25849383000967896,
+      "learning_rate": 0.0001999909596279112,
+      "loss": 1.1749,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28015440839970107,
+      "learning_rate": 0.0001999907456653115,
+      "loss": 1.1269,
+      "step": 1131
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26460543807236786,
+      "learning_rate": 0.00019999052920038053,
+      "loss": 1.1749,
+      "step": 1132
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27199116543714963,
+      "learning_rate": 0.0001999903102331237,
+      "loss": 1.0885,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22326728149419828,
+      "learning_rate": 0.00019999008876354658,
+      "loss": 1.106,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2957916257263048,
+      "learning_rate": 0.0001999898647916546,
+      "loss": 1.0432,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2926664019325837,
+      "learning_rate": 0.00019998963831745344,
+      "loss": 1.0905,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2560158671609372,
+      "learning_rate": 0.00019998940934094872,
+      "loss": 1.0585,
+      "step": 1137
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26739898061958195,
+      "learning_rate": 0.00019998917786214618,
+      "loss": 1.0375,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28363806649148315,
+      "learning_rate": 0.00019998894388105164,
+      "loss": 1.1372,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24811695289905492,
+      "learning_rate": 0.00019998870739767094,
+      "loss": 0.963,
+      "step": 1140
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28924944955268556,
+      "learning_rate": 0.00019998846841201,
+      "loss": 1.084,
+      "step": 1141
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2636078259266071,
+      "learning_rate": 0.00019998822692407478,
+      "loss": 1.0698,
+      "step": 1142
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2960715052303091,
+      "learning_rate": 0.0001999879829338714,
+      "loss": 1.2418,
+      "step": 1143
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23055523423338184,
+      "learning_rate": 0.00019998773644140584,
+      "loss": 1.169,
+      "step": 1144
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3043989132237114,
+      "learning_rate": 0.00019998748744668436,
+      "loss": 1.1707,
+      "step": 1145
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2652220574427321,
+      "learning_rate": 0.00019998723594971316,
+      "loss": 1.0908,
+      "step": 1146
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3564678490888735,
+      "learning_rate": 0.00019998698195049857,
+      "loss": 1.2161,
+      "step": 1147
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31147303496629464,
+      "learning_rate": 0.0001999867254490469,
+      "loss": 1.0795,
+      "step": 1148
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.33889241778400275,
+      "learning_rate": 0.00019998646644536457,
+      "loss": 1.2739,
+      "step": 1149
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27545717862181845,
+      "learning_rate": 0.00019998620493945807,
+      "loss": 1.074,
+      "step": 1150
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2576593733645889,
+      "learning_rate": 0.00019998594093133395,
+      "loss": 1.1171,
+      "step": 1151
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25688278200109543,
+      "learning_rate": 0.00019998567442099888,
+      "loss": 1.081,
+      "step": 1152
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3149168655482506,
+      "learning_rate": 0.0001999854054084594,
+      "loss": 1.1463,
+      "step": 1153
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28337563224538714,
+      "learning_rate": 0.00019998513389372233,
+      "loss": 1.1503,
+      "step": 1154
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28770905143410885,
+      "learning_rate": 0.00019998485987679447,
+      "loss": 1.0847,
+      "step": 1155
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2606963858756736,
+      "learning_rate": 0.00019998458335768264,
+      "loss": 1.2108,
+      "step": 1156
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2764798228490211,
+      "learning_rate": 0.00019998430433639376,
+      "loss": 1.1206,
+      "step": 1157
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3028071620221027,
+      "learning_rate": 0.00019998402281293484,
+      "loss": 1.1628,
+      "step": 1158
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.23132033284887418,
+      "learning_rate": 0.00019998373878731291,
+      "loss": 1.0603,
+      "step": 1159
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3197463940127305,
+      "learning_rate": 0.0001999834522595351,
+      "loss": 1.1337,
+      "step": 1160
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.258332321698546,
+      "learning_rate": 0.00019998316322960853,
+      "loss": 1.1347,
+      "step": 1161
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37002001593224093,
+      "learning_rate": 0.00019998287169754045,
+      "loss": 1.0973,
+      "step": 1162
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.35455352761567094,
+      "learning_rate": 0.00019998257766333822,
+      "loss": 1.0645,
+      "step": 1163
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25846010518779355,
+      "learning_rate": 0.00019998228112700912,
+      "loss": 1.099,
+      "step": 1164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45574094165617823,
+      "learning_rate": 0.00019998198208856058,
+      "loss": 1.2218,
+      "step": 1165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2806569349689396,
+      "learning_rate": 0.0001999816805480001,
+      "loss": 1.163,
+      "step": 1166
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3230556556910955,
+      "learning_rate": 0.00019998137650533527,
+      "loss": 1.0275,
+      "step": 1167
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.295834882980768,
+      "learning_rate": 0.0001999810699605736,
+      "loss": 1.0928,
+      "step": 1168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2838870309959414,
+      "learning_rate": 0.0001999807609137229,
+      "loss": 1.1008,
+      "step": 1169
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3164419453755688,
+      "learning_rate": 0.00019998044936479076,
+      "loss": 1.1307,
+      "step": 1170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.22581223994903243,
+      "learning_rate": 0.00019998013531378504,
+      "loss": 1.1228,
+      "step": 1171
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2611545463660394,
+      "learning_rate": 0.00019997981876071364,
+      "loss": 1.1299,
+      "step": 1172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2531091040846973,
+      "learning_rate": 0.00019997949970558437,
+      "loss": 1.2127,
+      "step": 1173
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3026109003824534,
+      "learning_rate": 0.00019997917814840537,
+      "loss": 1.1661,
+      "step": 1174
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3246636144590807,
+      "learning_rate": 0.00019997885408918454,
+      "loss": 1.0933,
+      "step": 1175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2574562133107501,
+      "learning_rate": 0.0001999785275279301,
+      "loss": 1.1417,
+      "step": 1176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2774322857021015,
+      "learning_rate": 0.00019997819846465014,
+      "loss": 1.2012,
+      "step": 1177
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2616318983640859,
+      "learning_rate": 0.00019997786689935292,
+      "loss": 1.064,
+      "step": 1178
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28254029261985597,
+      "learning_rate": 0.00019997753283204677,
+      "loss": 1.0777,
+      "step": 1179
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27452271119130867,
+      "learning_rate": 0.00019997719626274,
+      "loss": 1.2698,
+      "step": 1180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28281251810785174,
+      "learning_rate": 0.0001999768571914411,
+      "loss": 1.0866,
+      "step": 1181
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28877654959328175,
+      "learning_rate": 0.00019997651561815848,
+      "loss": 1.0607,
+      "step": 1182
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.31037965805002504,
+      "learning_rate": 0.00019997617154290077,
+      "loss": 1.1267,
+      "step": 1183
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.28875914132128616,
+      "learning_rate": 0.0001999758249656765,
+      "loss": 1.159,
+      "step": 1184
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26263865798211755,
+      "learning_rate": 0.00019997547588649438,
+      "loss": 1.1947,
+      "step": 1185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.25456732695723555,
+      "learning_rate": 0.00019997512430536314,
+      "loss": 1.1032,
+      "step": 1186
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.32332294350656676,
+      "learning_rate": 0.00019997477022229158,
+      "loss": 1.1283,
+      "step": 1187
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3066913219600098,
+      "learning_rate": 0.00019997441363728857,
+      "loss": 1.2178,
+      "step": 1188
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2982903122596879,
+      "learning_rate": 0.00019997405455036304,
+      "loss": 1.1613,
+      "step": 1189
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3072555573162715,
+      "learning_rate": 0.00019997369296152396,
+      "loss": 1.1927,
+      "step": 1190
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.27576655968710867,
+      "learning_rate": 0.00019997332887078034,
+      "loss": 1.148,
+      "step": 1191
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2651145511693654,
+      "learning_rate": 0.0001999729622781414,
+      "loss": 1.061,
+      "step": 1192
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.26492079668590324,
+      "learning_rate": 0.00019997259318361622,
+      "loss": 1.0942,
+      "step": 1193
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.268256812103858,
+      "learning_rate": 0.00019997222158721405,
+      "loss": 1.0478,
+      "step": 1194
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.29544097416774406,
+      "learning_rate": 0.00019997184748894422,
+      "loss": 1.084,
+      "step": 1195
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.24315111594316274,
+      "learning_rate": 0.00019997147088881607,
+      "loss": 1.1187,
+      "step": 1196
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2887864253869539,
+      "learning_rate": 0.00019997109178683905,
+      "loss": 1.1425,
+      "step": 1197
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.243613837120699,
+      "learning_rate": 0.0001999707101830226,
+      "loss": 1.2192,
+      "step": 1198
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2670339437152679,
+      "learning_rate": 0.00019997032607737633,
+      "loss": 0.9346,
+      "step": 1199
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.286415306705152,
+      "learning_rate": 0.0001999699394699098,
+      "loss": 1.2044,
+      "step": 1200
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2649888516882499,
+      "learning_rate": 0.0001999695503606327,
+      "loss": 1.1028,
+      "step": 1201
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.2784005327190465,
+      "learning_rate": 0.00019996915874955477,
+      "loss": 1.1883,
+      "step": 1202
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2827618352465213,
+      "learning_rate": 0.00019996876463668586,
+      "loss": 1.1373,
+      "step": 1203
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27252281665016315,
+      "learning_rate": 0.00019996836802203575,
+      "loss": 1.1434,
+      "step": 1204
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2829042974144935,
+      "learning_rate": 0.00019996796890561438,
+      "loss": 1.1242,
+      "step": 1205
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25919288560265524,
+      "learning_rate": 0.0001999675672874318,
+      "loss": 1.0836,
+      "step": 1206
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2462264710662166,
+      "learning_rate": 0.00019996716316749802,
+      "loss": 1.0824,
+      "step": 1207
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24248848464047051,
+      "learning_rate": 0.00019996675654582313,
+      "loss": 1.1398,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2542004323802939,
+      "learning_rate": 0.00019996634742241732,
+      "loss": 1.0721,
+      "step": 1209
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2665303881003603,
+      "learning_rate": 0.0001999659357972909,
+      "loss": 1.1183,
+      "step": 1210
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2776270813403137,
+      "learning_rate": 0.00019996552167045407,
+      "loss": 1.2601,
+      "step": 1211
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3169789236787061,
+      "learning_rate": 0.00019996510504191722,
+      "loss": 1.2331,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.247880062307769,
+      "learning_rate": 0.00019996468591169082,
+      "loss": 1.1088,
+      "step": 1213
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2743889662351891,
+      "learning_rate": 0.00019996426427978532,
+      "loss": 1.0541,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2911803551557875,
+      "learning_rate": 0.00019996384014621128,
+      "loss": 1.0826,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26095098186965116,
+      "learning_rate": 0.0001999634135109793,
+      "loss": 1.3256,
+      "step": 1216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2856385392323691,
+      "learning_rate": 0.0001999629843741001,
+      "loss": 1.1093,
+      "step": 1217
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2530173370522601,
+      "learning_rate": 0.00019996255273558436,
+      "loss": 1.0579,
+      "step": 1218
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2549728915097247,
+      "learning_rate": 0.00019996211859544296,
+      "loss": 0.9691,
+      "step": 1219
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2799442188519433,
+      "learning_rate": 0.00019996168195368668,
+      "loss": 1.0273,
+      "step": 1220
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30580035630418173,
+      "learning_rate": 0.0001999612428103265,
+      "loss": 1.2193,
+      "step": 1221
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2773508640042361,
+      "learning_rate": 0.00019996080116537339,
+      "loss": 1.059,
+      "step": 1222
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2917592835447524,
+      "learning_rate": 0.0001999603570188384,
+      "loss": 1.1454,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27511773778172,
+      "learning_rate": 0.00019995991037073267,
+      "loss": 1.0708,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2299812326773386,
+      "learning_rate": 0.00019995946122106735,
+      "loss": 0.9796,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28024776222182823,
+      "learning_rate": 0.00019995900956985369,
+      "loss": 1.1255,
+      "step": 1226
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26814475354624795,
+      "learning_rate": 0.000199958555417103,
+      "loss": 1.0957,
+      "step": 1227
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2872677212088015,
+      "learning_rate": 0.00019995809876282664,
+      "loss": 1.035,
+      "step": 1228
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2521840908110662,
+      "learning_rate": 0.00019995763960703605,
+      "loss": 1.0637,
+      "step": 1229
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.40673410489243833,
+      "learning_rate": 0.00019995717794974268,
+      "loss": 1.0153,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31023442713481986,
+      "learning_rate": 0.0001999567137909581,
+      "loss": 1.1233,
+      "step": 1231
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2712959430519531,
+      "learning_rate": 0.000199956247130694,
+      "loss": 1.1379,
+      "step": 1232
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28625127223535557,
+      "learning_rate": 0.0001999557779689619,
+      "loss": 1.2184,
+      "step": 1233
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30076306094072325,
+      "learning_rate": 0.0001999553063057737,
+      "loss": 1.1837,
+      "step": 1234
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2653252660175731,
+      "learning_rate": 0.00019995483214114114,
+      "loss": 1.0155,
+      "step": 1235
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2752439328777632,
+      "learning_rate": 0.0001999543554750761,
+      "loss": 1.1711,
+      "step": 1236
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27211117879122465,
+      "learning_rate": 0.00019995387630759046,
+      "loss": 1.1123,
+      "step": 1237
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26487361645343366,
+      "learning_rate": 0.00019995339463869626,
+      "loss": 1.074,
+      "step": 1238
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3021321790418319,
+      "learning_rate": 0.00019995291046840554,
+      "loss": 1.119,
+      "step": 1239
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2361755130550678,
+      "learning_rate": 0.00019995242379673041,
+      "loss": 1.1614,
+      "step": 1240
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.254336538394881,
+      "learning_rate": 0.00019995193462368308,
+      "loss": 1.0516,
+      "step": 1241
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26752053595877906,
+      "learning_rate": 0.00019995144294927575,
+      "loss": 1.0767,
+      "step": 1242
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2597082485713151,
+      "learning_rate": 0.00019995094877352075,
+      "loss": 1.0225,
+      "step": 1243
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2455539656845191,
+      "learning_rate": 0.00019995045209643042,
+      "loss": 1.1359,
+      "step": 1244
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27400725474036985,
+      "learning_rate": 0.00019994995291801725,
+      "loss": 1.1361,
+      "step": 1245
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2924655478762353,
+      "learning_rate": 0.00019994945123829366,
+      "loss": 1.1821,
+      "step": 1246
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.29710840841514063,
+      "learning_rate": 0.00019994894705727224,
+      "loss": 1.2383,
+      "step": 1247
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2813768369686393,
+      "learning_rate": 0.0001999484403749656,
+      "loss": 1.0407,
+      "step": 1248
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.28144059564025686,
+      "learning_rate": 0.00019994793119138644,
+      "loss": 1.2073,
+      "step": 1249
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24328115129403632,
+      "learning_rate": 0.00019994741950654746,
+      "loss": 1.1233,
+      "step": 1250
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2779013414605166,
+      "learning_rate": 0.00019994690532046155,
+      "loss": 1.1908,
+      "step": 1251
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2939081176965555,
+      "learning_rate": 0.00019994638863314146,
+      "loss": 1.0965,
+      "step": 1252
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.267694708686987,
+      "learning_rate": 0.0001999458694446002,
+      "loss": 1.0733,
+      "step": 1253
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.31511116622951807,
+      "learning_rate": 0.00019994534775485075,
+      "loss": 1.0266,
+      "step": 1254
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2718983181420745,
+      "learning_rate": 0.00019994482356390617,
+      "loss": 1.1895,
+      "step": 1255
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2558317058922546,
+      "learning_rate": 0.00019994429687177957,
+      "loss": 1.1239,
+      "step": 1256
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2744763799941594,
+      "learning_rate": 0.00019994376767848407,
+      "loss": 1.2122,
+      "step": 1257
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2643964783968129,
+      "learning_rate": 0.00019994323598403302,
+      "loss": 1.0028,
+      "step": 1258
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25268007251056396,
+      "learning_rate": 0.0001999427017884397,
+      "loss": 1.1376,
+      "step": 1259
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.230312980862343,
+      "learning_rate": 0.0001999421650917174,
+      "loss": 1.0732,
+      "step": 1260
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2740089140496981,
+      "learning_rate": 0.00019994162589387964,
+      "loss": 0.9982,
+      "step": 1261
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2726058611551938,
+      "learning_rate": 0.0001999410841949399,
+      "loss": 1.039,
+      "step": 1262
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3233351751878856,
+      "learning_rate": 0.00019994053999491167,
+      "loss": 1.2084,
+      "step": 1263
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24718142863026807,
+      "learning_rate": 0.00019993999329380864,
+      "loss": 0.9947,
+      "step": 1264
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2877635724046095,
+      "learning_rate": 0.00019993944409164448,
+      "loss": 1.1189,
+      "step": 1265
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3002448997033507,
+      "learning_rate": 0.00019993889238843288,
+      "loss": 1.0936,
+      "step": 1266
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24948366812390463,
+      "learning_rate": 0.00019993833818418772,
+      "loss": 1.1574,
+      "step": 1267
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24037766838141317,
+      "learning_rate": 0.00019993778147892285,
+      "loss": 1.1475,
+      "step": 1268
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24578207537112048,
+      "learning_rate": 0.00019993722227265218,
+      "loss": 1.1365,
+      "step": 1269
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24088318104194462,
+      "learning_rate": 0.00019993666056538972,
+      "loss": 1.0947,
+      "step": 1270
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2921571119742658,
+      "learning_rate": 0.0001999360963571495,
+      "loss": 1.0772,
+      "step": 1271
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30049773628170273,
+      "learning_rate": 0.00019993552964794566,
+      "loss": 1.2072,
+      "step": 1272
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3160778257013834,
+      "learning_rate": 0.0001999349604377924,
+      "loss": 1.0676,
+      "step": 1273
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23884600224412095,
+      "learning_rate": 0.00019993438872670396,
+      "loss": 1.0855,
+      "step": 1274
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2615500721708398,
+      "learning_rate": 0.0001999338145146946,
+      "loss": 1.1958,
+      "step": 1275
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2591847182045251,
+      "learning_rate": 0.00019993323780177874,
+      "loss": 1.0991,
+      "step": 1276
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2800525884700228,
+      "learning_rate": 0.00019993265858797083,
+      "loss": 1.0018,
+      "step": 1277
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25703108671920066,
+      "learning_rate": 0.0001999320768732853,
+      "loss": 1.0842,
+      "step": 1278
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2814109826464174,
+      "learning_rate": 0.00019993149265773674,
+      "loss": 1.1056,
+      "step": 1279
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26560101203311826,
+      "learning_rate": 0.0001999309059413398,
+      "loss": 1.1028,
+      "step": 1280
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2592301570333206,
+      "learning_rate": 0.00019993031672410912,
+      "loss": 1.2395,
+      "step": 1281
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2903887529235589,
+      "learning_rate": 0.00019992972500605945,
+      "loss": 1.2269,
+      "step": 1282
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.30985749070799845,
+      "learning_rate": 0.00019992913078720559,
+      "loss": 1.0394,
+      "step": 1283
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2427582461596586,
+      "learning_rate": 0.00019992853406756246,
+      "loss": 1.0323,
+      "step": 1284
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2674764639506977,
+      "learning_rate": 0.00019992793484714495,
+      "loss": 1.0569,
+      "step": 1285
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27159902681019893,
+      "learning_rate": 0.00019992733312596808,
+      "loss": 1.0051,
+      "step": 1286
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27222923746834743,
+      "learning_rate": 0.00019992672890404689,
+      "loss": 1.1311,
+      "step": 1287
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2557430982813261,
+      "learning_rate": 0.0001999261221813965,
+      "loss": 1.121,
+      "step": 1288
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2927850888283984,
+      "learning_rate": 0.0001999255129580321,
+      "loss": 1.1497,
+      "step": 1289
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2859310412286254,
+      "learning_rate": 0.00019992490123396897,
+      "loss": 1.1786,
+      "step": 1290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25665923491793874,
+      "learning_rate": 0.00019992428700922236,
+      "loss": 1.0947,
+      "step": 1291
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.24916615055802435,
+      "learning_rate": 0.00019992367028380764,
+      "loss": 1.1687,
+      "step": 1292
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2731992285743491,
+      "learning_rate": 0.00019992305105774033,
+      "loss": 1.1253,
+      "step": 1293
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26082758101857634,
+      "learning_rate": 0.0001999224293310358,
+      "loss": 1.1551,
+      "step": 1294
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2677306810392585,
+      "learning_rate": 0.00019992180510370976,
+      "loss": 1.2005,
+      "step": 1295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.23143531134792353,
+      "learning_rate": 0.00019992117837577768,
+      "loss": 1.0862,
+      "step": 1296
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2655840524523936,
+      "learning_rate": 0.00019992054914725533,
+      "loss": 1.1536,
+      "step": 1297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2640787870858716,
+      "learning_rate": 0.00019991991741815849,
+      "loss": 1.1011,
+      "step": 1298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.26001791207694314,
+      "learning_rate": 0.00019991928318850285,
+      "loss": 1.191,
+      "step": 1299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34274104078637135,
+      "learning_rate": 0.0001999186464583044,
+      "loss": 1.1189,
+      "step": 1300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.25164999546725786,
+      "learning_rate": 0.000199918007227579,
+      "loss": 1.1755,
+      "step": 1301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2619736635468605,
+      "learning_rate": 0.00019991736549634267,
+      "loss": 1.0093,
+      "step": 1302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2610328681252208,
+      "learning_rate": 0.00019991672126461147,
+      "loss": 1.2091,
+      "step": 1303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.27609879841972346,
+      "learning_rate": 0.00019991607453240153,
+      "loss": 1.1545,
+      "step": 1304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.2438919416348471,
+      "learning_rate": 0.00019991542529972905,
+      "loss": 1.0793,
+      "step": 1305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.34769105624706653,
+      "learning_rate": 0.00019991477356661022,
+      "loss": 1.0689,
+      "step": 1306
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24988099416617487,
+      "learning_rate": 0.0001999141193330614,
+      "loss": 0.9971,
+      "step": 1307
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26587548843318687,
+      "learning_rate": 0.00019991346259909897,
+      "loss": 1.0847,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27946757540245054,
+      "learning_rate": 0.00019991280336473935,
+      "loss": 1.137,
+      "step": 1309
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2928663604215012,
+      "learning_rate": 0.000199912141629999,
+      "loss": 1.157,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24871198294790245,
+      "learning_rate": 0.00019991147739489455,
+      "loss": 1.0734,
+      "step": 1311
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2384639920280004,
+      "learning_rate": 0.00019991081065944254,
+      "loss": 1.0737,
+      "step": 1312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2656285356842293,
+      "learning_rate": 0.00019991014142365976,
+      "loss": 1.0513,
+      "step": 1313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28562521186701684,
+      "learning_rate": 0.00019990946968756286,
+      "loss": 1.1639,
+      "step": 1314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2648382753477716,
+      "learning_rate": 0.0001999087954511687,
+      "loss": 1.1691,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3183214028731789,
+      "learning_rate": 0.00019990811871449412,
+      "loss": 1.1523,
+      "step": 1316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27420941387779174,
+      "learning_rate": 0.0001999074394775561,
+      "loss": 1.0882,
+      "step": 1317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24879969547328998,
+      "learning_rate": 0.00019990675774037164,
+      "loss": 1.0498,
+      "step": 1318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2738940434550004,
+      "learning_rate": 0.00019990607350295776,
+      "loss": 1.145,
+      "step": 1319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25098756631874913,
+      "learning_rate": 0.0001999053867653316,
+      "loss": 1.0748,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.285510083056974,
+      "learning_rate": 0.00019990469752751032,
+      "loss": 1.0964,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2528776643597391,
+      "learning_rate": 0.00019990400578951125,
+      "loss": 1.1219,
+      "step": 1322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2737172954753307,
+      "learning_rate": 0.0001999033115513516,
+      "loss": 1.144,
+      "step": 1323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2788543402132724,
+      "learning_rate": 0.00019990261481304882,
+      "loss": 1.2348,
+      "step": 1324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26612476420962583,
+      "learning_rate": 0.00019990191557462032,
+      "loss": 1.1158,
+      "step": 1325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27339420148068405,
+      "learning_rate": 0.00019990121383608357,
+      "loss": 1.193,
+      "step": 1326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2660334700185283,
+      "learning_rate": 0.0001999005095974562,
+      "loss": 1.0692,
+      "step": 1327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2779459146879295,
+      "learning_rate": 0.00019989980285875576,
+      "loss": 1.1296,
+      "step": 1328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.30588845858874797,
+      "learning_rate": 0.00019989909361999998,
+      "loss": 1.102,
+      "step": 1329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27026409826467807,
+      "learning_rate": 0.0001998983818812066,
+      "loss": 1.0786,
+      "step": 1330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23833256583341253,
+      "learning_rate": 0.00019989766764239342,
+      "loss": 1.1167,
+      "step": 1331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21010010951243135,
+      "learning_rate": 0.00019989695090357832,
+      "loss": 0.9995,
+      "step": 1332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28853987325226016,
+      "learning_rate": 0.00019989623166477926,
+      "loss": 0.9722,
+      "step": 1333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27581828527816954,
+      "learning_rate": 0.0001998955099260142,
+      "loss": 1.1367,
+      "step": 1334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3266981204077468,
+      "learning_rate": 0.00019989478568730124,
+      "loss": 1.0954,
+      "step": 1335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26634585407279676,
+      "learning_rate": 0.00019989405894865848,
+      "loss": 1.0318,
+      "step": 1336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2862324766372512,
+      "learning_rate": 0.0001998933297101041,
+      "loss": 1.2006,
+      "step": 1337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3219899860952693,
+      "learning_rate": 0.0001998925979716564,
+      "loss": 1.2428,
+      "step": 1338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2593333848699128,
+      "learning_rate": 0.0001998918637333336,
+      "loss": 1.0661,
+      "step": 1339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25460421291647545,
+      "learning_rate": 0.00019989112699515417,
+      "loss": 1.1549,
+      "step": 1340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2806093932847469,
+      "learning_rate": 0.0001998903877571365,
+      "loss": 1.1396,
+      "step": 1341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28375827740366566,
+      "learning_rate": 0.00019988964601929911,
+      "loss": 1.1933,
+      "step": 1342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25635075031051086,
+      "learning_rate": 0.00019988890178166053,
+      "loss": 1.0908,
+      "step": 1343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2888250115582476,
+      "learning_rate": 0.00019988815504423942,
+      "loss": 1.1556,
+      "step": 1344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23088975772081866,
+      "learning_rate": 0.00019988740580705443,
+      "loss": 1.1304,
+      "step": 1345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27596330498232263,
+      "learning_rate": 0.00019988665407012435,
+      "loss": 1.11,
+      "step": 1346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24512542720942831,
+      "learning_rate": 0.00019988589983346798,
+      "loss": 1.1509,
+      "step": 1347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25605979489959907,
+      "learning_rate": 0.00019988514309710417,
+      "loss": 1.0923,
+      "step": 1348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2532094527125873,
+      "learning_rate": 0.0001998843838610519,
+      "loss": 1.0768,
+      "step": 1349
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2951636432930729,
+      "learning_rate": 0.00019988362212533013,
+      "loss": 1.1431,
+      "step": 1350
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25484076775146086,
+      "learning_rate": 0.000199882857889958,
+      "loss": 1.1345,
+      "step": 1351
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23726743985931845,
+      "learning_rate": 0.0001998820911549545,
+      "loss": 1.106,
+      "step": 1352
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2587569140280319,
+      "learning_rate": 0.00019988132192033892,
+      "loss": 1.1131,
+      "step": 1353
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26259253430908974,
+      "learning_rate": 0.0001998805501861305,
+      "loss": 1.1066,
+      "step": 1354
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24656763039460808,
+      "learning_rate": 0.00019987977595234852,
+      "loss": 1.1207,
+      "step": 1355
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27728609474508775,
+      "learning_rate": 0.0001998789992190124,
+      "loss": 1.0683,
+      "step": 1356
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.266917547376331,
+      "learning_rate": 0.00019987821998614154,
+      "loss": 1.1693,
+      "step": 1357
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2915661042761893,
+      "learning_rate": 0.00019987743825375544,
+      "loss": 1.1064,
+      "step": 1358
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24772664636633338,
+      "learning_rate": 0.00019987665402187367,
+      "loss": 0.9948,
+      "step": 1359
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2855087767458927,
+      "learning_rate": 0.0001998758672905159,
+      "loss": 1.1449,
+      "step": 1360
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2772392094730354,
+      "learning_rate": 0.00019987507805970176,
+      "loss": 1.1349,
+      "step": 1361
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2612704349828971,
+      "learning_rate": 0.000199874286329451,
+      "loss": 1.0995,
+      "step": 1362
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24900155104384822,
+      "learning_rate": 0.00019987349209978352,
+      "loss": 1.1157,
+      "step": 1363
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26136521449937644,
+      "learning_rate": 0.0001998726953707191,
+      "loss": 1.1324,
+      "step": 1364
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27010231238155247,
+      "learning_rate": 0.0001998718961422777,
+      "loss": 1.1295,
+      "step": 1365
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28056706441584167,
+      "learning_rate": 0.00019987109441447934,
+      "loss": 1.1236,
+      "step": 1366
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24673474376997923,
+      "learning_rate": 0.00019987029018734407,
+      "loss": 1.1493,
+      "step": 1367
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25553237785153865,
+      "learning_rate": 0.00019986948346089201,
+      "loss": 1.1698,
+      "step": 1368
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2370749606006542,
+      "learning_rate": 0.0001998686742351434,
+      "loss": 1.03,
+      "step": 1369
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2786623699017042,
+      "learning_rate": 0.00019986786251011842,
+      "loss": 1.0002,
+      "step": 1370
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2582710457594854,
+      "learning_rate": 0.0001998670482858374,
+      "loss": 1.1957,
+      "step": 1371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2414613658655144,
+      "learning_rate": 0.00019986623156232076,
+      "loss": 1.1471,
+      "step": 1372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28959428619565936,
+      "learning_rate": 0.0001998654123395889,
+      "loss": 1.1716,
+      "step": 1373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.22250465518264687,
+      "learning_rate": 0.00019986459061766234,
+      "loss": 1.146,
+      "step": 1374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26717391829997156,
+      "learning_rate": 0.00019986376639656163,
+      "loss": 1.1858,
+      "step": 1375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2580258297035148,
+      "learning_rate": 0.00019986293967630742,
+      "loss": 1.1115,
+      "step": 1376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2678948286097016,
+      "learning_rate": 0.0001998621104569204,
+      "loss": 1.1202,
+      "step": 1377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.291208905746566,
+      "learning_rate": 0.00019986127873842128,
+      "loss": 1.1424,
+      "step": 1378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.21902207523941578,
+      "learning_rate": 0.00019986044452083087,
+      "loss": 1.1687,
+      "step": 1379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2907782173595465,
+      "learning_rate": 0.00019985960780417012,
+      "loss": 1.068,
+      "step": 1380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2629226322026941,
+      "learning_rate": 0.0001998587685884599,
+      "loss": 1.0816,
+      "step": 1381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2828450761974692,
+      "learning_rate": 0.00019985792687372126,
+      "loss": 1.0958,
+      "step": 1382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26457153514550896,
+      "learning_rate": 0.00019985708265997523,
+      "loss": 1.1454,
+      "step": 1383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25857136554982324,
+      "learning_rate": 0.00019985623594724294,
+      "loss": 1.074,
+      "step": 1384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.24297032009355968,
+      "learning_rate": 0.00019985538673554558,
+      "loss": 1.2311,
+      "step": 1385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29342141362537133,
+      "learning_rate": 0.00019985453502490447,
+      "loss": 1.0765,
+      "step": 1386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27688151236884495,
+      "learning_rate": 0.0001998536808153408,
+      "loss": 1.1203,
+      "step": 1387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.288243343301449,
+      "learning_rate": 0.000199852824106876,
+      "loss": 1.0883,
+      "step": 1388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.252259697983258,
+      "learning_rate": 0.00019985196489953158,
+      "loss": 1.2147,
+      "step": 1389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26515095142272627,
+      "learning_rate": 0.00019985110319332896,
+      "loss": 1.0793,
+      "step": 1390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.28200900869959167,
+      "learning_rate": 0.00019985023898828972,
+      "loss": 1.0851,
+      "step": 1391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.22101241462701582,
+      "learning_rate": 0.0001998493722844355,
+      "loss": 1.0172,
+      "step": 1392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.29717507654200653,
+      "learning_rate": 0.000199848503081788,
+      "loss": 1.1634,
+      "step": 1393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2734078285310837,
+      "learning_rate": 0.00019984763138036893,
+      "loss": 1.206,
+      "step": 1394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.31430473861156777,
+      "learning_rate": 0.00019984675718020016,
+      "loss": 1.1355,
+      "step": 1395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25927920918090913,
+      "learning_rate": 0.00019984588048130352,
+      "loss": 1.0166,
+      "step": 1396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2943475145072853,
+      "learning_rate": 0.000199845001283701,
+      "loss": 1.0848,
+      "step": 1397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26595828195937876,
+      "learning_rate": 0.0001998441195874145,
+      "loss": 1.0692,
+      "step": 1398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27821823753565383,
+      "learning_rate": 0.00019984323539246624,
+      "loss": 1.2192,
+      "step": 1399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2678637821368645,
+      "learning_rate": 0.00019984234869887825,
+      "loss": 1.121,
+      "step": 1400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.25399068189204,
+      "learning_rate": 0.0001998414595066727,
+      "loss": 1.0925,
+      "step": 1401
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23859094414183193,
+      "learning_rate": 0.00019984056781587191,
+      "loss": 1.0955,
+      "step": 1402
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.23969916810545017,
+      "learning_rate": 0.00019983967362649814,
+      "loss": 1.0125,
+      "step": 1403
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2542524563052129,
+      "learning_rate": 0.0001998387769385738,
+      "loss": 1.0373,
+      "step": 1404
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.27487621829351494,
+      "learning_rate": 0.0001998378777521213,
+      "loss": 1.0952,
+      "step": 1405
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2654518507826389,
+      "learning_rate": 0.0001998369760671632,
+      "loss": 1.1369,
+      "step": 1406
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.273289916929377,
+      "learning_rate": 0.000199836071883722,
+      "loss": 1.1703,
+      "step": 1407
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.26544162174771085,
+      "learning_rate": 0.0001998351652018204,
+      "loss": 1.1784,
+      "step": 1408
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2746593440350384,
+      "learning_rate": 0.000199834256021481,
+      "loss": 1.0443,
+      "step": 1409
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.2501215693772626,
+      "learning_rate": 0.00019983334434272662,
+      "loss": 1.089,
+      "step": 1410
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.268614062421362,
+      "learning_rate": 0.00019983243016558007,
+      "loss": 0.9597,
+      "step": 1411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25030986160663105,
+      "learning_rate": 0.00019983151349006417,
+      "loss": 1.2056,
+      "step": 1412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2603381688146406,
+      "learning_rate": 0.00019983059431620195,
+      "loss": 1.1547,
+      "step": 1413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2421120755485211,
+      "learning_rate": 0.0001998296726440163,
+      "loss": 1.1084,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2521346105354822,
+      "learning_rate": 0.00019982874847353043,
+      "loss": 1.0624,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27862535658172966,
+      "learning_rate": 0.00019982782180476733,
+      "loss": 1.1186,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2947914398618809,
+      "learning_rate": 0.00019982689263775026,
+      "loss": 1.0843,
+      "step": 1417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28536726878487906,
+      "learning_rate": 0.0001998259609725025,
+      "loss": 1.0653,
+      "step": 1418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2854261291390814,
+      "learning_rate": 0.00019982502680904732,
+      "loss": 1.1266,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2788285385425596,
+      "learning_rate": 0.0001998240901474081,
+      "loss": 1.1592,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2919295150525652,
+      "learning_rate": 0.00019982315098760825,
+      "loss": 1.23,
+      "step": 1421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2667438613342299,
+      "learning_rate": 0.00019982220932967135,
+      "loss": 1.1653,
+      "step": 1422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2937856653774384,
+      "learning_rate": 0.00019982126517362092,
+      "loss": 1.054,
+      "step": 1423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2787585191603891,
+      "learning_rate": 0.0001998203185194806,
+      "loss": 1.1195,
+      "step": 1424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2808724466906523,
+      "learning_rate": 0.00019981936936727402,
+      "loss": 1.1324,
+      "step": 1425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24401430651001996,
+      "learning_rate": 0.00019981841771702505,
+      "loss": 1.138,
+      "step": 1426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25174511987554116,
+      "learning_rate": 0.00019981746356875744,
+      "loss": 1.1906,
+      "step": 1427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2538503054933492,
+      "learning_rate": 0.00019981650692249504,
+      "loss": 1.1029,
+      "step": 1428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2764211181550846,
+      "learning_rate": 0.00019981554777826185,
+      "loss": 1.058,
+      "step": 1429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2952219050552689,
+      "learning_rate": 0.00019981458613608182,
+      "loss": 1.0941,
+      "step": 1430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2787064009838231,
+      "learning_rate": 0.00019981362199597907,
+      "loss": 1.1565,
+      "step": 1431
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2737728015545912,
+      "learning_rate": 0.00019981265535797766,
+      "loss": 1.181,
+      "step": 1432
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24759285916434062,
+      "learning_rate": 0.00019981168622210184,
+      "loss": 1.1008,
+      "step": 1433
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25095588755500636,
+      "learning_rate": 0.00019981071458837586,
+      "loss": 1.1312,
+      "step": 1434
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24482769279016886,
+      "learning_rate": 0.00019980974045682399,
+      "loss": 1.0652,
+      "step": 1435
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3060859989141741,
+      "learning_rate": 0.00019980876382747064,
+      "loss": 1.157,
+      "step": 1436
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2880153757669946,
+      "learning_rate": 0.00019980778470034025,
+      "loss": 1.2203,
+      "step": 1437
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27454359894274216,
+      "learning_rate": 0.00019980680307545733,
+      "loss": 1.0434,
+      "step": 1438
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2883237368727595,
+      "learning_rate": 0.00019980581895284646,
+      "loss": 1.1272,
+      "step": 1439
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2703448143010634,
+      "learning_rate": 0.0001998048323325322,
+      "loss": 1.0698,
+      "step": 1440
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2583337496657767,
+      "learning_rate": 0.00019980384321453931,
+      "loss": 1.0424,
+      "step": 1441
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2698951572905667,
+      "learning_rate": 0.00019980285159889251,
+      "loss": 1.1694,
+      "step": 1442
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3289970000039555,
+      "learning_rate": 0.00019980185748561663,
+      "loss": 1.1242,
+      "step": 1443
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.33181434848256175,
+      "learning_rate": 0.00019980086087473655,
+      "loss": 1.2397,
+      "step": 1444
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3011135638992476,
+      "learning_rate": 0.0001997998617662772,
+      "loss": 1.1714,
+      "step": 1445
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26023630385034113,
+      "learning_rate": 0.0001997988601602636,
+      "loss": 1.1072,
+      "step": 1446
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27251940674998737,
+      "learning_rate": 0.00019979785605672078,
+      "loss": 1.0825,
+      "step": 1447
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2554656785010738,
+      "learning_rate": 0.0001997968494556739,
+      "loss": 1.0752,
+      "step": 1448
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23684498161731513,
+      "learning_rate": 0.00019979584035714813,
+      "loss": 1.06,
+      "step": 1449
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31872148847175163,
+      "learning_rate": 0.00019979482876116876,
+      "loss": 1.1667,
+      "step": 1450
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25697407458502153,
+      "learning_rate": 0.00019979381466776107,
+      "loss": 1.0822,
+      "step": 1451
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2680749561732018,
+      "learning_rate": 0.00019979279807695046,
+      "loss": 1.1473,
+      "step": 1452
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23738996733574236,
+      "learning_rate": 0.00019979177898876233,
+      "loss": 1.1218,
+      "step": 1453
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2518660152482727,
+      "learning_rate": 0.00019979075740322224,
+      "loss": 1.1364,
+      "step": 1454
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2843340075762548,
+      "learning_rate": 0.00019978973332035574,
+      "loss": 1.1052,
+      "step": 1455
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23668848969056627,
+      "learning_rate": 0.0001997887067401884,
+      "loss": 1.1692,
+      "step": 1456
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28502539435405605,
+      "learning_rate": 0.000199787677662746,
+      "loss": 1.2374,
+      "step": 1457
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25655559740700473,
+      "learning_rate": 0.00019978664608805423,
+      "loss": 1.1088,
+      "step": 1458
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2585840843019671,
+      "learning_rate": 0.00019978561201613895,
+      "loss": 1.1601,
+      "step": 1459
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2550422041426665,
+      "learning_rate": 0.00019978457544702602,
+      "loss": 1.1033,
+      "step": 1460
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2899278429933367,
+      "learning_rate": 0.00019978353638074137,
+      "loss": 1.1611,
+      "step": 1461
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.234358263645623,
+      "learning_rate": 0.000199782494817311,
+      "loss": 1.0151,
+      "step": 1462
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.24241424067059672,
+      "learning_rate": 0.00019978145075676097,
+      "loss": 1.1854,
+      "step": 1463
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2685159364940143,
+      "learning_rate": 0.00019978040419911744,
+      "loss": 1.0691,
+      "step": 1464
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.31415333163030407,
+      "learning_rate": 0.0001997793551444066,
+      "loss": 1.0344,
+      "step": 1465
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2366061685991993,
+      "learning_rate": 0.00019977830359265472,
+      "loss": 1.1004,
+      "step": 1466
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2485341188969311,
+      "learning_rate": 0.00019977724954388802,
+      "loss": 1.1078,
+      "step": 1467
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29387809184755664,
+      "learning_rate": 0.00019977619299813297,
+      "loss": 1.1114,
+      "step": 1468
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2652962832837201,
+      "learning_rate": 0.00019977513395541598,
+      "loss": 1.1429,
+      "step": 1469
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.296485600586474,
+      "learning_rate": 0.00019977407241576355,
+      "loss": 1.1725,
+      "step": 1470
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25206336596375595,
+      "learning_rate": 0.00019977300837920227,
+      "loss": 1.1611,
+      "step": 1471
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2550564024652486,
+      "learning_rate": 0.00019977194184575873,
+      "loss": 1.1339,
+      "step": 1472
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27675386051052386,
+      "learning_rate": 0.00019977087281545966,
+      "loss": 1.1179,
+      "step": 1473
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2615226908183446,
+      "learning_rate": 0.00019976980128833178,
+      "loss": 1.0797,
+      "step": 1474
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.268113224944456,
+      "learning_rate": 0.00019976872726440193,
+      "loss": 1.1163,
+      "step": 1475
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23912506420391394,
+      "learning_rate": 0.00019976765074369697,
+      "loss": 1.099,
+      "step": 1476
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28513459887618126,
+      "learning_rate": 0.00019976657172624383,
+      "loss": 1.0811,
+      "step": 1477
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28114711060770275,
+      "learning_rate": 0.00019976549021206958,
+      "loss": 1.2168,
+      "step": 1478
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2572724456046117,
+      "learning_rate": 0.0001997644062012012,
+      "loss": 1.0752,
+      "step": 1479
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3019482336468289,
+      "learning_rate": 0.00019976331969366587,
+      "loss": 1.1129,
+      "step": 1480
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2566195486989078,
+      "learning_rate": 0.00019976223068949076,
+      "loss": 1.1825,
+      "step": 1481
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27980667259100733,
+      "learning_rate": 0.00019976113918870314,
+      "loss": 1.1828,
+      "step": 1482
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.30150809201942436,
+      "learning_rate": 0.0001997600451913303,
+      "loss": 1.1247,
+      "step": 1483
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2716498140574145,
+      "learning_rate": 0.0001997589486973996,
+      "loss": 1.1254,
+      "step": 1484
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2625386839698671,
+      "learning_rate": 0.00019975784970693855,
+      "loss": 1.1081,
+      "step": 1485
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29792002537853113,
+      "learning_rate": 0.00019975674821997463,
+      "loss": 1.1597,
+      "step": 1486
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2301276867707596,
+      "learning_rate": 0.00019975564423653538,
+      "loss": 1.1342,
+      "step": 1487
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27119741051691004,
+      "learning_rate": 0.00019975453775664844,
+      "loss": 1.0453,
+      "step": 1488
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25640967867486597,
+      "learning_rate": 0.0001997534287803415,
+      "loss": 1.0959,
+      "step": 1489
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2817881208330018,
+      "learning_rate": 0.00019975231730764227,
+      "loss": 1.1004,
+      "step": 1490
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2632972390210799,
+      "learning_rate": 0.00019975120333857866,
+      "loss": 1.0682,
+      "step": 1491
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.256251122254461,
+      "learning_rate": 0.0001997500868731785,
+      "loss": 1.1663,
+      "step": 1492
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26242859708220295,
+      "learning_rate": 0.0001997489679114697,
+      "loss": 1.0843,
+      "step": 1493
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.28009382446859793,
+      "learning_rate": 0.0001997478464534803,
+      "loss": 1.1744,
+      "step": 1494
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2448192879010323,
+      "learning_rate": 0.0001997467224992383,
+      "loss": 1.1268,
+      "step": 1495
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.23428824139945728,
+      "learning_rate": 0.00019974559604877195,
+      "loss": 1.0997,
+      "step": 1496
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2833166934983789,
+      "learning_rate": 0.00019974446710210934,
+      "loss": 1.0867,
+      "step": 1497
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.25022212744952455,
+      "learning_rate": 0.00019974333565927878,
+      "loss": 1.0903,
+      "step": 1498
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26855972657496696,
+      "learning_rate": 0.00019974220172030852,
+      "loss": 1.0304,
+      "step": 1499
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.26473122846509034,
+      "learning_rate": 0.000199741065285227,
+      "loss": 1.0981,
+      "step": 1500
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2915333904654244,
+      "learning_rate": 0.00019973992635406265,
+      "loss": 1.157,
+      "step": 1501
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.37217781009539613,
+      "learning_rate": 0.00019973878492684393,
+      "loss": 1.1401,
+      "step": 1502
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2745430144825149,
+      "learning_rate": 0.00019973764100359942,
+      "loss": 1.1318,
+      "step": 1503
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2755977808247509,
+      "learning_rate": 0.0001997364945843578,
+      "loss": 1.1609,
+      "step": 1504
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2915733276431069,
+      "learning_rate": 0.00019973534566914772,
+      "loss": 1.086,
+      "step": 1505
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2599783190286603,
+      "learning_rate": 0.00019973419425799792,
+      "loss": 1.1108,
+      "step": 1506
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.280753933923634,
+      "learning_rate": 0.0001997330403509372,
+      "loss": 1.147,
+      "step": 1507
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27055165527737557,
+      "learning_rate": 0.0001997318839479945,
+      "loss": 1.0952,
+      "step": 1508
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2491475388699138,
+      "learning_rate": 0.00019973072504919875,
+      "loss": 1.0893,
+      "step": 1509
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.22529007248240643,
+      "learning_rate": 0.00019972956365457887,
+      "loss": 0.9453,
+      "step": 1510
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.29669549846388793,
+      "learning_rate": 0.000199728399764164,
+      "loss": 1.2225,
+      "step": 1511
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.27716026577726227,
+      "learning_rate": 0.00019972723337798327,
+      "loss": 1.2158,
+      "step": 1512
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2615727080933855,
+      "learning_rate": 0.00019972606449606583,
+      "loss": 1.1745,
+      "step": 1513
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2613301650841691,
+      "learning_rate": 0.00019972489311844097,
+      "loss": 1.0105,
+      "step": 1514
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.2637350666912871,
+      "learning_rate": 0.00019972371924513796,
+      "loss": 1.2445,
+      "step": 1515
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23442111345358183,
+      "learning_rate": 0.0001997225428761862,
+      "loss": 1.0897,
+      "step": 1516
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2790398743404188,
+      "learning_rate": 0.00019972136401161516,
+      "loss": 1.0976,
+      "step": 1517
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24161573374816125,
+      "learning_rate": 0.00019972018265145428,
+      "loss": 1.0778,
+      "step": 1518
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2599798006860897,
+      "learning_rate": 0.00019971899879573317,
+      "loss": 1.0797,
+      "step": 1519
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24106641987255334,
+      "learning_rate": 0.00019971781244448145,
+      "loss": 1.0863,
+      "step": 1520
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28462310372801436,
+      "learning_rate": 0.0001997166235977288,
+      "loss": 1.2261,
+      "step": 1521
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2715980154291507,
+      "learning_rate": 0.00019971543225550498,
+      "loss": 1.074,
+      "step": 1522
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24961681655649778,
+      "learning_rate": 0.0001997142384178398,
+      "loss": 1.102,
+      "step": 1523
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27808524258122086,
+      "learning_rate": 0.00019971304208476313,
+      "loss": 1.0456,
+      "step": 1524
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.272393710564596,
+      "learning_rate": 0.0001997118432563049,
+      "loss": 0.9979,
+      "step": 1525
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2985225834683353,
+      "learning_rate": 0.00019971064193249517,
+      "loss": 1.1062,
+      "step": 1526
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31751577230140715,
+      "learning_rate": 0.0001997094381133639,
+      "loss": 1.1574,
+      "step": 1527
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2311208784177862,
+      "learning_rate": 0.00019970823179894134,
+      "loss": 1.0605,
+      "step": 1528
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27395265358450693,
+      "learning_rate": 0.00019970702298925756,
+      "loss": 1.0394,
+      "step": 1529
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.31050969831593717,
+      "learning_rate": 0.0001997058116843429,
+      "loss": 1.21,
+      "step": 1530
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29740261342037966,
+      "learning_rate": 0.00019970459788422762,
+      "loss": 1.0973,
+      "step": 1531
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.30710726934897853,
+      "learning_rate": 0.00019970338158894213,
+      "loss": 1.0899,
+      "step": 1532
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38054688041895546,
+      "learning_rate": 0.00019970216279851686,
+      "loss": 1.0608,
+      "step": 1533
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26727890820952926,
+      "learning_rate": 0.0001997009415129823,
+      "loss": 1.1018,
+      "step": 1534
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29384471950665425,
+      "learning_rate": 0.000199699717732369,
+      "loss": 1.088,
+      "step": 1535
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.245541461264911,
+      "learning_rate": 0.00019969849145670763,
+      "loss": 1.0829,
+      "step": 1536
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3031137108785239,
+      "learning_rate": 0.0001996972626860288,
+      "loss": 1.138,
+      "step": 1537
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27477763946901834,
+      "learning_rate": 0.0001996960314203634,
+      "loss": 1.1612,
+      "step": 1538
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2970092386217553,
+      "learning_rate": 0.0001996947976597421,
+      "loss": 1.0688,
+      "step": 1539
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3258862801086761,
+      "learning_rate": 0.00019969356140419584,
+      "loss": 1.1302,
+      "step": 1540
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2995257919569518,
+      "learning_rate": 0.00019969232265375556,
+      "loss": 1.0475,
+      "step": 1541
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26050135512792294,
+      "learning_rate": 0.00019969108140845224,
+      "loss": 1.1356,
+      "step": 1542
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.258577882146825,
+      "learning_rate": 0.00019968983766831695,
+      "loss": 1.0545,
+      "step": 1543
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2627007677252112,
+      "learning_rate": 0.00019968859143338084,
+      "loss": 1.1386,
+      "step": 1544
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2633365267212909,
+      "learning_rate": 0.00019968734270367505,
+      "loss": 1.1784,
+      "step": 1545
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28411120690910896,
+      "learning_rate": 0.0001996860914792309,
+      "loss": 1.0888,
+      "step": 1546
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26646479947544477,
+      "learning_rate": 0.00019968483776007962,
+      "loss": 1.1659,
+      "step": 1547
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24786146098955286,
+      "learning_rate": 0.00019968358154625265,
+      "loss": 1.2185,
+      "step": 1548
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23607348527883257,
+      "learning_rate": 0.0001996823228377814,
+      "loss": 1.0735,
+      "step": 1549
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2570468565891483,
+      "learning_rate": 0.00019968106163469735,
+      "loss": 1.0616,
+      "step": 1550
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25852020457845143,
+      "learning_rate": 0.00019967979793703212,
+      "loss": 1.093,
+      "step": 1551
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28278864970040046,
+      "learning_rate": 0.00019967853174481727,
+      "loss": 1.0388,
+      "step": 1552
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2941535073545967,
+      "learning_rate": 0.00019967726305808453,
+      "loss": 1.0706,
+      "step": 1553
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2622528773036849,
+      "learning_rate": 0.00019967599187686562,
+      "loss": 1.0883,
+      "step": 1554
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27742024001398574,
+      "learning_rate": 0.00019967471820119242,
+      "loss": 1.0728,
+      "step": 1555
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3052593236911406,
+      "learning_rate": 0.00019967344203109671,
+      "loss": 1.1789,
+      "step": 1556
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28599434696605697,
+      "learning_rate": 0.0001996721633666105,
+      "loss": 1.1562,
+      "step": 1557
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24494683600954692,
+      "learning_rate": 0.00019967088220776573,
+      "loss": 1.2048,
+      "step": 1558
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25067138267095485,
+      "learning_rate": 0.0001996695985545945,
+      "loss": 1.1256,
+      "step": 1559
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3027501173185624,
+      "learning_rate": 0.00019966831240712893,
+      "loss": 1.1089,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29832312031178204,
+      "learning_rate": 0.0001996670237654012,
+      "loss": 1.1435,
+      "step": 1561
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2602686000520101,
+      "learning_rate": 0.00019966573262944357,
+      "loss": 1.2332,
+      "step": 1562
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2805057631189005,
+      "learning_rate": 0.00019966443899928831,
+      "loss": 1.0837,
+      "step": 1563
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38166616945412846,
+      "learning_rate": 0.00019966314287496787,
+      "loss": 1.059,
+      "step": 1564
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2764938914379305,
+      "learning_rate": 0.00019966184425651464,
+      "loss": 1.1139,
+      "step": 1565
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24932842836394434,
+      "learning_rate": 0.0001996605431439611,
+      "loss": 1.1142,
+      "step": 1566
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29196882597427254,
+      "learning_rate": 0.00019965923953733987,
+      "loss": 1.1607,
+      "step": 1567
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2709530754699249,
+      "learning_rate": 0.00019965793343668347,
+      "loss": 1.0495,
+      "step": 1568
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2598432982826659,
+      "learning_rate": 0.0001996566248420247,
+      "loss": 1.1288,
+      "step": 1569
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.46312766196550453,
+      "learning_rate": 0.00019965531375339628,
+      "loss": 1.1307,
+      "step": 1570
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26262906846828266,
+      "learning_rate": 0.00019965400017083097,
+      "loss": 1.1543,
+      "step": 1571
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25314149624310883,
+      "learning_rate": 0.00019965268409436168,
+      "loss": 1.0466,
+      "step": 1572
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2687356408453828,
+      "learning_rate": 0.00019965136552402136,
+      "loss": 1.1159,
+      "step": 1573
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2708428789609682,
+      "learning_rate": 0.00019965004445984298,
+      "loss": 1.055,
+      "step": 1574
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3839155117397319,
+      "learning_rate": 0.0001996487209018596,
+      "loss": 1.0844,
+      "step": 1575
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7734934544648255,
+      "learning_rate": 0.00019964739485010436,
+      "loss": 1.1704,
+      "step": 1576
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2985808342595769,
+      "learning_rate": 0.00019964606630461042,
+      "loss": 1.1233,
+      "step": 1577
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28199357321934904,
+      "learning_rate": 0.00019964473526541107,
+      "loss": 1.1306,
+      "step": 1578
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2708719269403245,
+      "learning_rate": 0.0001996434017325396,
+      "loss": 1.2283,
+      "step": 1579
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2792985547025937,
+      "learning_rate": 0.00019964206570602936,
+      "loss": 1.1385,
+      "step": 1580
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23415050247231725,
+      "learning_rate": 0.0001996407271859138,
+      "loss": 1.0389,
+      "step": 1581
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2740160527199919,
+      "learning_rate": 0.00019963938617222643,
+      "loss": 1.1243,
+      "step": 1582
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27399387661110913,
+      "learning_rate": 0.0001996380426650008,
+      "loss": 1.1087,
+      "step": 1583
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2692759658009343,
+      "learning_rate": 0.0001996366966642705,
+      "loss": 1.1163,
+      "step": 1584
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9607464989916183,
+      "learning_rate": 0.0001996353481700693,
+      "loss": 1.0442,
+      "step": 1585
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24459484067831805,
+      "learning_rate": 0.00019963399718243084,
+      "loss": 1.0841,
+      "step": 1586
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2762521737160385,
+      "learning_rate": 0.00019963264370138903,
+      "loss": 1.1496,
+      "step": 1587
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25933505174178323,
+      "learning_rate": 0.0001996312877269777,
+      "loss": 1.2522,
+      "step": 1588
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.27189774149978774,
+      "learning_rate": 0.00019962992925923073,
+      "loss": 1.1413,
+      "step": 1589
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.576402306240423,
+      "learning_rate": 0.00019962856829818223,
+      "loss": 1.1702,
+      "step": 1590
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28151915869781446,
+      "learning_rate": 0.00019962720484386614,
+      "loss": 1.1608,
+      "step": 1591
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.487170407807799,
+      "learning_rate": 0.00019962583889631663,
+      "loss": 1.1038,
+      "step": 1592
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25611542436265444,
+      "learning_rate": 0.00019962447045556792,
+      "loss": 1.0629,
+      "step": 1593
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24414987846237798,
+      "learning_rate": 0.00019962309952165425,
+      "loss": 1.0264,
+      "step": 1594
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26011148045017146,
+      "learning_rate": 0.00019962172609460982,
+      "loss": 1.0993,
+      "step": 1595
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2618849192631391,
+      "learning_rate": 0.00019962035017446916,
+      "loss": 1.1054,
+      "step": 1596
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24456979964789494,
+      "learning_rate": 0.0001996189717612666,
+      "loss": 1.1605,
+      "step": 1597
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2832924711395636,
+      "learning_rate": 0.00019961759085503666,
+      "loss": 1.1245,
+      "step": 1598
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.24882178646084718,
+      "learning_rate": 0.00019961620745581387,
+      "loss": 1.0725,
+      "step": 1599
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.22925300602806323,
+      "learning_rate": 0.00019961482156363296,
+      "loss": 1.0953,
+      "step": 1600
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2531531358177607,
+      "learning_rate": 0.00019961343317852846,
+      "loss": 1.0204,
+      "step": 1601
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2738032960535575,
+      "learning_rate": 0.00019961204230053525,
+      "loss": 1.1092,
+      "step": 1602
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2731400838916497,
+      "learning_rate": 0.00019961064892968806,
+      "loss": 1.1173,
+      "step": 1603
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2493741546517925,
+      "learning_rate": 0.00019960925306602176,
+      "loss": 1.0769,
+      "step": 1604
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.25050950474007155,
+      "learning_rate": 0.0001996078547095713,
+      "loss": 1.133,
+      "step": 1605
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29817105037980673,
+      "learning_rate": 0.0001996064538603717,
+      "loss": 1.1855,
+      "step": 1606
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.28263632425868207,
+      "learning_rate": 0.00019960505051845796,
+      "loss": 1.136,
+      "step": 1607
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.23133396852491225,
+      "learning_rate": 0.00019960364468386526,
+      "loss": 0.9476,
+      "step": 1608
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2775442108767544,
+      "learning_rate": 0.00019960223635662874,
+      "loss": 1.1606,
+      "step": 1609
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2805090162171342,
+      "learning_rate": 0.00019960082553678365,
+      "loss": 1.143,
+      "step": 1610
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2537997461349602,
+      "learning_rate": 0.0001995994122243653,
+      "loss": 1.1834,
+      "step": 1611
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.32560983069595756,
+      "learning_rate": 0.00019959799641940907,
+      "loss": 0.9919,
+      "step": 1612
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2777624913037504,
+      "learning_rate": 0.0001995965781219504,
+      "loss": 1.1872,
+      "step": 1613
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2758857420110973,
+      "learning_rate": 0.00019959515733202477,
+      "loss": 1.137,
+      "step": 1614
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.256755743287946,
+      "learning_rate": 0.0001995937340496677,
+      "loss": 1.0922,
+      "step": 1615
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2517079524462437,
+      "learning_rate": 0.00019959230827491488,
+      "loss": 1.0859,
+      "step": 1616
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.2724013994730525,
+      "learning_rate": 0.00019959088000780193,
+      "loss": 1.0288,
+      "step": 1617
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3008600441450719,
+      "learning_rate": 0.00019958944924836463,
+      "loss": 1.044,
+      "step": 1618
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.29350285017468836,
+      "learning_rate": 0.00019958801599663877,
+      "loss": 1.1309,
+      "step": 1619
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.26935905493782025,
+      "learning_rate": 0.0001995865802526602,
+      "loss": 1.1497,
+      "step": 1620
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2619749727571851,
+      "learning_rate": 0.0001995851420164649,
+      "loss": 1.226,
+      "step": 1621
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2566559504175784,
+      "learning_rate": 0.00019958370128808883,
+      "loss": 1.126,
+      "step": 1622
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2950868154557528,
+      "learning_rate": 0.00019958225806756806,
+      "loss": 1.1165,
+      "step": 1623
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.262703858238673,
+      "learning_rate": 0.00019958081235493867,
+      "loss": 1.1535,
+      "step": 1624
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27043560750628914,
+      "learning_rate": 0.00019957936415023687,
+      "loss": 1.1192,
+      "step": 1625
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.263242304281921,
+      "learning_rate": 0.00019957791345349892,
+      "loss": 1.0326,
+      "step": 1626
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2815449484727248,
+      "learning_rate": 0.0001995764602647611,
+      "loss": 1.0835,
+      "step": 1627
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2868280654211211,
+      "learning_rate": 0.00019957500458405976,
+      "loss": 1.1983,
+      "step": 1628
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.31775561559603943,
+      "learning_rate": 0.00019957354641143136,
+      "loss": 1.163,
+      "step": 1629
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2962143104484358,
+      "learning_rate": 0.00019957208574691238,
+      "loss": 1.2085,
+      "step": 1630
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.33134737467969283,
+      "learning_rate": 0.0001995706225905394,
+      "loss": 1.1736,
+      "step": 1631
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2973914160477879,
+      "learning_rate": 0.00019956915694234895,
+      "loss": 1.1877,
+      "step": 1632
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23797305266943983,
+      "learning_rate": 0.00019956768880237781,
+      "loss": 1.13,
+      "step": 1633
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28430135940967705,
+      "learning_rate": 0.0001995662181706627,
+      "loss": 1.1628,
+      "step": 1634
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2606274355294148,
+      "learning_rate": 0.00019956474504724038,
+      "loss": 1.1124,
+      "step": 1635
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26540643513755174,
+      "learning_rate": 0.00019956326943214775,
+      "loss": 1.0509,
+      "step": 1636
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25782459518811457,
+      "learning_rate": 0.00019956179132542173,
+      "loss": 1.0932,
+      "step": 1637
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24697183220142635,
+      "learning_rate": 0.00019956031072709932,
+      "loss": 1.2274,
+      "step": 1638
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2609501474414402,
+      "learning_rate": 0.0001995588276372175,
+      "loss": 1.0552,
+      "step": 1639
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2649049635321031,
+      "learning_rate": 0.00019955734205581352,
+      "loss": 0.9704,
+      "step": 1640
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2593017975988553,
+      "learning_rate": 0.00019955585398292447,
+      "loss": 1.1551,
+      "step": 1641
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25956189155243276,
+      "learning_rate": 0.0001995543634185876,
+      "loss": 1.1989,
+      "step": 1642
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2560637809135954,
+      "learning_rate": 0.0001995528703628402,
+      "loss": 1.1312,
+      "step": 1643
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27697911347482446,
+      "learning_rate": 0.00019955137481571968,
+      "loss": 1.2054,
+      "step": 1644
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24778928635339068,
+      "learning_rate": 0.00019954987677726343,
+      "loss": 1.1358,
+      "step": 1645
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26610781310481363,
+      "learning_rate": 0.00019954837624750895,
+      "loss": 1.1007,
+      "step": 1646
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2799296986376897,
+      "learning_rate": 0.0001995468732264938,
+      "loss": 1.0621,
+      "step": 1647
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26342475793330683,
+      "learning_rate": 0.00019954536771425556,
+      "loss": 1.1325,
+      "step": 1648
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2809166549010256,
+      "learning_rate": 0.00019954385971083193,
+      "loss": 1.0778,
+      "step": 1649
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28395265589643537,
+      "learning_rate": 0.00019954234921626068,
+      "loss": 0.9792,
+      "step": 1650
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3037040271080715,
+      "learning_rate": 0.00019954083623057955,
+      "loss": 1.1754,
+      "step": 1651
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23645934008105318,
+      "learning_rate": 0.00019953932075382646,
+      "loss": 1.1307,
+      "step": 1652
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2723890058669645,
+      "learning_rate": 0.00019953780278603932,
+      "loss": 1.1161,
+      "step": 1653
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2607964386960627,
+      "learning_rate": 0.00019953628232725608,
+      "loss": 1.1741,
+      "step": 1654
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27487552426540823,
+      "learning_rate": 0.0001995347593775148,
+      "loss": 1.131,
+      "step": 1655
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24819637701868438,
+      "learning_rate": 0.00019953323393685367,
+      "loss": 1.1246,
+      "step": 1656
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2880503601951672,
+      "learning_rate": 0.00019953170600531074,
+      "loss": 1.1414,
+      "step": 1657
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2661262998136285,
+      "learning_rate": 0.00019953017558292438,
+      "loss": 1.1857,
+      "step": 1658
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2660986350757843,
+      "learning_rate": 0.00019952864266973278,
+      "loss": 1.1092,
+      "step": 1659
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2713659116488214,
+      "learning_rate": 0.00019952710726577435,
+      "loss": 1.0772,
+      "step": 1660
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29223182156019084,
+      "learning_rate": 0.00019952556937108753,
+      "loss": 1.0789,
+      "step": 1661
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2582755115758311,
+      "learning_rate": 0.00019952402898571077,
+      "loss": 1.1875,
+      "step": 1662
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2654852860667591,
+      "learning_rate": 0.00019952248610968264,
+      "loss": 1.0323,
+      "step": 1663
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2935124336052021,
+      "learning_rate": 0.00019952094074304175,
+      "loss": 0.9859,
+      "step": 1664
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2546178333136953,
+      "learning_rate": 0.00019951939288582676,
+      "loss": 1.0756,
+      "step": 1665
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3002327247319587,
+      "learning_rate": 0.0001995178425380764,
+      "loss": 1.1356,
+      "step": 1666
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2792232404834516,
+      "learning_rate": 0.00019951628969982953,
+      "loss": 1.1251,
+      "step": 1667
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28042254510601033,
+      "learning_rate": 0.00019951473437112495,
+      "loss": 1.0406,
+      "step": 1668
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2869293910747367,
+      "learning_rate": 0.0001995131765520016,
+      "loss": 1.0321,
+      "step": 1669
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26731773733638436,
+      "learning_rate": 0.00019951161624249844,
+      "loss": 1.1865,
+      "step": 1670
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26494146088310505,
+      "learning_rate": 0.00019951005344265462,
+      "loss": 1.105,
+      "step": 1671
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26633126397973417,
+      "learning_rate": 0.0001995084881525091,
+      "loss": 1.0366,
+      "step": 1672
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27563381490550426,
+      "learning_rate": 0.00019950692037210113,
+      "loss": 1.1346,
+      "step": 1673
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2808572907767721,
+      "learning_rate": 0.00019950535010146994,
+      "loss": 1.1304,
+      "step": 1674
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.249082820217972,
+      "learning_rate": 0.00019950377734065486,
+      "loss": 1.1375,
+      "step": 1675
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.242030163260417,
+      "learning_rate": 0.00019950220208969519,
+      "loss": 1.0647,
+      "step": 1676
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23873021173771233,
+      "learning_rate": 0.00019950062434863038,
+      "loss": 1.1427,
+      "step": 1677
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2526857334515431,
+      "learning_rate": 0.00019949904411749995,
+      "loss": 1.0652,
+      "step": 1678
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22005647549892257,
+      "learning_rate": 0.00019949746139634336,
+      "loss": 1.2141,
+      "step": 1679
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29830008364083893,
+      "learning_rate": 0.0001994958761852003,
+      "loss": 0.9724,
+      "step": 1680
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2336810810860809,
+      "learning_rate": 0.00019949428848411036,
+      "loss": 1.0718,
+      "step": 1681
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2596500357999934,
+      "learning_rate": 0.00019949269829311336,
+      "loss": 1.1449,
+      "step": 1682
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28287000384537997,
+      "learning_rate": 0.00019949110561224905,
+      "loss": 1.1513,
+      "step": 1683
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27286503590527955,
+      "learning_rate": 0.00019948951044155728,
+      "loss": 1.095,
+      "step": 1684
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26619367233558344,
+      "learning_rate": 0.000199487912781078,
+      "loss": 1.0861,
+      "step": 1685
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29588681836665753,
+      "learning_rate": 0.0001994863126308512,
+      "loss": 1.0361,
+      "step": 1686
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27036346205815504,
+      "learning_rate": 0.00019948470999091685,
+      "loss": 1.1066,
+      "step": 1687
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24182695997990206,
+      "learning_rate": 0.00019948310486131513,
+      "loss": 0.9643,
+      "step": 1688
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26214221449152814,
+      "learning_rate": 0.0001994814972420862,
+      "loss": 1.1246,
+      "step": 1689
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2804122217034823,
+      "learning_rate": 0.0001994798871332703,
+      "loss": 1.1795,
+      "step": 1690
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2594172377288309,
+      "learning_rate": 0.00019947827453490767,
+      "loss": 1.1388,
+      "step": 1691
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2893954998932536,
+      "learning_rate": 0.0001994766594470387,
+      "loss": 1.0079,
+      "step": 1692
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2936972888903826,
+      "learning_rate": 0.0001994750418697038,
+      "loss": 1.1258,
+      "step": 1693
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26900132587121695,
+      "learning_rate": 0.00019947342180294346,
+      "loss": 1.2546,
+      "step": 1694
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26577176120374074,
+      "learning_rate": 0.00019947179924679825,
+      "loss": 1.1571,
+      "step": 1695
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2734445337941885,
+      "learning_rate": 0.00019947017420130872,
+      "loss": 1.0859,
+      "step": 1696
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24000154406623206,
+      "learning_rate": 0.0001994685466665156,
+      "loss": 1.1031,
+      "step": 1697
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.22779641522136282,
+      "learning_rate": 0.00019946691664245956,
+      "loss": 1.0854,
+      "step": 1698
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2479814495549998,
+      "learning_rate": 0.0001994652841291814,
+      "loss": 1.1121,
+      "step": 1699
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.250268222311568,
+      "learning_rate": 0.00019946364912672203,
+      "loss": 1.0779,
+      "step": 1700
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.24912184436362922,
+      "learning_rate": 0.00019946201163512233,
+      "loss": 1.0067,
+      "step": 1701
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3207294751787716,
+      "learning_rate": 0.00019946037165442327,
+      "loss": 1.303,
+      "step": 1702
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28907607252287926,
+      "learning_rate": 0.0001994587291846659,
+      "loss": 1.1877,
+      "step": 1703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.25848051240176273,
+      "learning_rate": 0.0001994570842258913,
+      "loss": 1.0081,
+      "step": 1704
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2756315421016595,
+      "learning_rate": 0.00019945543677814067,
+      "loss": 1.0329,
+      "step": 1705
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2505304319664102,
+      "learning_rate": 0.00019945378684145526,
+      "loss": 1.1502,
+      "step": 1706
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3140138337651653,
+      "learning_rate": 0.00019945213441587633,
+      "loss": 1.108,
+      "step": 1707
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2650198799790111,
+      "learning_rate": 0.0001994504795014452,
+      "loss": 1.0843,
+      "step": 1708
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.27741715785219834,
+      "learning_rate": 0.00019944882209820333,
+      "loss": 1.0471,
+      "step": 1709
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28688576394081394,
+      "learning_rate": 0.0001994471622061922,
+      "loss": 1.0745,
+      "step": 1710
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3160625561308828,
+      "learning_rate": 0.0001994454998254533,
+      "loss": 1.0412,
+      "step": 1711
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.29390242181574067,
+      "learning_rate": 0.0001994438349560283,
+      "loss": 1.0297,
+      "step": 1712
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2662667067917202,
+      "learning_rate": 0.00019944216759795885,
+      "loss": 1.0189,
+      "step": 1713
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2652127225373576,
+      "learning_rate": 0.00019944049775128661,
+      "loss": 1.0433,
+      "step": 1714
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.28198238645107127,
+      "learning_rate": 0.00019943882541605343,
+      "loss": 1.1984,
+      "step": 1715
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2930815699661024,
+      "learning_rate": 0.00019943715059230117,
+      "loss": 1.0741,
+      "step": 1716
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.32027322958606264,
+      "learning_rate": 0.0001994354732800717,
+      "loss": 1.097,
+      "step": 1717
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23663660962171493,
+      "learning_rate": 0.00019943379347940704,
+      "loss": 1.0879,
+      "step": 1718
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26819927151645284,
+      "learning_rate": 0.0001994321111903492,
+      "loss": 1.1009,
+      "step": 1719
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23575422949387753,
+      "learning_rate": 0.00019943042641294028,
+      "loss": 1.1155,
+      "step": 1720
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23615429167267799,
+      "learning_rate": 0.00019942873914722243,
+      "loss": 1.1978,
+      "step": 1721
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.23075710530915816,
+      "learning_rate": 0.00019942704939323794,
+      "loss": 1.1802,
+      "step": 1722
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.26619797669604706,
+      "learning_rate": 0.00019942535715102903,
+      "loss": 1.2049,
+      "step": 1723
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.2766557794492051,
+      "learning_rate": 0.00019942366242063807,
+      "loss": 1.1062,
+      "step": 1724
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2650480870677421,
+      "learning_rate": 0.00019942196520210748,
+      "loss": 1.2134,
+      "step": 1725
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25567802198203454,
+      "learning_rate": 0.00019942026549547973,
+      "loss": 1.1747,
+      "step": 1726
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27010898900806307,
+      "learning_rate": 0.00019941856330079732,
+      "loss": 1.2341,
+      "step": 1727
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2672243888343421,
+      "learning_rate": 0.0001994168586181029,
+      "loss": 1.038,
+      "step": 1728
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27037675730748634,
+      "learning_rate": 0.00019941515144743913,
+      "loss": 1.1484,
+      "step": 1729
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27423511003272355,
+      "learning_rate": 0.00019941344178884868,
+      "loss": 1.1925,
+      "step": 1730
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25846818596543564,
+      "learning_rate": 0.0001994117296423744,
+      "loss": 1.0063,
+      "step": 1731
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2686589532404364,
+      "learning_rate": 0.0001994100150080591,
+      "loss": 1.0285,
+      "step": 1732
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2766088906498271,
+      "learning_rate": 0.00019940829788594569,
+      "loss": 1.0765,
+      "step": 1733
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26339315358801624,
+      "learning_rate": 0.00019940657827607715,
+      "loss": 1.1692,
+      "step": 1734
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2603560134384319,
+      "learning_rate": 0.0001994048561784965,
+      "loss": 1.0406,
+      "step": 1735
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2904525258177709,
+      "learning_rate": 0.0001994031315932469,
+      "loss": 0.9866,
+      "step": 1736
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.254967597751732,
+      "learning_rate": 0.00019940140452037142,
+      "loss": 1.1711,
+      "step": 1737
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2587539173182572,
+      "learning_rate": 0.00019939967495991332,
+      "loss": 1.1377,
+      "step": 1738
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26239329017748414,
+      "learning_rate": 0.0001993979429119159,
+      "loss": 1.0846,
+      "step": 1739
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2641030306092387,
+      "learning_rate": 0.00019939620837642247,
+      "loss": 1.1515,
+      "step": 1740
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29629448536301656,
+      "learning_rate": 0.00019939447135347647,
+      "loss": 1.1464,
+      "step": 1741
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.259586106165695,
+      "learning_rate": 0.00019939273184312137,
+      "loss": 1.0899,
+      "step": 1742
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26296382387837713,
+      "learning_rate": 0.0001993909898454007,
+      "loss": 1.0482,
+      "step": 1743
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28051523053769833,
+      "learning_rate": 0.000199389245360358,
+      "loss": 1.1547,
+      "step": 1744
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2671245457129155,
+      "learning_rate": 0.00019938749838803696,
+      "loss": 1.0592,
+      "step": 1745
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3181482392262407,
+      "learning_rate": 0.00019938574892848135,
+      "loss": 1.0635,
+      "step": 1746
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2899294422265618,
+      "learning_rate": 0.0001993839969817349,
+      "loss": 1.0768,
+      "step": 1747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25002771184882505,
+      "learning_rate": 0.00019938224254784147,
+      "loss": 1.1619,
+      "step": 1748
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2568140049401059,
+      "learning_rate": 0.00019938048562684495,
+      "loss": 1.0229,
+      "step": 1749
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.36292519666598,
+      "learning_rate": 0.00019937872621878934,
+      "loss": 1.1794,
+      "step": 1750
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2932029976728667,
+      "learning_rate": 0.0001993769643237186,
+      "loss": 1.1902,
+      "step": 1751
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2771533942225469,
+      "learning_rate": 0.00019937519994167694,
+      "loss": 1.0731,
+      "step": 1752
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.30070233491956233,
+      "learning_rate": 0.00019937343307270842,
+      "loss": 1.1186,
+      "step": 1753
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23520740536849988,
+      "learning_rate": 0.00019937166371685727,
+      "loss": 1.0768,
+      "step": 1754
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2689615596060478,
+      "learning_rate": 0.0001993698918741678,
+      "loss": 1.0889,
+      "step": 1755
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2667096465377454,
+      "learning_rate": 0.0001993681175446843,
+      "loss": 1.1273,
+      "step": 1756
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2774024784883662,
+      "learning_rate": 0.00019936634072845126,
+      "loss": 1.0687,
+      "step": 1757
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2948555113393165,
+      "learning_rate": 0.00019936456142551306,
+      "loss": 1.1369,
+      "step": 1758
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2783762471401322,
+      "learning_rate": 0.00019936277963591428,
+      "loss": 1.1218,
+      "step": 1759
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24719261245618435,
+      "learning_rate": 0.00019936099535969946,
+      "loss": 1.1024,
+      "step": 1760
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3182614470412637,
+      "learning_rate": 0.00019935920859691332,
+      "loss": 1.1849,
+      "step": 1761
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23710818113640875,
+      "learning_rate": 0.00019935741934760053,
+      "loss": 1.0466,
+      "step": 1762
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2638247973270966,
+      "learning_rate": 0.00019935562761180586,
+      "loss": 1.046,
+      "step": 1763
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29051740614485994,
+      "learning_rate": 0.0001993538333895742,
+      "loss": 1.0124,
+      "step": 1764
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.344576484247007,
+      "learning_rate": 0.0001993520366809504,
+      "loss": 1.0553,
+      "step": 1765
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2734046389788186,
+      "learning_rate": 0.00019935023748597942,
+      "loss": 1.1175,
+      "step": 1766
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2658344263426795,
+      "learning_rate": 0.00019934843580470633,
+      "loss": 1.1288,
+      "step": 1767
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24386557141888807,
+      "learning_rate": 0.0001993466316371762,
+      "loss": 1.0556,
+      "step": 1768
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2903532438407926,
+      "learning_rate": 0.00019934482498343417,
+      "loss": 1.1214,
+      "step": 1769
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23273429637205256,
+      "learning_rate": 0.00019934301584352543,
+      "loss": 1.1819,
+      "step": 1770
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26436150903465644,
+      "learning_rate": 0.0001993412042174953,
+      "loss": 1.2129,
+      "step": 1771
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.272699055168821,
+      "learning_rate": 0.00019933939010538914,
+      "loss": 1.0845,
+      "step": 1772
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24313514050010182,
+      "learning_rate": 0.00019933757350725227,
+      "loss": 1.0947,
+      "step": 1773
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2743855476662608,
+      "learning_rate": 0.00019933575442313022,
+      "loss": 1.0721,
+      "step": 1774
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27484803850628725,
+      "learning_rate": 0.00019933393285306847,
+      "loss": 1.086,
+      "step": 1775
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2862674448464463,
+      "learning_rate": 0.0001993321087971126,
+      "loss": 1.1687,
+      "step": 1776
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2654433158884731,
+      "learning_rate": 0.00019933028225530832,
+      "loss": 1.1524,
+      "step": 1777
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28570985555687595,
+      "learning_rate": 0.00019932845322770127,
+      "loss": 1.1032,
+      "step": 1778
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2398384669725776,
+      "learning_rate": 0.00019932662171433726,
+      "loss": 1.1805,
+      "step": 1779
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27132648388882097,
+      "learning_rate": 0.00019932478771526212,
+      "loss": 1.1706,
+      "step": 1780
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2490907055488986,
+      "learning_rate": 0.00019932295123052175,
+      "loss": 1.0303,
+      "step": 1781
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2483810784888273,
+      "learning_rate": 0.0001993211122601621,
+      "loss": 1.183,
+      "step": 1782
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.23864167216181773,
+      "learning_rate": 0.00019931927080422921,
+      "loss": 1.0438,
+      "step": 1783
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27530929268313675,
+      "learning_rate": 0.0001993174268627691,
+      "loss": 1.0515,
+      "step": 1784
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24646180451212257,
+      "learning_rate": 0.00019931558043582802,
+      "loss": 1.1064,
+      "step": 1785
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3072867379708468,
+      "learning_rate": 0.00019931373152345206,
+      "loss": 1.0433,
+      "step": 1786
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2699984246251364,
+      "learning_rate": 0.0001993118801256876,
+      "loss": 1.2135,
+      "step": 1787
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29491554910347884,
+      "learning_rate": 0.00019931002624258093,
+      "loss": 1.0451,
+      "step": 1788
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2592446364845811,
+      "learning_rate": 0.00019930816987417843,
+      "loss": 1.0678,
+      "step": 1789
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27551160462797714,
+      "learning_rate": 0.00019930631102052656,
+      "loss": 1.099,
+      "step": 1790
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3004312874610158,
+      "learning_rate": 0.00019930444968167184,
+      "loss": 1.2349,
+      "step": 1791
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2902276940371733,
+      "learning_rate": 0.00019930258585766083,
+      "loss": 1.0191,
+      "step": 1792
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2880266394681308,
+      "learning_rate": 0.00019930071954854026,
+      "loss": 1.0946,
+      "step": 1793
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27739961453298684,
+      "learning_rate": 0.00019929885075435673,
+      "loss": 1.0883,
+      "step": 1794
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26237306161853147,
+      "learning_rate": 0.00019929697947515705,
+      "loss": 1.13,
+      "step": 1795
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.25732915392100997,
+      "learning_rate": 0.0001992951057109881,
+      "loss": 1.0223,
+      "step": 1796
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26231079106015476,
+      "learning_rate": 0.00019929322946189669,
+      "loss": 1.2334,
+      "step": 1797
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2689310820422752,
+      "learning_rate": 0.00019929135072792979,
+      "loss": 1.0859,
+      "step": 1798
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28231810159636933,
+      "learning_rate": 0.00019928946950913446,
+      "loss": 1.1499,
+      "step": 1799
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26114756115035764,
+      "learning_rate": 0.00019928758580555777,
+      "loss": 1.0692,
+      "step": 1800
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.21499560470777954,
+      "learning_rate": 0.00019928569961724684,
+      "loss": 1.0246,
+      "step": 1801
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2499670112727543,
+      "learning_rate": 0.00019928381094424887,
+      "loss": 1.0571,
+      "step": 1802
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2668951326275685,
+      "learning_rate": 0.00019928191978661112,
+      "loss": 1.0914,
+      "step": 1803
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2436534609761875,
+      "learning_rate": 0.00019928002614438096,
+      "loss": 1.115,
+      "step": 1804
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24543054408218462,
+      "learning_rate": 0.00019927813001760573,
+      "loss": 1.1398,
+      "step": 1805
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.275192488837919,
+      "learning_rate": 0.0001992762314063329,
+      "loss": 1.157,
+      "step": 1806
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.24910350712903606,
+      "learning_rate": 0.00019927433031061,
+      "loss": 1.1897,
+      "step": 1807
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2440628864265639,
+      "learning_rate": 0.0001992724267304846,
+      "loss": 1.1515,
+      "step": 1808
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2878309034150722,
+      "learning_rate": 0.0001992705206660043,
+      "loss": 1.2264,
+      "step": 1809
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2910816442764025,
+      "learning_rate": 0.00019926861211721684,
+      "loss": 1.0837,
+      "step": 1810
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.26250435365108804,
+      "learning_rate": 0.00019926670108416997,
+      "loss": 1.2048,
+      "step": 1811
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2870815342486622,
+      "learning_rate": 0.00019926478756691153,
+      "loss": 1.1559,
+      "step": 1812
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27090400410537946,
+      "learning_rate": 0.0001992628715654894,
+      "loss": 1.2068,
+      "step": 1813
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28120763276720057,
+      "learning_rate": 0.0001992609530799515,
+      "loss": 1.0354,
+      "step": 1814
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2894660442297703,
+      "learning_rate": 0.0001992590321103459,
+      "loss": 1.0584,
+      "step": 1815
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2780933045178876,
+      "learning_rate": 0.00019925710865672063,
+      "loss": 1.2506,
+      "step": 1816
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.28692036245587266,
+      "learning_rate": 0.0001992551827191238,
+      "loss": 1.1621,
+      "step": 1817
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2982135909302484,
+      "learning_rate": 0.00019925325429760368,
+      "loss": 1.0948,
+      "step": 1818
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2820173648746567,
+      "learning_rate": 0.0001992513233922085,
+      "loss": 1.1477,
+      "step": 1819
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2639471711581256,
+      "learning_rate": 0.00019924939000298656,
+      "loss": 1.0992,
+      "step": 1820
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3258628877911058,
+      "learning_rate": 0.00019924745412998625,
+      "loss": 1.0337,
+      "step": 1821
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32533873777010336,
+      "learning_rate": 0.00019924551577325605,
+      "loss": 1.036,
+      "step": 1822
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29923641265792245,
+      "learning_rate": 0.00019924357493284443,
+      "loss": 1.1199,
+      "step": 1823
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2871942608589218,
+      "learning_rate": 0.00019924163160879997,
+      "loss": 1.111,
+      "step": 1824
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.32570726041176123,
+      "learning_rate": 0.0001992396858011713,
+      "loss": 1.208,
+      "step": 1825
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.2745227229675084,
+      "learning_rate": 0.00019923773751000714,
+      "loss": 1.0936,
+      "step": 1826
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.29321444725480744,
+      "learning_rate": 0.00019923578673535622,
+      "loss": 1.1939,
+      "step": 1827
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.31320266548147835,
+      "learning_rate": 0.0001992338334772674,
+      "loss": 1.092,
+      "step": 1828
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.27134903395742416,
+      "learning_rate": 0.0001992318777357895,
+      "loss": 1.1936,
+      "step": 1829
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25932835062459336,
+      "learning_rate": 0.0001992299195109715,
+      "loss": 1.1709,
+      "step": 1830
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2821426140168247,
+      "learning_rate": 0.0001992279588028624,
+      "loss": 1.0848,
+      "step": 1831
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27780532527873364,
+      "learning_rate": 0.00019922599561151126,
+      "loss": 1.0701,
+      "step": 1832
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3213413443644143,
+      "learning_rate": 0.00019922402993696725,
+      "loss": 1.2066,
+      "step": 1833
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24152642775118963,
+      "learning_rate": 0.00019922206177927948,
+      "loss": 1.0779,
+      "step": 1834
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26358149683834353,
+      "learning_rate": 0.00019922009113849728,
+      "loss": 1.0631,
+      "step": 1835
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2771938797580255,
+      "learning_rate": 0.00019921811801466995,
+      "loss": 1.0627,
+      "step": 1836
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24390434386104548,
+      "learning_rate": 0.00019921614240784688,
+      "loss": 1.0826,
+      "step": 1837
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2883536207722251,
+      "learning_rate": 0.00019921416431807748,
+      "loss": 1.0587,
+      "step": 1838
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26058794322850115,
+      "learning_rate": 0.00019921218374541124,
+      "loss": 1.0926,
+      "step": 1839
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.253199761945334,
+      "learning_rate": 0.00019921020068989776,
+      "loss": 1.0659,
+      "step": 1840
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27605817699682705,
+      "learning_rate": 0.00019920821515158666,
+      "loss": 1.0807,
+      "step": 1841
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.34296685061041043,
+      "learning_rate": 0.0001992062271305276,
+      "loss": 1.0399,
+      "step": 1842
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24791123495573145,
+      "learning_rate": 0.0001992042366267704,
+      "loss": 1.0986,
+      "step": 1843
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2730759255331824,
+      "learning_rate": 0.0001992022436403648,
+      "loss": 1.1249,
+      "step": 1844
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25825382605152414,
+      "learning_rate": 0.0001992002481713607,
+      "loss": 1.1515,
+      "step": 1845
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2709947998255462,
+      "learning_rate": 0.0001991982502198081,
+      "loss": 1.0644,
+      "step": 1846
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3041130032344396,
+      "learning_rate": 0.0001991962497857569,
+      "loss": 1.2027,
+      "step": 1847
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2514054383827712,
+      "learning_rate": 0.00019919424686925722,
+      "loss": 1.1181,
+      "step": 1848
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2671075622096655,
+      "learning_rate": 0.00019919224147035914,
+      "loss": 1.0748,
+      "step": 1849
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2563779371324541,
+      "learning_rate": 0.00019919023358911292,
+      "loss": 1.1708,
+      "step": 1850
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27518701288612535,
+      "learning_rate": 0.00019918822322556877,
+      "loss": 1.068,
+      "step": 1851
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23464980973730304,
+      "learning_rate": 0.00019918621037977693,
+      "loss": 1.115,
+      "step": 1852
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22754310999457095,
+      "learning_rate": 0.0001991841950517879,
+      "loss": 1.143,
+      "step": 1853
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2402610477626132,
+      "learning_rate": 0.00019918217724165205,
+      "loss": 1.045,
+      "step": 1854
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2466867783180613,
+      "learning_rate": 0.00019918015694941988,
+      "loss": 1.1413,
+      "step": 1855
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27021081362108734,
+      "learning_rate": 0.00019917813417514194,
+      "loss": 1.1366,
+      "step": 1856
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2290266939618748,
+      "learning_rate": 0.00019917610891886884,
+      "loss": 1.077,
+      "step": 1857
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2912099506203328,
+      "learning_rate": 0.0001991740811806513,
+      "loss": 1.2381,
+      "step": 1858
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25710098640782725,
+      "learning_rate": 0.00019917205096054005,
+      "loss": 1.1494,
+      "step": 1859
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23444875824179745,
+      "learning_rate": 0.00019917001825858592,
+      "loss": 1.1993,
+      "step": 1860
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27972502638604985,
+      "learning_rate": 0.00019916798307483973,
+      "loss": 1.0881,
+      "step": 1861
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.22927317304767958,
+      "learning_rate": 0.00019916594540935246,
+      "loss": 1.1226,
+      "step": 1862
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2777824230056472,
+      "learning_rate": 0.00019916390526217507,
+      "loss": 1.2791,
+      "step": 1863
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29153525342371706,
+      "learning_rate": 0.0001991618626333586,
+      "loss": 1.1707,
+      "step": 1864
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26506907160815607,
+      "learning_rate": 0.00019915981752295422,
+      "loss": 1.1309,
+      "step": 1865
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24281279217203466,
+      "learning_rate": 0.00019915776993101311,
+      "loss": 1.117,
+      "step": 1866
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2915237131661395,
+      "learning_rate": 0.00019915571985758645,
+      "loss": 1.1615,
+      "step": 1867
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26598319618634586,
+      "learning_rate": 0.00019915366730272562,
+      "loss": 1.2443,
+      "step": 1868
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25083927497077174,
+      "learning_rate": 0.00019915161226648193,
+      "loss": 1.091,
+      "step": 1869
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24871843858798745,
+      "learning_rate": 0.00019914955474890683,
+      "loss": 1.2225,
+      "step": 1870
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2807205864271564,
+      "learning_rate": 0.00019914749475005182,
+      "loss": 1.0856,
+      "step": 1871
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2564207640831039,
+      "learning_rate": 0.00019914543226996846,
+      "loss": 1.1381,
+      "step": 1872
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2711542979084927,
+      "learning_rate": 0.00019914336730870828,
+      "loss": 1.1482,
+      "step": 1873
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24605707766197607,
+      "learning_rate": 0.00019914129986632308,
+      "loss": 1.0468,
+      "step": 1874
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25895981792303296,
+      "learning_rate": 0.00019913922994286453,
+      "loss": 1.1124,
+      "step": 1875
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26071698636999885,
+      "learning_rate": 0.00019913715753838444,
+      "loss": 1.0977,
+      "step": 1876
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26270406723403217,
+      "learning_rate": 0.00019913508265293468,
+      "loss": 0.9724,
+      "step": 1877
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2729621283459525,
+      "learning_rate": 0.00019913300528656718,
+      "loss": 1.0379,
+      "step": 1878
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2667009672221466,
+      "learning_rate": 0.00019913092543933392,
+      "loss": 1.201,
+      "step": 1879
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3062682088725694,
+      "learning_rate": 0.00019912884311128692,
+      "loss": 1.2133,
+      "step": 1880
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2471603219169982,
+      "learning_rate": 0.00019912675830247834,
+      "loss": 1.0426,
+      "step": 1881
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2706944754253154,
+      "learning_rate": 0.00019912467101296035,
+      "loss": 1.0739,
+      "step": 1882
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23877771124529687,
+      "learning_rate": 0.00019912258124278517,
+      "loss": 1.1371,
+      "step": 1883
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2501800385911424,
+      "learning_rate": 0.00019912048899200507,
+      "loss": 0.985,
+      "step": 1884
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2543490544283289,
+      "learning_rate": 0.00019911839426067245,
+      "loss": 1.0922,
+      "step": 1885
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2610356534075215,
+      "learning_rate": 0.0001991162970488397,
+      "loss": 1.062,
+      "step": 1886
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31809853336105126,
+      "learning_rate": 0.0001991141973565594,
+      "loss": 1.0172,
+      "step": 1887
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3037942696710212,
+      "learning_rate": 0.00019911209518388393,
+      "loss": 1.2201,
+      "step": 1888
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2612653091466676,
+      "learning_rate": 0.00019910999053086604,
+      "loss": 1.1529,
+      "step": 1889
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25927362339534166,
+      "learning_rate": 0.00019910788339755833,
+      "loss": 1.1727,
+      "step": 1890
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26339036209636246,
+      "learning_rate": 0.00019910577378401355,
+      "loss": 1.0759,
+      "step": 1891
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.26180032269848985,
+      "learning_rate": 0.00019910366169028452,
+      "loss": 1.0782,
+      "step": 1892
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27889254421508974,
+      "learning_rate": 0.00019910154711642403,
+      "loss": 1.1011,
+      "step": 1893
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2549984542896029,
+      "learning_rate": 0.00019909943006248505,
+      "loss": 1.15,
+      "step": 1894
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2274655301809786,
+      "learning_rate": 0.0001990973105285206,
+      "loss": 1.1843,
+      "step": 1895
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.259196981448463,
+      "learning_rate": 0.00019909518851458363,
+      "loss": 1.0451,
+      "step": 1896
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.37452396398591586,
+      "learning_rate": 0.0001990930640207273,
+      "loss": 1.1319,
+      "step": 1897
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30511870306850003,
+      "learning_rate": 0.00019909093704700473,
+      "loss": 1.1613,
+      "step": 1898
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29479478078757926,
+      "learning_rate": 0.00019908880759346925,
+      "loss": 1.1725,
+      "step": 1899
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3086910524148668,
+      "learning_rate": 0.00019908667566017406,
+      "loss": 1.1686,
+      "step": 1900
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2974264651570825,
+      "learning_rate": 0.0001990845412471725,
+      "loss": 1.1257,
+      "step": 1901
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.256728492572231,
+      "learning_rate": 0.00019908240435451805,
+      "loss": 1.0166,
+      "step": 1902
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.23937868610540455,
+      "learning_rate": 0.00019908026498226418,
+      "loss": 1.1205,
+      "step": 1903
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.272693028176549,
+      "learning_rate": 0.00019907812313046437,
+      "loss": 1.1055,
+      "step": 1904
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27354401730685546,
+      "learning_rate": 0.00019907597879917227,
+      "loss": 1.1253,
+      "step": 1905
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27305416183963116,
+      "learning_rate": 0.00019907383198844157,
+      "loss": 1.0841,
+      "step": 1906
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.31852024294594067,
+      "learning_rate": 0.00019907168269832592,
+      "loss": 1.1546,
+      "step": 1907
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.29709692567412604,
+      "learning_rate": 0.00019906953092887916,
+      "loss": 1.2313,
+      "step": 1908
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2617821076298669,
+      "learning_rate": 0.00019906737668015515,
+      "loss": 0.943,
+      "step": 1909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2636246176724674,
+      "learning_rate": 0.00019906521995220774,
+      "loss": 1.0627,
+      "step": 1910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2679413444345284,
+      "learning_rate": 0.00019906306074509095,
+      "loss": 1.1503,
+      "step": 1911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24358831048061988,
+      "learning_rate": 0.0001990608990588588,
+      "loss": 1.0467,
+      "step": 1912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2873239577441084,
+      "learning_rate": 0.0001990587348935654,
+      "loss": 1.2148,
+      "step": 1913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2970349427468681,
+      "learning_rate": 0.00019905656824926492,
+      "loss": 1.1718,
+      "step": 1914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.328114785718262,
+      "learning_rate": 0.00019905439912601156,
+      "loss": 1.0894,
+      "step": 1915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2768655282500371,
+      "learning_rate": 0.00019905222752385958,
+      "loss": 0.9798,
+      "step": 1916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24581149927304233,
+      "learning_rate": 0.00019905005344286338,
+      "loss": 1.1947,
+      "step": 1917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24905142815716402,
+      "learning_rate": 0.00019904787688307735,
+      "loss": 1.0603,
+      "step": 1918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25006568481196073,
+      "learning_rate": 0.00019904569784455592,
+      "loss": 1.1451,
+      "step": 1919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.24640239497730002,
+      "learning_rate": 0.0001990435163273537,
+      "loss": 1.1513,
+      "step": 1920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3158678665771197,
+      "learning_rate": 0.00019904133233152518,
+      "loss": 1.1675,
+      "step": 1921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2925864535264506,
+      "learning_rate": 0.0001990391458571251,
+      "loss": 1.0303,
+      "step": 1922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2919025168815423,
+      "learning_rate": 0.00019903695690420817,
+      "loss": 1.2033,
+      "step": 1923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2537949728039277,
+      "learning_rate": 0.00019903476547282914,
+      "loss": 1.144,
+      "step": 1924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2583763269969566,
+      "learning_rate": 0.00019903257156304285,
+      "loss": 1.1037,
+      "step": 1925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2613319587588986,
+      "learning_rate": 0.00019903037517490422,
+      "loss": 1.0958,
+      "step": 1926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2595089644687432,
+      "learning_rate": 0.00019902817630846822,
+      "loss": 1.1155,
+      "step": 1927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.25297594336763907,
+      "learning_rate": 0.00019902597496378985,
+      "loss": 1.1028,
+      "step": 1928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.30513489135176775,
+      "learning_rate": 0.00019902377114092425,
+      "loss": 1.1394,
+      "step": 1929
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.3021684093102426,
+      "learning_rate": 0.00019902156483992653,
+      "loss": 0.9847,
+      "step": 1930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2590031991887478,
+      "learning_rate": 0.00019901935606085193,
+      "loss": 1.036,
+      "step": 1931
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.27555119779022413,
+      "learning_rate": 0.00019901714480375572,
+      "loss": 1.0828,
+      "step": 1932
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.2590125935515052,
+      "learning_rate": 0.0001990149310686932,
+      "loss": 1.0109,
+      "step": 1933
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27535539380394297,
+      "learning_rate": 0.0001990127148557198,
+      "loss": 1.0645,
+      "step": 1934
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2604038154667027,
+      "learning_rate": 0.000199010496164891,
+      "loss": 1.1085,
+      "step": 1935
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24979308709369744,
+      "learning_rate": 0.0001990082749962623,
+      "loss": 1.084,
+      "step": 1936
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3032327587240259,
+      "learning_rate": 0.0001990060513498893,
+      "loss": 1.1131,
+      "step": 1937
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2678255255634505,
+      "learning_rate": 0.00019900382522582765,
+      "loss": 1.1368,
+      "step": 1938
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2180195498119384,
+      "learning_rate": 0.00019900159662413305,
+      "loss": 1.0595,
+      "step": 1939
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23825147073097672,
+      "learning_rate": 0.00019899936554486128,
+      "loss": 1.1574,
+      "step": 1940
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23730357351239173,
+      "learning_rate": 0.00019899713198806812,
+      "loss": 1.1184,
+      "step": 1941
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2805064829854749,
+      "learning_rate": 0.00019899489595380957,
+      "loss": 1.2007,
+      "step": 1942
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2473758737702457,
+      "learning_rate": 0.00019899265744214152,
+      "loss": 1.0602,
+      "step": 1943
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2718909446342109,
+      "learning_rate": 0.00019899041645312,
+      "loss": 1.1384,
+      "step": 1944
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2541735443801485,
+      "learning_rate": 0.0001989881729868011,
+      "loss": 1.0676,
+      "step": 1945
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2500138203165472,
+      "learning_rate": 0.00019898592704324094,
+      "loss": 1.0983,
+      "step": 1946
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2593377151457589,
+      "learning_rate": 0.00019898367862249575,
+      "loss": 1.0257,
+      "step": 1947
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2495745540042679,
+      "learning_rate": 0.00019898142772462182,
+      "loss": 1.0384,
+      "step": 1948
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2969538903134302,
+      "learning_rate": 0.00019897917434967544,
+      "loss": 1.1127,
+      "step": 1949
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26632242247343,
+      "learning_rate": 0.00019897691849771301,
+      "loss": 1.1186,
+      "step": 1950
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29353632758131,
+      "learning_rate": 0.00019897466016879098,
+      "loss": 1.0999,
+      "step": 1951
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2888464592603483,
+      "learning_rate": 0.00019897239936296588,
+      "loss": 1.0546,
+      "step": 1952
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2576028803170492,
+      "learning_rate": 0.00019897013608029428,
+      "loss": 1.0409,
+      "step": 1953
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.310447192134622,
+      "learning_rate": 0.00019896787032083285,
+      "loss": 1.1755,
+      "step": 1954
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28394813161935395,
+      "learning_rate": 0.00019896560208463825,
+      "loss": 1.071,
+      "step": 1955
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23857662880995645,
+      "learning_rate": 0.00019896333137176726,
+      "loss": 0.9972,
+      "step": 1956
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27471986726786446,
+      "learning_rate": 0.00019896105818227673,
+      "loss": 1.1453,
+      "step": 1957
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26890257495666114,
+      "learning_rate": 0.00019895878251622348,
+      "loss": 1.0331,
+      "step": 1958
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2663478542354145,
+      "learning_rate": 0.00019895650437366452,
+      "loss": 1.0474,
+      "step": 1959
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2604279634361445,
+      "learning_rate": 0.00019895422375465686,
+      "loss": 1.1096,
+      "step": 1960
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2606384507083339,
+      "learning_rate": 0.00019895194065925754,
+      "loss": 1.0248,
+      "step": 1961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27935688700196437,
+      "learning_rate": 0.00019894965508752375,
+      "loss": 1.2211,
+      "step": 1962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2745748469246835,
+      "learning_rate": 0.00019894736703951263,
+      "loss": 1.0072,
+      "step": 1963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23091984266360946,
+      "learning_rate": 0.00019894507651528148,
+      "loss": 1.043,
+      "step": 1964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22878065896236086,
+      "learning_rate": 0.00019894278351488757,
+      "loss": 1.0798,
+      "step": 1965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2554600035367144,
+      "learning_rate": 0.00019894048803838834,
+      "loss": 1.1627,
+      "step": 1966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2737920223861705,
+      "learning_rate": 0.00019893819008584123,
+      "loss": 1.1778,
+      "step": 1967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2708368675723523,
+      "learning_rate": 0.0001989358896573037,
+      "loss": 0.9277,
+      "step": 1968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24550243352794962,
+      "learning_rate": 0.00019893358675283337,
+      "loss": 1.1226,
+      "step": 1969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2642424699796039,
+      "learning_rate": 0.00019893128137248787,
+      "loss": 1.1078,
+      "step": 1970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2573419222534839,
+      "learning_rate": 0.00019892897351632484,
+      "loss": 1.2793,
+      "step": 1971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27224735814088064,
+      "learning_rate": 0.00019892666318440213,
+      "loss": 1.0788,
+      "step": 1972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2766376630314068,
+      "learning_rate": 0.00019892435037677746,
+      "loss": 1.1132,
+      "step": 1973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3067049268603965,
+      "learning_rate": 0.00019892203509350875,
+      "loss": 1.0906,
+      "step": 1974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27698131540064885,
+      "learning_rate": 0.00019891971733465395,
+      "loss": 1.1791,
+      "step": 1975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2874873201823343,
+      "learning_rate": 0.00019891739710027105,
+      "loss": 1.1604,
+      "step": 1976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28015375821188654,
+      "learning_rate": 0.00019891507439041814,
+      "loss": 1.1313,
+      "step": 1977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2690366043454793,
+      "learning_rate": 0.0001989127492051533,
+      "loss": 1.199,
+      "step": 1978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23644122244331564,
+      "learning_rate": 0.00019891042154453477,
+      "loss": 1.0604,
+      "step": 1979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.23968738721131738,
+      "learning_rate": 0.00019890809140862077,
+      "loss": 1.0409,
+      "step": 1980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24461539048772374,
+      "learning_rate": 0.0001989057587974696,
+      "loss": 1.103,
+      "step": 1981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2826392500904235,
+      "learning_rate": 0.0001989034237111397,
+      "loss": 1.1352,
+      "step": 1982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2772470869290075,
+      "learning_rate": 0.0001989010861496894,
+      "loss": 1.1124,
+      "step": 1983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2848620471084776,
+      "learning_rate": 0.00019889874611317732,
+      "loss": 1.0845,
+      "step": 1984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2797268047479781,
+      "learning_rate": 0.00019889640360166194,
+      "loss": 1.1135,
+      "step": 1985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2569875278431773,
+      "learning_rate": 0.00019889405861520188,
+      "loss": 1.1096,
+      "step": 1986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2617687295928765,
+      "learning_rate": 0.0001988917111538559,
+      "loss": 1.0682,
+      "step": 1987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28290160215390237,
+      "learning_rate": 0.00019888936121768266,
+      "loss": 1.1322,
+      "step": 1988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2626538158195342,
+      "learning_rate": 0.00019888700880674103,
+      "loss": 1.1404,
+      "step": 1989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27468104620820544,
+      "learning_rate": 0.00019888465392108986,
+      "loss": 1.211,
+      "step": 1990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2684528257690631,
+      "learning_rate": 0.00019888229656078808,
+      "loss": 1.086,
+      "step": 1991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2701652519028749,
+      "learning_rate": 0.00019887993672589466,
+      "loss": 1.1998,
+      "step": 1992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26483806146239974,
+      "learning_rate": 0.00019887757441646868,
+      "loss": 1.0015,
+      "step": 1993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.25776041537869526,
+      "learning_rate": 0.00019887520963256927,
+      "loss": 1.1646,
+      "step": 1994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2857416415267135,
+      "learning_rate": 0.00019887284237425558,
+      "loss": 1.1295,
+      "step": 1995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2720535216646917,
+      "learning_rate": 0.00019887047264158692,
+      "loss": 1.0362,
+      "step": 1996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2556899383886345,
+      "learning_rate": 0.0001988681004346225,
+      "loss": 1.239,
+      "step": 1997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29555721414719655,
+      "learning_rate": 0.00019886572575342174,
+      "loss": 1.1347,
+      "step": 1998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28921149656501194,
+      "learning_rate": 0.00019886334859804406,
+      "loss": 1.1826,
+      "step": 1999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.274012668267724,
+      "learning_rate": 0.00019886096896854896,
+      "loss": 1.0865,
+      "step": 2000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2422334744946127,
+      "learning_rate": 0.00019885858686499594,
+      "loss": 1.0813,
+      "step": 2001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30780112818652483,
+      "learning_rate": 0.00019885620228744468,
+      "loss": 1.0997,
+      "step": 2002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24632043466385958,
+      "learning_rate": 0.00019885381523595484,
+      "loss": 1.0984,
+      "step": 2003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27662537602831366,
+      "learning_rate": 0.00019885142571058614,
+      "loss": 1.1465,
+      "step": 2004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24285561219836202,
+      "learning_rate": 0.00019884903371139838,
+      "loss": 1.0971,
+      "step": 2005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24656395139808884,
+      "learning_rate": 0.00019884663923845142,
+      "loss": 1.143,
+      "step": 2006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2629422784962355,
+      "learning_rate": 0.0001988442422918052,
+      "loss": 1.1407,
+      "step": 2007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2811274440605523,
+      "learning_rate": 0.0001988418428715197,
+      "loss": 1.159,
+      "step": 2008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28599512163427754,
+      "learning_rate": 0.00019883944097765497,
+      "loss": 1.2293,
+      "step": 2009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2637511197971286,
+      "learning_rate": 0.0001988370366102711,
+      "loss": 1.0897,
+      "step": 2010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2621878190951607,
+      "learning_rate": 0.00019883462976942826,
+      "loss": 1.0737,
+      "step": 2011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2762514383867885,
+      "learning_rate": 0.0001988322204551867,
+      "loss": 1.1701,
+      "step": 2012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24133383394031283,
+      "learning_rate": 0.00019882980866760673,
+      "loss": 1.0147,
+      "step": 2013
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26354770911760284,
+      "learning_rate": 0.00019882739440674863,
+      "loss": 1.1734,
+      "step": 2014
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28989400649087416,
+      "learning_rate": 0.00019882497767267294,
+      "loss": 0.9902,
+      "step": 2015
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.27932170163937037,
+      "learning_rate": 0.00019882255846544005,
+      "loss": 1.1016,
+      "step": 2016
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.25887068615609143,
+      "learning_rate": 0.00019882013678511052,
+      "loss": 1.0908,
+      "step": 2017
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28911454498855693,
+      "learning_rate": 0.000198817712631745,
+      "loss": 1.1229,
+      "step": 2018
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2546367208219804,
+      "learning_rate": 0.00019881528600540404,
+      "loss": 1.1906,
+      "step": 2019
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2736344196999536,
+      "learning_rate": 0.0001988128569061485,
+      "loss": 1.1621,
+      "step": 2020
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2818744152091652,
+      "learning_rate": 0.0001988104253340391,
+      "loss": 1.0773,
+      "step": 2021
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2826631297854352,
+      "learning_rate": 0.00019880799128913672,
+      "loss": 1.043,
+      "step": 2022
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.22842355541316772,
+      "learning_rate": 0.00019880555477150223,
+      "loss": 0.8834,
+      "step": 2023
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3137419246856175,
+      "learning_rate": 0.00019880311578119667,
+      "loss": 1.1337,
+      "step": 2024
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2733714490698032,
+      "learning_rate": 0.00019880067431828102,
+      "loss": 1.183,
+      "step": 2025
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26580523535512374,
+      "learning_rate": 0.00019879823038281642,
+      "loss": 1.121,
+      "step": 2026
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.26858458548658276,
+      "learning_rate": 0.000198795783974864,
+      "loss": 1.0527,
+      "step": 2027
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2763698286676362,
+      "learning_rate": 0.00019879333509448496,
+      "loss": 1.0042,
+      "step": 2028
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.24269742693449786,
+      "learning_rate": 0.00019879088374174066,
+      "loss": 1.1615,
+      "step": 2029
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.30073652024639724,
+      "learning_rate": 0.0001987884299166924,
+      "loss": 1.121,
+      "step": 2030
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2556748923041058,
+      "learning_rate": 0.00019878597361940161,
+      "loss": 0.8961,
+      "step": 2031
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2591634855819998,
+      "learning_rate": 0.00019878351484992974,
+      "loss": 1.1487,
+      "step": 2032
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.2600632642281181,
+      "learning_rate": 0.00019878105360833832,
+      "loss": 1.1916,
+      "step": 2033
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.28116821489552873,
+      "learning_rate": 0.00019877858989468894,
+      "loss": 1.0512,
+      "step": 2034
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.20261666938146597,
+      "learning_rate": 0.0001987761237090433,
+      "loss": 1.1838,
+      "step": 2035
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29561896466102555,
+      "learning_rate": 0.00019877365505146304,
+      "loss": 1.0852,
+      "step": 2036
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.29527708211041853,
+      "learning_rate": 0.00019877118392201,
+      "loss": 1.1186,
+      "step": 2037
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.250583627663694,
+      "learning_rate": 0.00019876871032074603,
+      "loss": 1.045,
+      "step": 2038
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3097749132789377,
+      "learning_rate": 0.000198766234247733,
+      "loss": 1.1162,
+      "step": 2039
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2557816245578032,
+      "learning_rate": 0.0001987637557030329,
+      "loss": 1.0323,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28034027348239304,
+      "learning_rate": 0.00019876127468670772,
+      "loss": 1.1111,
+      "step": 2041
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27069634417343275,
+      "learning_rate": 0.00019875879119881957,
+      "loss": 1.0432,
+      "step": 2042
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2579239339061907,
+      "learning_rate": 0.00019875630523943062,
+      "loss": 1.0104,
+      "step": 2043
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28379405935029695,
+      "learning_rate": 0.00019875381680860304,
+      "loss": 1.1044,
+      "step": 2044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25276024384054346,
+      "learning_rate": 0.00019875132590639917,
+      "loss": 1.0816,
+      "step": 2045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2529581499654312,
+      "learning_rate": 0.00019874883253288126,
+      "loss": 1.0982,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27524747771098035,
+      "learning_rate": 0.00019874633668811177,
+      "loss": 1.1365,
+      "step": 2047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28979122397869284,
+      "learning_rate": 0.00019874383837215314,
+      "loss": 1.1472,
+      "step": 2048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26208338201886544,
+      "learning_rate": 0.00019874133758506792,
+      "loss": 1.0844,
+      "step": 2049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23244493332423305,
+      "learning_rate": 0.00019873883432691868,
+      "loss": 1.0652,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26326298210214855,
+      "learning_rate": 0.000198736328597768,
+      "loss": 1.1114,
+      "step": 2051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25775993234870526,
+      "learning_rate": 0.0001987338203976787,
+      "loss": 0.9868,
+      "step": 2052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2558696856240754,
+      "learning_rate": 0.00019873130972671347,
+      "loss": 1.0485,
+      "step": 2053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25810577145871305,
+      "learning_rate": 0.00019872879658493515,
+      "loss": 1.0948,
+      "step": 2054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26764993308160495,
+      "learning_rate": 0.00019872628097240667,
+      "loss": 1.1752,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2844642341098131,
+      "learning_rate": 0.00019872376288919093,
+      "loss": 1.1397,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27934363473211593,
+      "learning_rate": 0.00019872124233535102,
+      "loss": 1.2909,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28305414844226917,
+      "learning_rate": 0.00019871871931094996,
+      "loss": 1.1307,
+      "step": 2058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2547448994013944,
+      "learning_rate": 0.0001987161938160509,
+      "loss": 1.0631,
+      "step": 2059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2457309364326485,
+      "learning_rate": 0.00019871366585071706,
+      "loss": 1.0862,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2892837695062467,
+      "learning_rate": 0.00019871113541501168,
+      "loss": 1.1139,
+      "step": 2061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.297033511165508,
+      "learning_rate": 0.0001987086025089981,
+      "loss": 1.158,
+      "step": 2062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28710128200720647,
+      "learning_rate": 0.00019870606713273968,
+      "loss": 1.0244,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29989733418311626,
+      "learning_rate": 0.00019870352928629993,
+      "loss": 1.0888,
+      "step": 2064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2698108293542839,
+      "learning_rate": 0.00019870098896974234,
+      "loss": 1.0765,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2833643780662451,
+      "learning_rate": 0.00019869844618313046,
+      "loss": 1.0251,
+      "step": 2066
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3349028897187998,
+      "learning_rate": 0.00019869590092652791,
+      "loss": 1.1527,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26736402567126216,
+      "learning_rate": 0.0001986933531999984,
+      "loss": 1.1222,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24629779305771238,
+      "learning_rate": 0.00019869080300360576,
+      "loss": 1.0542,
+      "step": 2069
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2659353555548773,
+      "learning_rate": 0.00019868825033741373,
+      "loss": 1.1196,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2880977765710642,
+      "learning_rate": 0.00019868569520148618,
+      "loss": 1.1662,
+      "step": 2071
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26777705281811776,
+      "learning_rate": 0.0001986831375958871,
+      "loss": 1.1153,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2446161703167875,
+      "learning_rate": 0.0001986805775206805,
+      "loss": 1.1845,
+      "step": 2073
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2873479752588241,
+      "learning_rate": 0.00019867801497593042,
+      "loss": 1.19,
+      "step": 2074
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2574173866976346,
+      "learning_rate": 0.000198675449961701,
+      "loss": 1.1004,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.31037055152728826,
+      "learning_rate": 0.00019867288247805642,
+      "loss": 1.1266,
+      "step": 2076
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25230875618755544,
+      "learning_rate": 0.00019867031252506095,
+      "loss": 1.0861,
+      "step": 2077
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28027907388788925,
+      "learning_rate": 0.0001986677401027789,
+      "loss": 1.0899,
+      "step": 2078
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2629017191349244,
+      "learning_rate": 0.00019866516521127462,
+      "loss": 1.1268,
+      "step": 2079
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2858944012914975,
+      "learning_rate": 0.0001986625878506126,
+      "loss": 1.1248,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3010750896726883,
+      "learning_rate": 0.00019866000802085728,
+      "loss": 1.109,
+      "step": 2081
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27335673435624314,
+      "learning_rate": 0.0001986574257220733,
+      "loss": 1.0929,
+      "step": 2082
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25277983760592904,
+      "learning_rate": 0.0001986548409543252,
+      "loss": 1.0946,
+      "step": 2083
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2522955007250379,
+      "learning_rate": 0.00019865225371767773,
+      "loss": 1.1279,
+      "step": 2084
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2876473859106391,
+      "learning_rate": 0.00019864966401219559,
+      "loss": 1.044,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5481907916561444,
+      "learning_rate": 0.00019864707183794362,
+      "loss": 1.3456,
+      "step": 2086
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26281784160346944,
+      "learning_rate": 0.00019864447719498667,
+      "loss": 1.2029,
+      "step": 2087
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27656267392775247,
+      "learning_rate": 0.00019864188008338968,
+      "loss": 1.1244,
+      "step": 2088
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2712507307178155,
+      "learning_rate": 0.00019863928050321765,
+      "loss": 1.2326,
+      "step": 2089
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27024578206691424,
+      "learning_rate": 0.00019863667845453563,
+      "loss": 1.1642,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.305876067098806,
+      "learning_rate": 0.00019863407393740876,
+      "loss": 1.2,
+      "step": 2091
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2665786067215833,
+      "learning_rate": 0.00019863146695190217,
+      "loss": 1.1217,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23808439538640014,
+      "learning_rate": 0.00019862885749808115,
+      "loss": 1.089,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27236917331109767,
+      "learning_rate": 0.00019862624557601103,
+      "loss": 1.1333,
+      "step": 2094
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26305710425726253,
+      "learning_rate": 0.00019862363118575705,
+      "loss": 1.1396,
+      "step": 2095
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2302881958735561,
+      "learning_rate": 0.00019862101432738475,
+      "loss": 1.0263,
+      "step": 2096
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2954108631439019,
+      "learning_rate": 0.0001986183950009596,
+      "loss": 1.1058,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2948895500433461,
+      "learning_rate": 0.00019861577320654712,
+      "loss": 1.1621,
+      "step": 2098
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24612354772654924,
+      "learning_rate": 0.00019861314894421294,
+      "loss": 1.1682,
+      "step": 2099
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2770485117180903,
+      "learning_rate": 0.00019861052221402275,
+      "loss": 1.0537,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2803003393924788,
+      "learning_rate": 0.00019860789301604222,
+      "loss": 1.1575,
+      "step": 2101
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.263398275519541,
+      "learning_rate": 0.00019860526135033723,
+      "loss": 1.1161,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2735697531308213,
+      "learning_rate": 0.0001986026272169736,
+      "loss": 1.1304,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2837690209815238,
+      "learning_rate": 0.00019859999061601726,
+      "loss": 0.9939,
+      "step": 2104
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2611549781543971,
+      "learning_rate": 0.00019859735154753418,
+      "loss": 1.0968,
+      "step": 2105
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2858960886543411,
+      "learning_rate": 0.0001985947100115904,
+      "loss": 1.1623,
+      "step": 2106
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3657978801967696,
+      "learning_rate": 0.00019859206600825207,
+      "loss": 1.2114,
+      "step": 2107
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24528859351726237,
+      "learning_rate": 0.0001985894195375853,
+      "loss": 1.1096,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.309781272595587,
+      "learning_rate": 0.00019858677059965632,
+      "loss": 1.1382,
+      "step": 2109
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3015108916795954,
+      "learning_rate": 0.0001985841191945315,
+      "loss": 1.0789,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27510018422236365,
+      "learning_rate": 0.0001985814653222771,
+      "loss": 1.1214,
+      "step": 2111
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2504556220073607,
+      "learning_rate": 0.0001985788089829596,
+      "loss": 1.1829,
+      "step": 2112
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27607247184581263,
+      "learning_rate": 0.00019857615017664543,
+      "loss": 1.2014,
+      "step": 2113
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28257879262143415,
+      "learning_rate": 0.00019857348890340117,
+      "loss": 1.1302,
+      "step": 2114
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2961265516298664,
+      "learning_rate": 0.0001985708251632934,
+      "loss": 1.0324,
+      "step": 2115
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2942307299808682,
+      "learning_rate": 0.00019856815895638876,
+      "loss": 1.0799,
+      "step": 2116
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2541933617332,
+      "learning_rate": 0.000198565490282754,
+      "loss": 1.1498,
+      "step": 2117
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28011641730308906,
+      "learning_rate": 0.0001985628191424559,
+      "loss": 1.1392,
+      "step": 2118
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2807759455450216,
+      "learning_rate": 0.0001985601455355613,
+      "loss": 1.1776,
+      "step": 2119
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.32430654597893255,
+      "learning_rate": 0.00019855746946213714,
+      "loss": 1.1778,
+      "step": 2120
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2525816278621571,
+      "learning_rate": 0.00019855479092225037,
+      "loss": 1.1537,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.26640266147857056,
+      "learning_rate": 0.00019855210991596796,
+      "loss": 1.096,
+      "step": 2122
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.25884771414681745,
+      "learning_rate": 0.00019854942644335712,
+      "loss": 1.1562,
+      "step": 2123
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.27617724462201587,
+      "learning_rate": 0.00019854674050448493,
+      "loss": 1.1385,
+      "step": 2124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2919548651872331,
+      "learning_rate": 0.00019854405209941863,
+      "loss": 1.0791,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.24993403620332835,
+      "learning_rate": 0.00019854136122822547,
+      "loss": 1.0431,
+      "step": 2126
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.23940290308480794,
+      "learning_rate": 0.0001985386678909728,
+      "loss": 1.0944,
+      "step": 2127
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2852028804707256,
+      "learning_rate": 0.00019853597208772808,
+      "loss": 1.0735,
+      "step": 2128
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2799825280793891,
+      "learning_rate": 0.0001985332738185587,
+      "loss": 1.1108,
+      "step": 2129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2803961566058768,
+      "learning_rate": 0.00019853057308353225,
+      "loss": 1.1428,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2681024606858511,
+      "learning_rate": 0.00019852786988271628,
+      "loss": 1.1777,
+      "step": 2131
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28971370065149094,
+      "learning_rate": 0.0001985251642161784,
+      "loss": 1.1166,
+      "step": 2132
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2809462172886824,
+      "learning_rate": 0.0001985224560839864,
+      "loss": 1.1337,
+      "step": 2133
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2662105547019178,
+      "learning_rate": 0.00019851974548620803,
+      "loss": 1.2131,
+      "step": 2134
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2689850661970803,
+      "learning_rate": 0.0001985170324229111,
+      "loss": 1.1857,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2831472805779883,
+      "learning_rate": 0.00019851431689416353,
+      "loss": 1.1575,
+      "step": 2136
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2877033555483126,
+      "learning_rate": 0.00019851159890003323,
+      "loss": 1.0868,
+      "step": 2137
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.29781126767542937,
+      "learning_rate": 0.00019850887844058827,
+      "loss": 1.1535,
+      "step": 2138
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2528619996193506,
+      "learning_rate": 0.00019850615551589672,
+      "loss": 1.0632,
+      "step": 2139
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2605060917972941,
+      "learning_rate": 0.00019850343012602672,
+      "loss": 1.1709,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.2773145775379898,
+      "learning_rate": 0.0001985007022710465,
+      "loss": 1.1957,
+      "step": 2141
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.28927051493387645,
+      "learning_rate": 0.00019849797195102426,
+      "loss": 1.0608,
+      "step": 2142
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.31331821900541645,
+      "learning_rate": 0.0001984952391660284,
+      "loss": 1.1371,
+      "step": 2143
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2785760487723139,
+      "learning_rate": 0.00019849250391612726,
+      "loss": 1.1553,
+      "step": 2144
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28557512665641493,
+      "learning_rate": 0.0001984897662013893,
+      "loss": 1.1505,
+      "step": 2145
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2640270431957491,
+      "learning_rate": 0.00019848702602188304,
+      "loss": 1.0196,
+      "step": 2146
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26301343020312196,
+      "learning_rate": 0.00019848428337767708,
+      "loss": 1.0716,
+      "step": 2147
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24955497958144957,
+      "learning_rate": 0.00019848153826884004,
+      "loss": 1.1068,
+      "step": 2148
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25592481094445924,
+      "learning_rate": 0.00019847879069544058,
+      "loss": 1.0493,
+      "step": 2149
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2690607872687816,
+      "learning_rate": 0.0001984760406575475,
+      "loss": 1.1645,
+      "step": 2150
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29603325449239903,
+      "learning_rate": 0.00019847328815522964,
+      "loss": 1.0333,
+      "step": 2151
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25742762041890327,
+      "learning_rate": 0.00019847053318855582,
+      "loss": 1.2017,
+      "step": 2152
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30645008656891,
+      "learning_rate": 0.00019846777575759504,
+      "loss": 1.1346,
+      "step": 2153
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27044205667054494,
+      "learning_rate": 0.00019846501586241627,
+      "loss": 1.097,
+      "step": 2154
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23220679441493658,
+      "learning_rate": 0.00019846225350308864,
+      "loss": 1.0664,
+      "step": 2155
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26546624754158665,
+      "learning_rate": 0.00019845948867968117,
+      "loss": 1.0479,
+      "step": 2156
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2798970718045841,
+      "learning_rate": 0.00019845672139226316,
+      "loss": 1.0244,
+      "step": 2157
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2784787690728781,
+      "learning_rate": 0.00019845395164090382,
+      "loss": 1.1114,
+      "step": 2158
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24443020869424956,
+      "learning_rate": 0.00019845117942567244,
+      "loss": 1.1341,
+      "step": 2159
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.244756739484968,
+      "learning_rate": 0.00019844840474663843,
+      "loss": 1.0807,
+      "step": 2160
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2702201314078314,
+      "learning_rate": 0.00019844562760387122,
+      "loss": 1.1269,
+      "step": 2161
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29077240998538223,
+      "learning_rate": 0.00019844284799744032,
+      "loss": 1.1688,
+      "step": 2162
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2683927419703879,
+      "learning_rate": 0.00019844006592741525,
+      "loss": 1.0173,
+      "step": 2163
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3109361300962534,
+      "learning_rate": 0.0001984372813938657,
+      "loss": 1.148,
+      "step": 2164
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2501468531327423,
+      "learning_rate": 0.00019843449439686128,
+      "loss": 1.1907,
+      "step": 2165
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2664858040953975,
+      "learning_rate": 0.0001984317049364718,
+      "loss": 1.1097,
+      "step": 2166
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2549104447589198,
+      "learning_rate": 0.00019842891301276704,
+      "loss": 1.0737,
+      "step": 2167
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2908504505180112,
+      "learning_rate": 0.00019842611862581685,
+      "loss": 1.0539,
+      "step": 2168
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2959941920542236,
+      "learning_rate": 0.00019842332177569122,
+      "loss": 1.1418,
+      "step": 2169
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26475147358616613,
+      "learning_rate": 0.00019842052246246008,
+      "loss": 1.0866,
+      "step": 2170
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28262317137702664,
+      "learning_rate": 0.0001984177206861935,
+      "loss": 1.1431,
+      "step": 2171
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2615494504849684,
+      "learning_rate": 0.00019841491644696164,
+      "loss": 1.1576,
+      "step": 2172
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27419165343889973,
+      "learning_rate": 0.00019841210974483464,
+      "loss": 1.1325,
+      "step": 2173
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24040329342282296,
+      "learning_rate": 0.0001984093005798827,
+      "loss": 1.1437,
+      "step": 2174
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29409408115598895,
+      "learning_rate": 0.00019840648895217623,
+      "loss": 1.1064,
+      "step": 2175
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2523665722905447,
+      "learning_rate": 0.00019840367486178548,
+      "loss": 1.07,
+      "step": 2176
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25402746178920604,
+      "learning_rate": 0.00019840085830878095,
+      "loss": 1.0573,
+      "step": 2177
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24450222623833068,
+      "learning_rate": 0.00019839803929323305,
+      "loss": 1.1127,
+      "step": 2178
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24554337014527297,
+      "learning_rate": 0.00019839521781521245,
+      "loss": 1.0781,
+      "step": 2179
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29644405786503714,
+      "learning_rate": 0.00019839239387478962,
+      "loss": 1.1072,
+      "step": 2180
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23532298664328116,
+      "learning_rate": 0.00019838956747203533,
+      "loss": 0.9529,
+      "step": 2181
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.255567234762623,
+      "learning_rate": 0.00019838673860702027,
+      "loss": 1.2165,
+      "step": 2182
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2565392465769596,
+      "learning_rate": 0.00019838390727981527,
+      "loss": 1.049,
+      "step": 2183
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28869850717775036,
+      "learning_rate": 0.00019838107349049111,
+      "loss": 1.2043,
+      "step": 2184
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2717398071388341,
+      "learning_rate": 0.0001983782372391188,
+      "loss": 1.1689,
+      "step": 2185
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2714997167115452,
+      "learning_rate": 0.00019837539852576923,
+      "loss": 1.0412,
+      "step": 2186
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2528223715764014,
+      "learning_rate": 0.0001983725573505135,
+      "loss": 1.0636,
+      "step": 2187
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24457509080188328,
+      "learning_rate": 0.0001983697137134227,
+      "loss": 1.0427,
+      "step": 2188
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2647502073171626,
+      "learning_rate": 0.00019836686761456803,
+      "loss": 1.1109,
+      "step": 2189
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25621160412291943,
+      "learning_rate": 0.00019836401905402062,
+      "loss": 1.1426,
+      "step": 2190
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2875669800942636,
+      "learning_rate": 0.00019836116803185184,
+      "loss": 1.0843,
+      "step": 2191
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26793735322362255,
+      "learning_rate": 0.000198358314548133,
+      "loss": 1.2198,
+      "step": 2192
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30265207091393975,
+      "learning_rate": 0.00019835545860293551,
+      "loss": 0.9996,
+      "step": 2193
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25384247043679864,
+      "learning_rate": 0.0001983526001963309,
+      "loss": 1.1222,
+      "step": 2194
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25800061296353,
+      "learning_rate": 0.00019834973932839062,
+      "loss": 1.0905,
+      "step": 2195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27769824178444574,
+      "learning_rate": 0.00019834687599918632,
+      "loss": 1.0538,
+      "step": 2196
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2772132245071213,
+      "learning_rate": 0.00019834401020878963,
+      "loss": 1.0624,
+      "step": 2197
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26363087570930127,
+      "learning_rate": 0.0001983411419572723,
+      "loss": 0.9887,
+      "step": 2198
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2815107276805014,
+      "learning_rate": 0.00019833827124470608,
+      "loss": 1.1811,
+      "step": 2199
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2785544171403854,
+      "learning_rate": 0.0001983353980711628,
+      "loss": 1.1437,
+      "step": 2200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3011117821316356,
+      "learning_rate": 0.0001983325224367144,
+      "loss": 1.0398,
+      "step": 2201
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2445670553372607,
+      "learning_rate": 0.00019832964434143282,
+      "loss": 1.101,
+      "step": 2202
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25914004062255874,
+      "learning_rate": 0.00019832676378539005,
+      "loss": 1.1808,
+      "step": 2203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2754672779595424,
+      "learning_rate": 0.00019832388076865826,
+      "loss": 1.0929,
+      "step": 2204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2565507943348922,
+      "learning_rate": 0.00019832099529130959,
+      "loss": 1.0699,
+      "step": 2205
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3343174008427606,
+      "learning_rate": 0.00019831810735341618,
+      "loss": 1.0145,
+      "step": 2206
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2804796337948149,
+      "learning_rate": 0.00019831521695505035,
+      "loss": 1.0897,
+      "step": 2207
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2525345630451486,
+      "learning_rate": 0.00019831232409628445,
+      "loss": 1.0794,
+      "step": 2208
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24602578478108195,
+      "learning_rate": 0.0001983094287771908,
+      "loss": 1.1439,
+      "step": 2209
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.26742569319862003,
+      "learning_rate": 0.00019830653099784195,
+      "loss": 1.1399,
+      "step": 2210
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.30626650715688947,
+      "learning_rate": 0.00019830363075831037,
+      "loss": 1.2276,
+      "step": 2211
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2978427240509176,
+      "learning_rate": 0.00019830072805866866,
+      "loss": 1.215,
+      "step": 2212
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2795761909559458,
+      "learning_rate": 0.00019829782289898943,
+      "loss": 1.2044,
+      "step": 2213
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28310847050083876,
+      "learning_rate": 0.0001982949152793454,
+      "loss": 1.1433,
+      "step": 2214
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2796955122824297,
+      "learning_rate": 0.00019829200519980937,
+      "loss": 1.0606,
+      "step": 2215
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2528523703698838,
+      "learning_rate": 0.0001982890926604541,
+      "loss": 1.068,
+      "step": 2216
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2523940552498862,
+      "learning_rate": 0.00019828617766135255,
+      "loss": 1.0647,
+      "step": 2217
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2970231511295705,
+      "learning_rate": 0.0001982832602025776,
+      "loss": 1.2357,
+      "step": 2218
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27974233774133495,
+      "learning_rate": 0.00019828034028420232,
+      "loss": 1.0735,
+      "step": 2219
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.269451164229955,
+      "learning_rate": 0.00019827741790629975,
+      "loss": 1.0784,
+      "step": 2220
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2658199878337128,
+      "learning_rate": 0.00019827449306894304,
+      "loss": 1.0841,
+      "step": 2221
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.257731802421506,
+      "learning_rate": 0.00019827156577220537,
+      "loss": 1.2333,
+      "step": 2222
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.32039613850942644,
+      "learning_rate": 0.00019826863601616,
+      "loss": 1.1436,
+      "step": 2223
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.23336247900235474,
+      "learning_rate": 0.00019826570380088025,
+      "loss": 1.1719,
+      "step": 2224
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.28395673225685364,
+      "learning_rate": 0.0001982627691264395,
+      "loss": 1.183,
+      "step": 2225
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2848669382303132,
+      "learning_rate": 0.00019825983199291122,
+      "loss": 1.1098,
+      "step": 2226
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27510914460004676,
+      "learning_rate": 0.0001982568924003689,
+      "loss": 1.1285,
+      "step": 2227
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2921265346308141,
+      "learning_rate": 0.00019825395034888605,
+      "loss": 1.1692,
+      "step": 2228
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2761631545730766,
+      "learning_rate": 0.00019825100583853637,
+      "loss": 1.1872,
+      "step": 2229
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2592595323871488,
+      "learning_rate": 0.00019824805886939353,
+      "loss": 1.0289,
+      "step": 2230
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.29309277372861997,
+      "learning_rate": 0.00019824510944153125,
+      "loss": 1.1123,
+      "step": 2231
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.27306258003376527,
+      "learning_rate": 0.00019824215755502337,
+      "loss": 1.1453,
+      "step": 2232
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.263424668387752,
+      "learning_rate": 0.00019823920320994373,
+      "loss": 1.1002,
+      "step": 2233
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2938351047841059,
+      "learning_rate": 0.00019823624640636633,
+      "loss": 1.127,
+      "step": 2234
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24927930137531826,
+      "learning_rate": 0.0001982332871443651,
+      "loss": 1.0708,
+      "step": 2235
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2786877424279345,
+      "learning_rate": 0.00019823032542401413,
+      "loss": 1.0868,
+      "step": 2236
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2798063724399594,
+      "learning_rate": 0.00019822736124538754,
+      "loss": 1.1573,
+      "step": 2237
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2633597183766863,
+      "learning_rate": 0.00019822439460855947,
+      "loss": 1.1058,
+      "step": 2238
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.24785877289941977,
+      "learning_rate": 0.00019822142551360422,
+      "loss": 1.0471,
+      "step": 2239
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2713868952406667,
+      "learning_rate": 0.00019821845396059606,
+      "loss": 1.0428,
+      "step": 2240
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2811381014767392,
+      "learning_rate": 0.0001982154799496094,
+      "loss": 1.0762,
+      "step": 2241
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2725029043198944,
+      "learning_rate": 0.00019821250348071856,
+      "loss": 1.1293,
+      "step": 2242
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2827100509016029,
+      "learning_rate": 0.00019820952455399814,
+      "loss": 1.1447,
+      "step": 2243
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2758022374658988,
+      "learning_rate": 0.00019820654316952263,
+      "loss": 1.1659,
+      "step": 2244
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2633396491372797,
+      "learning_rate": 0.00019820355932736666,
+      "loss": 1.0462,
+      "step": 2245
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.2731226166685037,
+      "learning_rate": 0.00019820057302760488,
+      "loss": 0.9548,
+      "step": 2246
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.25556114167089006,
+      "learning_rate": 0.00019819758427031206,
+      "loss": 1.2312,
+      "step": 2247
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28848997593382414,
+      "learning_rate": 0.00019819459305556297,
+      "loss": 1.0739,
+      "step": 2248
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23351405165987268,
+      "learning_rate": 0.0001981915993834325,
+      "loss": 1.0641,
+      "step": 2249
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26791227089364905,
+      "learning_rate": 0.00019818860325399552,
+      "loss": 1.1015,
+      "step": 2250
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2820213434051579,
+      "learning_rate": 0.00019818560466732706,
+      "loss": 1.063,
+      "step": 2251
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24527078074648306,
+      "learning_rate": 0.00019818260362350213,
+      "loss": 1.1702,
+      "step": 2252
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2720420411260554,
+      "learning_rate": 0.0001981796001225958,
+      "loss": 1.0912,
+      "step": 2253
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2713012314046693,
+      "learning_rate": 0.00019817659416468332,
+      "loss": 1.0524,
+      "step": 2254
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26924822640795093,
+      "learning_rate": 0.00019817358574983983,
+      "loss": 1.0871,
+      "step": 2255
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27363733386951783,
+      "learning_rate": 0.0001981705748781407,
+      "loss": 1.0598,
+      "step": 2256
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24189809697792714,
+      "learning_rate": 0.0001981675615496612,
+      "loss": 1.084,
+      "step": 2257
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2706773118754261,
+      "learning_rate": 0.0001981645457644768,
+      "loss": 1.0637,
+      "step": 2258
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27820113281091335,
+      "learning_rate": 0.00019816152752266292,
+      "loss": 1.1624,
+      "step": 2259
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23093374468477146,
+      "learning_rate": 0.00019815850682429516,
+      "loss": 1.1735,
+      "step": 2260
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26705391917092386,
+      "learning_rate": 0.00019815548366944904,
+      "loss": 1.049,
+      "step": 2261
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28355313423369083,
+      "learning_rate": 0.00019815245805820028,
+      "loss": 1.0949,
+      "step": 2262
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2395712965169708,
+      "learning_rate": 0.00019814942999062457,
+      "loss": 1.05,
+      "step": 2263
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.285804478616941,
+      "learning_rate": 0.00019814639946679768,
+      "loss": 1.1369,
+      "step": 2264
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25061529704124386,
+      "learning_rate": 0.00019814336648679546,
+      "loss": 1.0655,
+      "step": 2265
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2909795934470434,
+      "learning_rate": 0.0001981403310506938,
+      "loss": 1.1807,
+      "step": 2266
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23297851873356334,
+      "learning_rate": 0.00019813729315856869,
+      "loss": 1.152,
+      "step": 2267
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25889655866668293,
+      "learning_rate": 0.00019813425281049613,
+      "loss": 1.1054,
+      "step": 2268
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23977654506120644,
+      "learning_rate": 0.00019813121000655223,
+      "loss": 1.1002,
+      "step": 2269
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2549715877517098,
+      "learning_rate": 0.00019812816474681314,
+      "loss": 1.057,
+      "step": 2270
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26247547673776234,
+      "learning_rate": 0.00019812511703135504,
+      "loss": 1.0619,
+      "step": 2271
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2867669044848128,
+      "learning_rate": 0.00019812206686025424,
+      "loss": 1.1794,
+      "step": 2272
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27304725331072943,
+      "learning_rate": 0.000198119014233587,
+      "loss": 1.1895,
+      "step": 2273
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25831673662414345,
+      "learning_rate": 0.00019811595915142979,
+      "loss": 1.088,
+      "step": 2274
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23021022674148214,
+      "learning_rate": 0.00019811290161385906,
+      "loss": 1.0841,
+      "step": 2275
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2575794159303839,
+      "learning_rate": 0.00019810984162095129,
+      "loss": 1.0906,
+      "step": 2276
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2609400179291492,
+      "learning_rate": 0.00019810677917278305,
+      "loss": 1.1717,
+      "step": 2277
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28398870543915045,
+      "learning_rate": 0.00019810371426943105,
+      "loss": 1.1347,
+      "step": 2278
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26309231855699067,
+      "learning_rate": 0.0001981006469109719,
+      "loss": 1.1804,
+      "step": 2279
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2902283356657511,
+      "learning_rate": 0.00019809757709748243,
+      "loss": 1.1167,
+      "step": 2280
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25748170041372764,
+      "learning_rate": 0.00019809450482903942,
+      "loss": 1.1476,
+      "step": 2281
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2667597093950612,
+      "learning_rate": 0.0001980914301057198,
+      "loss": 1.1277,
+      "step": 2282
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27836946786170796,
+      "learning_rate": 0.0001980883529276005,
+      "loss": 1.1525,
+      "step": 2283
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3083167991873422,
+      "learning_rate": 0.0001980852732947585,
+      "loss": 1.1216,
+      "step": 2284
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2676745480686396,
+      "learning_rate": 0.00019808219120727086,
+      "loss": 1.1328,
+      "step": 2285
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25527058852259726,
+      "learning_rate": 0.0001980791066652148,
+      "loss": 1.065,
+      "step": 2286
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28337351811282757,
+      "learning_rate": 0.00019807601966866746,
+      "loss": 1.1723,
+      "step": 2287
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27319098266987507,
+      "learning_rate": 0.00019807293021770604,
+      "loss": 1.0549,
+      "step": 2288
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.30841736311542484,
+      "learning_rate": 0.00019806983831240795,
+      "loss": 1.1445,
+      "step": 2289
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2555510965247522,
+      "learning_rate": 0.0001980667439528505,
+      "loss": 1.1424,
+      "step": 2290
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2641571799003314,
+      "learning_rate": 0.00019806364713911116,
+      "loss": 1.033,
+      "step": 2291
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2838900082793651,
+      "learning_rate": 0.0001980605478712674,
+      "loss": 1.0774,
+      "step": 2292
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.31407713147896055,
+      "learning_rate": 0.00019805744614939682,
+      "loss": 1.2683,
+      "step": 2293
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27082803879903133,
+      "learning_rate": 0.00019805434197357703,
+      "loss": 1.1711,
+      "step": 2294
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27007517574821516,
+      "learning_rate": 0.0001980512353438857,
+      "loss": 1.1142,
+      "step": 2295
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25200965101215933,
+      "learning_rate": 0.00019804812626040056,
+      "loss": 1.1365,
+      "step": 2296
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2482335861309017,
+      "learning_rate": 0.00019804501472319946,
+      "loss": 1.0387,
+      "step": 2297
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27093750047305093,
+      "learning_rate": 0.0001980419007323602,
+      "loss": 1.0562,
+      "step": 2298
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2823581467965368,
+      "learning_rate": 0.00019803878428796082,
+      "loss": 1.2542,
+      "step": 2299
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27114630287941716,
+      "learning_rate": 0.00019803566539007924,
+      "loss": 1.1863,
+      "step": 2300
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27533847009087203,
+      "learning_rate": 0.0001980325440387935,
+      "loss": 0.9211,
+      "step": 2301
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24736978230602902,
+      "learning_rate": 0.00019802942023418175,
+      "loss": 1.1474,
+      "step": 2302
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24528908482065118,
+      "learning_rate": 0.00019802629397632212,
+      "loss": 1.0203,
+      "step": 2303
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3102766690223985,
+      "learning_rate": 0.00019802316526529293,
+      "loss": 1.1166,
+      "step": 2304
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2645211126197188,
+      "learning_rate": 0.00019802003410117238,
+      "loss": 1.09,
+      "step": 2305
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24314341601375852,
+      "learning_rate": 0.0001980169004840389,
+      "loss": 1.067,
+      "step": 2306
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28901370914218866,
+      "learning_rate": 0.00019801376441397087,
+      "loss": 1.125,
+      "step": 2307
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.22977734256634133,
+      "learning_rate": 0.00019801062589104676,
+      "loss": 1.1017,
+      "step": 2308
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2763909341056602,
+      "learning_rate": 0.00019800748491534517,
+      "loss": 1.1466,
+      "step": 2309
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2638965585187832,
+      "learning_rate": 0.00019800434148694468,
+      "loss": 1.0884,
+      "step": 2310
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25365372028597144,
+      "learning_rate": 0.00019800119560592393,
+      "loss": 1.1063,
+      "step": 2311
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2946212685821378,
+      "learning_rate": 0.0001979980472723617,
+      "loss": 1.0209,
+      "step": 2312
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28135433865494547,
+      "learning_rate": 0.00019799489648633675,
+      "loss": 1.1626,
+      "step": 2313
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3064233206742547,
+      "learning_rate": 0.00019799174324792787,
+      "loss": 1.1433,
+      "step": 2314
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2800872851032662,
+      "learning_rate": 0.00019798858755721405,
+      "loss": 1.039,
+      "step": 2315
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25898676357853834,
+      "learning_rate": 0.00019798542941427426,
+      "loss": 1.1401,
+      "step": 2316
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26001355286806555,
+      "learning_rate": 0.00019798226881918753,
+      "loss": 1.0741,
+      "step": 2317
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.31195060460939816,
+      "learning_rate": 0.00019797910577203293,
+      "loss": 1.155,
+      "step": 2318
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2853360314799912,
+      "learning_rate": 0.00019797594027288963,
+      "loss": 1.1006,
+      "step": 2319
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2282844809122414,
+      "learning_rate": 0.00019797277232183684,
+      "loss": 1.0532,
+      "step": 2320
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28551517768089857,
+      "learning_rate": 0.00019796960191895385,
+      "loss": 1.0486,
+      "step": 2321
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2590261403859847,
+      "learning_rate": 0.00019796642906432004,
+      "loss": 1.0397,
+      "step": 2322
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23480527152036285,
+      "learning_rate": 0.0001979632537580147,
+      "loss": 1.0072,
+      "step": 2323
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26973107870500546,
+      "learning_rate": 0.00019796007600011742,
+      "loss": 1.1077,
+      "step": 2324
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26492201462618237,
+      "learning_rate": 0.0001979568957907077,
+      "loss": 1.1157,
+      "step": 2325
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.279425391462633,
+      "learning_rate": 0.00019795371312986504,
+      "loss": 1.0199,
+      "step": 2326
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2836141793634087,
+      "learning_rate": 0.00019795052801766915,
+      "loss": 1.1172,
+      "step": 2327
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2544431831518797,
+      "learning_rate": 0.0001979473404541998,
+      "loss": 1.0867,
+      "step": 2328
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2402666227462647,
+      "learning_rate": 0.0001979441504395366,
+      "loss": 1.0139,
+      "step": 2329
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2503197635519736,
+      "learning_rate": 0.00019794095797375953,
+      "loss": 1.0556,
+      "step": 2330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2581127509866034,
+      "learning_rate": 0.00019793776305694846,
+      "loss": 0.9451,
+      "step": 2331
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24403281634446966,
+      "learning_rate": 0.0001979345656891833,
+      "loss": 1.1031,
+      "step": 2332
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2532440703626388,
+      "learning_rate": 0.00019793136587054405,
+      "loss": 1.0738,
+      "step": 2333
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2597097391411967,
+      "learning_rate": 0.00019792816360111087,
+      "loss": 1.1359,
+      "step": 2334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.27991032105875546,
+      "learning_rate": 0.00019792495888096382,
+      "loss": 0.9373,
+      "step": 2335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2713405323422737,
+      "learning_rate": 0.00019792175171018313,
+      "loss": 1.0818,
+      "step": 2336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2942256218211258,
+      "learning_rate": 0.00019791854208884907,
+      "loss": 1.1304,
+      "step": 2337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.24670357822710387,
+      "learning_rate": 0.00019791533001704194,
+      "loss": 1.0164,
+      "step": 2338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25797879496591175,
+      "learning_rate": 0.00019791211549484216,
+      "loss": 1.0922,
+      "step": 2339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.28874847418524446,
+      "learning_rate": 0.00019790889852233016,
+      "loss": 1.2126,
+      "step": 2340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2831045631432414,
+      "learning_rate": 0.00019790567909958644,
+      "loss": 1.1913,
+      "step": 2341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25052218126412557,
+      "learning_rate": 0.00019790245722669153,
+      "loss": 1.1242,
+      "step": 2342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2980993557396919,
+      "learning_rate": 0.00019789923290372614,
+      "loss": 1.0499,
+      "step": 2343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26321626908133683,
+      "learning_rate": 0.00019789600613077092,
+      "loss": 1.0864,
+      "step": 2344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.26596294842052304,
+      "learning_rate": 0.0001978927769079066,
+      "loss": 1.0538,
+      "step": 2345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25422065022309154,
+      "learning_rate": 0.00019788954523521402,
+      "loss": 1.2115,
+      "step": 2346
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.23598171182692104,
+      "learning_rate": 0.00019788631111277406,
+      "loss": 1.0686,
+      "step": 2347
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2847052874921601,
+      "learning_rate": 0.00019788307454066763,
+      "loss": 1.0641,
+      "step": 2348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25357051014186355,
+      "learning_rate": 0.00019787983551897576,
+      "loss": 1.0484,
+      "step": 2349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.2411514518029808,
+      "learning_rate": 0.00019787659404777946,
+      "loss": 1.133,
+      "step": 2350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.25446854039811434,
+      "learning_rate": 0.0001978733501271599,
+      "loss": 1.1672,
+      "step": 2351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26777410057209505,
+      "learning_rate": 0.00019787010375719826,
+      "loss": 1.138,
+      "step": 2352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27808638420934434,
+      "learning_rate": 0.0001978668549379757,
+      "loss": 1.1782,
+      "step": 2353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2592972824957828,
+      "learning_rate": 0.00019786360366957367,
+      "loss": 1.1013,
+      "step": 2354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25628836959733703,
+      "learning_rate": 0.0001978603499520734,
+      "loss": 0.9967,
+      "step": 2355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2636173075017777,
+      "learning_rate": 0.0001978570937855564,
+      "loss": 1.1131,
+      "step": 2356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2965203958377406,
+      "learning_rate": 0.0001978538351701041,
+      "loss": 1.1412,
+      "step": 2357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2551730645828324,
+      "learning_rate": 0.0001978505741057981,
+      "loss": 1.1347,
+      "step": 2358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3173189097283464,
+      "learning_rate": 0.00019784731059271996,
+      "loss": 1.1354,
+      "step": 2359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.287566565828461,
+      "learning_rate": 0.00019784404463095144,
+      "loss": 1.0052,
+      "step": 2360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24804179463343937,
+      "learning_rate": 0.00019784077622057416,
+      "loss": 1.1699,
+      "step": 2361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3073578708099378,
+      "learning_rate": 0.00019783750536166993,
+      "loss": 1.1715,
+      "step": 2362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22734700217911738,
+      "learning_rate": 0.0001978342320543207,
+      "loss": 0.9563,
+      "step": 2363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2590258564685437,
+      "learning_rate": 0.0001978309562986083,
+      "loss": 1.0816,
+      "step": 2364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27557855270603093,
+      "learning_rate": 0.00019782767809461475,
+      "loss": 1.077,
+      "step": 2365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25603874246559705,
+      "learning_rate": 0.00019782439744242205,
+      "loss": 1.146,
+      "step": 2366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2761244679810728,
+      "learning_rate": 0.00019782111434211235,
+      "loss": 1.107,
+      "step": 2367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26568572466382623,
+      "learning_rate": 0.00019781782879376775,
+      "loss": 1.1306,
+      "step": 2368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2723644753582144,
+      "learning_rate": 0.00019781454079747054,
+      "loss": 1.1705,
+      "step": 2369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26902633260641967,
+      "learning_rate": 0.00019781125035330297,
+      "loss": 1.0754,
+      "step": 2370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2711384423626026,
+      "learning_rate": 0.0001978079574613474,
+      "loss": 1.1202,
+      "step": 2371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28204033368789444,
+      "learning_rate": 0.0001978046621216862,
+      "loss": 1.1706,
+      "step": 2372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26064152152322406,
+      "learning_rate": 0.00019780136433440184,
+      "loss": 1.0596,
+      "step": 2373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2907116807628749,
+      "learning_rate": 0.00019779806409957692,
+      "loss": 1.1855,
+      "step": 2374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2739091926383725,
+      "learning_rate": 0.00019779476141729396,
+      "loss": 1.1093,
+      "step": 2375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24773766770558472,
+      "learning_rate": 0.00019779145628763564,
+      "loss": 1.0423,
+      "step": 2376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.31039070403873764,
+      "learning_rate": 0.00019778814871068465,
+      "loss": 1.125,
+      "step": 2377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25711704080946696,
+      "learning_rate": 0.0001977848386865238,
+      "loss": 1.1752,
+      "step": 2378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24583579381340756,
+      "learning_rate": 0.0001977815262152359,
+      "loss": 1.0188,
+      "step": 2379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28912054275819865,
+      "learning_rate": 0.00019777821129690387,
+      "loss": 1.0406,
+      "step": 2380
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27568971892193483,
+      "learning_rate": 0.0001977748939316106,
+      "loss": 1.1377,
+      "step": 2381
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2703312754633334,
+      "learning_rate": 0.0001977715741194392,
+      "loss": 1.0889,
+      "step": 2382
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3244168867345363,
+      "learning_rate": 0.00019776825186047268,
+      "loss": 1.2365,
+      "step": 2383
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26694652233140037,
+      "learning_rate": 0.00019776492715479428,
+      "loss": 0.9792,
+      "step": 2384
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25774010566880007,
+      "learning_rate": 0.00019776160000248706,
+      "loss": 1.0835,
+      "step": 2385
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24706317723666119,
+      "learning_rate": 0.0001977582704036344,
+      "loss": 1.0586,
+      "step": 2386
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2990804163988819,
+      "learning_rate": 0.00019775493835831959,
+      "loss": 1.0996,
+      "step": 2387
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25726590762789603,
+      "learning_rate": 0.00019775160386662597,
+      "loss": 1.117,
+      "step": 2388
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2650577582941275,
+      "learning_rate": 0.00019774826692863705,
+      "loss": 1.0128,
+      "step": 2389
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2668462355680707,
+      "learning_rate": 0.00019774492754443635,
+      "loss": 1.0804,
+      "step": 2390
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2992539843834909,
+      "learning_rate": 0.00019774158571410737,
+      "loss": 1.2372,
+      "step": 2391
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2864988762111991,
+      "learning_rate": 0.00019773824143773377,
+      "loss": 1.0984,
+      "step": 2392
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.304882593233717,
+      "learning_rate": 0.00019773489471539926,
+      "loss": 1.1334,
+      "step": 2393
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28019107709650354,
+      "learning_rate": 0.00019773154554718762,
+      "loss": 1.0708,
+      "step": 2394
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.22866457976217683,
+      "learning_rate": 0.00019772819393318262,
+      "loss": 1.0887,
+      "step": 2395
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26481598326652184,
+      "learning_rate": 0.00019772483987346812,
+      "loss": 1.1488,
+      "step": 2396
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27476032047452265,
+      "learning_rate": 0.0001977214833681281,
+      "loss": 1.0884,
+      "step": 2397
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2671685597445549,
+      "learning_rate": 0.00019771812441724652,
+      "loss": 1.021,
+      "step": 2398
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2777768402686966,
+      "learning_rate": 0.00019771476302090754,
+      "loss": 1.0786,
+      "step": 2399
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25286519332057394,
+      "learning_rate": 0.00019771139917919512,
+      "loss": 0.9957,
+      "step": 2400
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25899496107583564,
+      "learning_rate": 0.00019770803289219355,
+      "loss": 1.232,
+      "step": 2401
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2554893399049988,
+      "learning_rate": 0.00019770466415998706,
+      "loss": 1.1702,
+      "step": 2402
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26615765059825036,
+      "learning_rate": 0.00019770129298265994,
+      "loss": 1.1589,
+      "step": 2403
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2867731760890007,
+      "learning_rate": 0.00019769791936029657,
+      "loss": 1.0231,
+      "step": 2404
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25865612120055026,
+      "learning_rate": 0.00019769454329298134,
+      "loss": 1.0742,
+      "step": 2405
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27168406194912653,
+      "learning_rate": 0.00019769116478079876,
+      "loss": 1.11,
+      "step": 2406
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.29872484561116197,
+      "learning_rate": 0.00019768778382383344,
+      "loss": 1.0637,
+      "step": 2407
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27164584328580743,
+      "learning_rate": 0.0001976844004221699,
+      "loss": 1.0909,
+      "step": 2408
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2739762871753536,
+      "learning_rate": 0.00019768101457589283,
+      "loss": 1.0961,
+      "step": 2409
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25357273039005795,
+      "learning_rate": 0.00019767762628508702,
+      "loss": 1.0625,
+      "step": 2410
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26504750469407357,
+      "learning_rate": 0.00019767423554983718,
+      "loss": 1.0843,
+      "step": 2411
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.274182183625254,
+      "learning_rate": 0.00019767084237022823,
+      "loss": 1.144,
+      "step": 2412
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.282485486193323,
+      "learning_rate": 0.00019766744674634508,
+      "loss": 1.1495,
+      "step": 2413
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24729551379445136,
+      "learning_rate": 0.00019766404867827269,
+      "loss": 1.1148,
+      "step": 2414
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23229018534328089,
+      "learning_rate": 0.00019766064816609607,
+      "loss": 1.0309,
+      "step": 2415
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.252259979313865,
+      "learning_rate": 0.00019765724520990038,
+      "loss": 1.1207,
+      "step": 2416
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23700561147186552,
+      "learning_rate": 0.00019765383980977074,
+      "loss": 1.1039,
+      "step": 2417
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26556810321631696,
+      "learning_rate": 0.0001976504319657924,
+      "loss": 1.1749,
+      "step": 2418
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23143332904854963,
+      "learning_rate": 0.00019764702167805064,
+      "loss": 1.1775,
+      "step": 2419
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27696732231432797,
+      "learning_rate": 0.00019764360894663076,
+      "loss": 1.0399,
+      "step": 2420
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27237322850717344,
+      "learning_rate": 0.00019764019377161823,
+      "loss": 1.0703,
+      "step": 2421
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2535297853960644,
+      "learning_rate": 0.00019763677615309847,
+      "loss": 1.0938,
+      "step": 2422
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26018058934136795,
+      "learning_rate": 0.00019763335609115703,
+      "loss": 1.0601,
+      "step": 2423
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3439807499205691,
+      "learning_rate": 0.0001976299335858795,
+      "loss": 1.04,
+      "step": 2424
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24885428737327994,
+      "learning_rate": 0.0001976265086373515,
+      "loss": 1.0878,
+      "step": 2425
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2518594237134699,
+      "learning_rate": 0.0001976230812456588,
+      "loss": 1.1054,
+      "step": 2426
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27045653240229783,
+      "learning_rate": 0.0001976196514108871,
+      "loss": 0.9974,
+      "step": 2427
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27897290427116067,
+      "learning_rate": 0.0001976162191331223,
+      "loss": 1.1238,
+      "step": 2428
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26488970207228696,
+      "learning_rate": 0.00019761278441245023,
+      "loss": 1.0545,
+      "step": 2429
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2676698182906196,
+      "learning_rate": 0.00019760934724895692,
+      "loss": 1.1392,
+      "step": 2430
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2554504815510488,
+      "learning_rate": 0.00019760590764272834,
+      "loss": 1.1628,
+      "step": 2431
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23803751811537502,
+      "learning_rate": 0.0001976024655938506,
+      "loss": 1.1082,
+      "step": 2432
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2843747176664285,
+      "learning_rate": 0.00019759902110240977,
+      "loss": 1.0482,
+      "step": 2433
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.23996308884375614,
+      "learning_rate": 0.00019759557416849214,
+      "loss": 1.0972,
+      "step": 2434
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26431108924886854,
+      "learning_rate": 0.00019759212479218393,
+      "loss": 1.1708,
+      "step": 2435
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26205638413878823,
+      "learning_rate": 0.0001975886729735714,
+      "loss": 1.1489,
+      "step": 2436
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26602482241211184,
+      "learning_rate": 0.00019758521871274107,
+      "loss": 1.1432,
+      "step": 2437
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2611287789955635,
+      "learning_rate": 0.00019758176200977928,
+      "loss": 1.0747,
+      "step": 2438
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2512098131069008,
+      "learning_rate": 0.00019757830286477258,
+      "loss": 1.1516,
+      "step": 2439
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.300485183834668,
+      "learning_rate": 0.0001975748412778075,
+      "loss": 1.0355,
+      "step": 2440
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24888115485358228,
+      "learning_rate": 0.00019757137724897073,
+      "loss": 1.1842,
+      "step": 2441
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24322023911969542,
+      "learning_rate": 0.0001975679107783489,
+      "loss": 1.1434,
+      "step": 2442
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25278550537075395,
+      "learning_rate": 0.00019756444186602877,
+      "loss": 1.1414,
+      "step": 2443
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2629822500133049,
+      "learning_rate": 0.0001975609705120972,
+      "loss": 1.1144,
+      "step": 2444
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.25330431801210734,
+      "learning_rate": 0.00019755749671664102,
+      "loss": 1.1006,
+      "step": 2445
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27277917321975415,
+      "learning_rate": 0.00019755402047974717,
+      "loss": 1.1218,
+      "step": 2446
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24954891050559894,
+      "learning_rate": 0.00019755054180150262,
+      "loss": 1.168,
+      "step": 2447
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24441850562430112,
+      "learning_rate": 0.00019754706068199446,
+      "loss": 0.9717,
+      "step": 2448
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.26980275249400415,
+      "learning_rate": 0.00019754357712130984,
+      "loss": 1.0781,
+      "step": 2449
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.2644012899065647,
+      "learning_rate": 0.00019754009111953586,
+      "loss": 1.2219,
+      "step": 2450
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28502445348284167,
+      "learning_rate": 0.00019753660267675982,
+      "loss": 1.1411,
+      "step": 2451
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24105312734962844,
+      "learning_rate": 0.000197533111793069,
+      "loss": 1.0395,
+      "step": 2452
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.28402705301365877,
+      "learning_rate": 0.0001975296184685507,
+      "loss": 1.1438,
+      "step": 2453
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.251590448441616,
+      "learning_rate": 0.00019752612270329247,
+      "loss": 1.0017,
+      "step": 2454
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.27385681060181105,
+      "learning_rate": 0.0001975226244973817,
+      "loss": 1.2012,
+      "step": 2455
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.24837745278204337,
+      "learning_rate": 0.000197519123850906,
+      "loss": 1.1111,
+      "step": 2456
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25263755603908267,
+      "learning_rate": 0.0001975156207639529,
+      "loss": 1.1382,
+      "step": 2457
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2913062612133686,
+      "learning_rate": 0.0001975121152366101,
+      "loss": 1.144,
+      "step": 2458
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.279969123492799,
+      "learning_rate": 0.00019750860726896536,
+      "loss": 1.1385,
+      "step": 2459
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24114428531190943,
+      "learning_rate": 0.00019750509686110643,
+      "loss": 1.0758,
+      "step": 2460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24865099555297668,
+      "learning_rate": 0.00019750158401312117,
+      "loss": 1.0137,
+      "step": 2461
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30049591909461865,
+      "learning_rate": 0.0001974980687250975,
+      "loss": 1.324,
+      "step": 2462
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2535091796023302,
+      "learning_rate": 0.00019749455099712332,
+      "loss": 1.1444,
+      "step": 2463
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24893191882319649,
+      "learning_rate": 0.00019749103082928682,
+      "loss": 0.8933,
+      "step": 2464
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2930991676695541,
+      "learning_rate": 0.00019748750822167594,
+      "loss": 1.0118,
+      "step": 2465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2436760336989602,
+      "learning_rate": 0.00019748398317437894,
+      "loss": 1.0733,
+      "step": 2466
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2815078161913315,
+      "learning_rate": 0.00019748045568748396,
+      "loss": 1.1311,
+      "step": 2467
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2707645703704046,
+      "learning_rate": 0.00019747692576107935,
+      "loss": 1.1313,
+      "step": 2468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2605533882651966,
+      "learning_rate": 0.00019747339339525337,
+      "loss": 1.0691,
+      "step": 2469
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25404802860541503,
+      "learning_rate": 0.00019746985859009448,
+      "loss": 1.1801,
+      "step": 2470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2769596865608125,
+      "learning_rate": 0.00019746632134569114,
+      "loss": 1.0646,
+      "step": 2471
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28755610236991974,
+      "learning_rate": 0.0001974627816621318,
+      "loss": 1.0567,
+      "step": 2472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2561327873358053,
+      "learning_rate": 0.00019745923953950516,
+      "loss": 1.1097,
+      "step": 2473
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27123359888401705,
+      "learning_rate": 0.00019745569497789975,
+      "loss": 1.0804,
+      "step": 2474
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3011360941789264,
+      "learning_rate": 0.00019745214797740437,
+      "loss": 0.9762,
+      "step": 2475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23621703283724582,
+      "learning_rate": 0.00019744859853810772,
+      "loss": 1.2314,
+      "step": 2476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2585948578333519,
+      "learning_rate": 0.00019744504666009864,
+      "loss": 1.1219,
+      "step": 2477
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.23886114739760986,
+      "learning_rate": 0.00019744149234346604,
+      "loss": 1.0854,
+      "step": 2478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27090142697493064,
+      "learning_rate": 0.00019743793558829885,
+      "loss": 1.1247,
+      "step": 2479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26214502250978616,
+      "learning_rate": 0.00019743437639468606,
+      "loss": 1.0928,
+      "step": 2480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26399673853102246,
+      "learning_rate": 0.00019743081476271675,
+      "loss": 1.229,
+      "step": 2481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3147492977324059,
+      "learning_rate": 0.00019742725069248014,
+      "loss": 1.1473,
+      "step": 2482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26237262208219775,
+      "learning_rate": 0.0001974236841840653,
+      "loss": 1.0796,
+      "step": 2483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2556663951288371,
+      "learning_rate": 0.00019742011523756154,
+      "loss": 1.0103,
+      "step": 2484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2780031870200213,
+      "learning_rate": 0.0001974165438530582,
+      "loss": 1.1362,
+      "step": 2485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27186052495109936,
+      "learning_rate": 0.0001974129700306446,
+      "loss": 1.0304,
+      "step": 2486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28105904722620484,
+      "learning_rate": 0.0001974093937704102,
+      "loss": 1.1844,
+      "step": 2487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25863957400723964,
+      "learning_rate": 0.00019740581507244449,
+      "loss": 1.1119,
+      "step": 2488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28394214465783835,
+      "learning_rate": 0.00019740223393683706,
+      "loss": 1.1119,
+      "step": 2489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26485094517848373,
+      "learning_rate": 0.00019739865036367751,
+      "loss": 1.1412,
+      "step": 2490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28470688440068453,
+      "learning_rate": 0.0001973950643530555,
+      "loss": 1.0705,
+      "step": 2491
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26331398514634713,
+      "learning_rate": 0.00019739147590506085,
+      "loss": 1.052,
+      "step": 2492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25097699443135785,
+      "learning_rate": 0.00019738788501978325,
+      "loss": 1.0758,
+      "step": 2493
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24766896766086546,
+      "learning_rate": 0.00019738429169731262,
+      "loss": 1.0952,
+      "step": 2494
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28022175227433616,
+      "learning_rate": 0.00019738069593773893,
+      "loss": 1.0738,
+      "step": 2495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2522932834715685,
+      "learning_rate": 0.0001973770977411521,
+      "loss": 1.1787,
+      "step": 2496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2926767224083834,
+      "learning_rate": 0.0001973734971076422,
+      "loss": 1.1114,
+      "step": 2497
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2915842765013931,
+      "learning_rate": 0.00019736989403729935,
+      "loss": 1.0235,
+      "step": 2498
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2462394398490279,
+      "learning_rate": 0.0001973662885302137,
+      "loss": 1.0616,
+      "step": 2499
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2608600216455592,
+      "learning_rate": 0.00019736268058647547,
+      "loss": 1.2348,
+      "step": 2500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27127694754468956,
+      "learning_rate": 0.000197359070206175,
+      "loss": 1.2243,
+      "step": 2501
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27181980704622394,
+      "learning_rate": 0.00019735545738940258,
+      "loss": 1.0685,
+      "step": 2502
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26463599414343286,
+      "learning_rate": 0.00019735184213624866,
+      "loss": 1.083,
+      "step": 2503
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2858932841784855,
+      "learning_rate": 0.00019734822444680372,
+      "loss": 1.1853,
+      "step": 2504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29334417625161935,
+      "learning_rate": 0.00019734460432115826,
+      "loss": 1.149,
+      "step": 2505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30013981243424426,
+      "learning_rate": 0.00019734098175940292,
+      "loss": 1.1454,
+      "step": 2506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25388848360987426,
+      "learning_rate": 0.00019733735676162833,
+      "loss": 1.0328,
+      "step": 2507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2911267465827057,
+      "learning_rate": 0.0001973337293279252,
+      "loss": 1.0848,
+      "step": 2508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3184503213445064,
+      "learning_rate": 0.00019733009945838435,
+      "loss": 1.1812,
+      "step": 2509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24860610501636035,
+      "learning_rate": 0.00019732646715309656,
+      "loss": 1.0944,
+      "step": 2510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2523314374067835,
+      "learning_rate": 0.00019732283241215276,
+      "loss": 1.0563,
+      "step": 2511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26493511638695577,
+      "learning_rate": 0.00019731919523564395,
+      "loss": 1.0797,
+      "step": 2512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.34257964691618786,
+      "learning_rate": 0.00019731555562366108,
+      "loss": 1.0442,
+      "step": 2513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.278969034097831,
+      "learning_rate": 0.0001973119135762953,
+      "loss": 1.085,
+      "step": 2514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2821218023861396,
+      "learning_rate": 0.00019730826909363771,
+      "loss": 1.1072,
+      "step": 2515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2550975254765698,
+      "learning_rate": 0.00019730462217577955,
+      "loss": 1.0796,
+      "step": 2516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24932942154107635,
+      "learning_rate": 0.00019730097282281202,
+      "loss": 1.0744,
+      "step": 2517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.278741555046221,
+      "learning_rate": 0.00019729732103482652,
+      "loss": 1.2485,
+      "step": 2518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28093405939217236,
+      "learning_rate": 0.0001972936668119144,
+      "loss": 1.107,
+      "step": 2519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25703049143132695,
+      "learning_rate": 0.00019729001015416714,
+      "loss": 1.1391,
+      "step": 2520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26010024870942694,
+      "learning_rate": 0.00019728635106167622,
+      "loss": 1.0808,
+      "step": 2521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2640446473647651,
+      "learning_rate": 0.00019728268953453324,
+      "loss": 1.0537,
+      "step": 2522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26949297336392625,
+      "learning_rate": 0.0001972790255728298,
+      "loss": 1.0032,
+      "step": 2523
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27078666739320834,
+      "learning_rate": 0.00019727535917665764,
+      "loss": 1.0726,
+      "step": 2524
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2270614158675183,
+      "learning_rate": 0.00019727169034610843,
+      "loss": 1.084,
+      "step": 2525
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2772909155669702,
+      "learning_rate": 0.00019726801908127403,
+      "loss": 1.0104,
+      "step": 2526
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26805528081435065,
+      "learning_rate": 0.00019726434538224638,
+      "loss": 1.0985,
+      "step": 2527
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3021364267615302,
+      "learning_rate": 0.00019726066924911732,
+      "loss": 1.1961,
+      "step": 2528
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2742459883930204,
+      "learning_rate": 0.0001972569906819789,
+      "loss": 0.8326,
+      "step": 2529
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2965953559002404,
+      "learning_rate": 0.00019725330968092315,
+      "loss": 1.107,
+      "step": 2530
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2684835136930391,
+      "learning_rate": 0.0001972496262460422,
+      "loss": 1.0751,
+      "step": 2531
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2718707210872007,
+      "learning_rate": 0.00019724594037742824,
+      "loss": 1.0515,
+      "step": 2532
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25725745750943513,
+      "learning_rate": 0.00019724225207517354,
+      "loss": 1.1485,
+      "step": 2533
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2890461500320697,
+      "learning_rate": 0.0001972385613393703,
+      "loss": 1.1419,
+      "step": 2534
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2440614465912653,
+      "learning_rate": 0.000197234868170111,
+      "loss": 1.0766,
+      "step": 2535
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2631890740774793,
+      "learning_rate": 0.00019723117256748802,
+      "loss": 1.0583,
+      "step": 2536
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2654794914552144,
+      "learning_rate": 0.0001972274745315938,
+      "loss": 1.103,
+      "step": 2537
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25772820159308557,
+      "learning_rate": 0.00019722377406252095,
+      "loss": 1.1091,
+      "step": 2538
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2759349478590902,
+      "learning_rate": 0.00019722007116036204,
+      "loss": 1.0997,
+      "step": 2539
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29182935407079413,
+      "learning_rate": 0.00019721636582520978,
+      "loss": 0.9975,
+      "step": 2540
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2706259485506027,
+      "learning_rate": 0.00019721265805715686,
+      "loss": 1.0848,
+      "step": 2541
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27778012967187665,
+      "learning_rate": 0.00019720894785629604,
+      "loss": 1.1421,
+      "step": 2542
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24224758551615272,
+      "learning_rate": 0.00019720523522272023,
+      "loss": 1.1094,
+      "step": 2543
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27897485799673416,
+      "learning_rate": 0.0001972015201565223,
+      "loss": 0.9736,
+      "step": 2544
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25755355735125685,
+      "learning_rate": 0.00019719780265779527,
+      "loss": 1.1512,
+      "step": 2545
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26839822619910775,
+      "learning_rate": 0.00019719408272663211,
+      "loss": 1.0875,
+      "step": 2546
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3230859730143896,
+      "learning_rate": 0.00019719036036312595,
+      "loss": 1.1185,
+      "step": 2547
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3000203530525595,
+      "learning_rate": 0.00019718663556736997,
+      "loss": 0.9699,
+      "step": 2548
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2673877046628615,
+      "learning_rate": 0.00019718290833945732,
+      "loss": 1.0584,
+      "step": 2549
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.28638413313248523,
+      "learning_rate": 0.00019717917867948136,
+      "loss": 1.1215,
+      "step": 2550
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.27932539137807955,
+      "learning_rate": 0.00019717544658753533,
+      "loss": 1.0187,
+      "step": 2551
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29577216947674934,
+      "learning_rate": 0.00019717171206371268,
+      "loss": 1.0782,
+      "step": 2552
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.29132618922289466,
+      "learning_rate": 0.00019716797510810688,
+      "loss": 1.1716,
+      "step": 2553
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.30978702723114215,
+      "learning_rate": 0.00019716423572081144,
+      "loss": 1.176,
+      "step": 2554
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26767419284947913,
+      "learning_rate": 0.0001971604939019199,
+      "loss": 0.9729,
+      "step": 2555
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24950178862061026,
+      "learning_rate": 0.000197156749651526,
+      "loss": 1.1032,
+      "step": 2556
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.2676153474407372,
+      "learning_rate": 0.00019715300296972333,
+      "loss": 1.0011,
+      "step": 2557
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.24233870876913635,
+      "learning_rate": 0.00019714925385660572,
+      "loss": 1.0203,
+      "step": 2558
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.26922786191679665,
+      "learning_rate": 0.00019714550231226697,
+      "loss": 1.0598,
+      "step": 2559
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.25818023471817164,
+      "learning_rate": 0.000197141748336801,
+      "loss": 1.1347,
+      "step": 2560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2618807690441127,
+      "learning_rate": 0.00019713799193030166,
+      "loss": 1.1211,
+      "step": 2561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29741704565126115,
+      "learning_rate": 0.00019713423309286309,
+      "loss": 1.0597,
+      "step": 2562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28884125218560003,
+      "learning_rate": 0.00019713047182457928,
+      "loss": 1.0358,
+      "step": 2563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26649025648335334,
+      "learning_rate": 0.00019712670812554434,
+      "loss": 1.162,
+      "step": 2564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2947735363660406,
+      "learning_rate": 0.00019712294199585248,
+      "loss": 1.0586,
+      "step": 2565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25677075754484635,
+      "learning_rate": 0.000197119173435598,
+      "loss": 1.1841,
+      "step": 2566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2637205544668129,
+      "learning_rate": 0.00019711540244487515,
+      "loss": 1.0261,
+      "step": 2567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24823082695362084,
+      "learning_rate": 0.0001971116290237783,
+      "loss": 1.0277,
+      "step": 2568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25229952638255043,
+      "learning_rate": 0.0001971078531724019,
+      "loss": 1.0751,
+      "step": 2569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30254391196407815,
+      "learning_rate": 0.00019710407489084047,
+      "loss": 1.1231,
+      "step": 2570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3021948667516261,
+      "learning_rate": 0.00019710029417918854,
+      "loss": 1.0033,
+      "step": 2571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23588733965722716,
+      "learning_rate": 0.00019709651103754067,
+      "loss": 1.1007,
+      "step": 2572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24382613129524622,
+      "learning_rate": 0.00019709272546599164,
+      "loss": 1.0672,
+      "step": 2573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2733353690652961,
+      "learning_rate": 0.00019708893746463613,
+      "loss": 1.1596,
+      "step": 2574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26912681562116336,
+      "learning_rate": 0.00019708514703356894,
+      "loss": 1.1428,
+      "step": 2575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2956110006161434,
+      "learning_rate": 0.00019708135417288491,
+      "loss": 1.0804,
+      "step": 2576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2630689064650718,
+      "learning_rate": 0.000197077558882679,
+      "loss": 1.0586,
+      "step": 2577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2689950863610693,
+      "learning_rate": 0.00019707376116304617,
+      "loss": 1.1105,
+      "step": 2578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2910234057476756,
+      "learning_rate": 0.00019706996101408146,
+      "loss": 1.0988,
+      "step": 2579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27854677846223636,
+      "learning_rate": 0.00019706615843587995,
+      "loss": 1.1806,
+      "step": 2580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24191714015622726,
+      "learning_rate": 0.00019706235342853683,
+      "loss": 1.0358,
+      "step": 2581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2654110236015743,
+      "learning_rate": 0.00019705854599214734,
+      "loss": 1.1535,
+      "step": 2582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24454579378976135,
+      "learning_rate": 0.0001970547361268067,
+      "loss": 1.0781,
+      "step": 2583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.24819542572155914,
+      "learning_rate": 0.00019705092383261028,
+      "loss": 1.1531,
+      "step": 2584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31712338038836446,
+      "learning_rate": 0.00019704710910965352,
+      "loss": 1.2182,
+      "step": 2585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2770504057135462,
+      "learning_rate": 0.00019704329195803188,
+      "loss": 0.9587,
+      "step": 2586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2381187276476534,
+      "learning_rate": 0.00019703947237784087,
+      "loss": 0.9862,
+      "step": 2587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29302550861588544,
+      "learning_rate": 0.00019703565036917605,
+      "loss": 1.0806,
+      "step": 2588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26005985628816053,
+      "learning_rate": 0.0001970318259321331,
+      "loss": 0.9799,
+      "step": 2589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27882886267706336,
+      "learning_rate": 0.0001970279990668077,
+      "loss": 1.1686,
+      "step": 2590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26205774855080516,
+      "learning_rate": 0.0001970241697732957,
+      "loss": 1.1234,
+      "step": 2591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26668517374001244,
+      "learning_rate": 0.00019702033805169285,
+      "loss": 1.0884,
+      "step": 2592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25715862286863395,
+      "learning_rate": 0.00019701650390209504,
+      "loss": 1.2244,
+      "step": 2593
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3058673752020222,
+      "learning_rate": 0.00019701266732459827,
+      "loss": 1.0458,
+      "step": 2594
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2529497741528999,
+      "learning_rate": 0.00019700882831929852,
+      "loss": 1.3244,
+      "step": 2595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2943742754766903,
+      "learning_rate": 0.0001970049868862919,
+      "loss": 1.0521,
+      "step": 2596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25522996903552037,
+      "learning_rate": 0.0001970011430256745,
+      "loss": 1.0895,
+      "step": 2597
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2671836556503752,
+      "learning_rate": 0.00019699729673754255,
+      "loss": 1.1154,
+      "step": 2598
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3007590432393228,
+      "learning_rate": 0.00019699344802199224,
+      "loss": 1.2871,
+      "step": 2599
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2991338944551984,
+      "learning_rate": 0.00019698959687911998,
+      "loss": 1.0738,
+      "step": 2600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26091410208452626,
+      "learning_rate": 0.00019698574330902208,
+      "loss": 1.2443,
+      "step": 2601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27563048614294183,
+      "learning_rate": 0.00019698188731179502,
+      "loss": 1.0806,
+      "step": 2602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26585816626226483,
+      "learning_rate": 0.00019697802888753526,
+      "loss": 0.9916,
+      "step": 2603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27522846970536974,
+      "learning_rate": 0.0001969741680363394,
+      "loss": 1.0998,
+      "step": 2604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2602657313030757,
+      "learning_rate": 0.00019697030475830402,
+      "loss": 1.0419,
+      "step": 2605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2453226312120246,
+      "learning_rate": 0.00019696643905352582,
+      "loss": 0.9631,
+      "step": 2606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29471825123420187,
+      "learning_rate": 0.00019696257092210155,
+      "loss": 1.0666,
+      "step": 2607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2902443561039826,
+      "learning_rate": 0.000196958700364128,
+      "loss": 1.0988,
+      "step": 2608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2538912185129694,
+      "learning_rate": 0.00019695482737970202,
+      "loss": 1.1586,
+      "step": 2609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28579871153918274,
+      "learning_rate": 0.0001969509519689206,
+      "loss": 1.0839,
+      "step": 2610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23553918792552284,
+      "learning_rate": 0.00019694707413188062,
+      "loss": 1.0165,
+      "step": 2611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28381053637003834,
+      "learning_rate": 0.0001969431938686792,
+      "loss": 1.0899,
+      "step": 2612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2605318408738508,
+      "learning_rate": 0.00019693931117941346,
+      "loss": 1.1033,
+      "step": 2613
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.1338403224945068,
+      "eval_runtime": 4230.6339,
+      "eval_samples_per_second": 19.765,
+      "eval_steps_per_second": 2.471,
+      "step": 2613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.255252333620048,
+      "learning_rate": 0.00019693542606418052,
+      "loss": 1.1488,
+      "step": 2614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.30633153587103285,
+      "learning_rate": 0.00019693153852307757,
+      "loss": 1.0757,
+      "step": 2615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27505703870757664,
+      "learning_rate": 0.000196927648556202,
+      "loss": 1.1299,
+      "step": 2616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2713935955775416,
+      "learning_rate": 0.00019692375616365112,
+      "loss": 1.0189,
+      "step": 2617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2684321316986978,
+      "learning_rate": 0.00019691986134552227,
+      "loss": 1.1241,
+      "step": 2618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25621412932428106,
+      "learning_rate": 0.00019691596410191303,
+      "loss": 1.0671,
+      "step": 2619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26289530369330816,
+      "learning_rate": 0.00019691206443292085,
+      "loss": 0.9654,
+      "step": 2620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2650174904695066,
+      "learning_rate": 0.00019690816233864337,
+      "loss": 1.0996,
+      "step": 2621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2575327507563281,
+      "learning_rate": 0.0001969042578191782,
+      "loss": 1.1166,
+      "step": 2622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26454046915240054,
+      "learning_rate": 0.00019690035087462307,
+      "loss": 0.9949,
+      "step": 2623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2805971262563047,
+      "learning_rate": 0.0001968964415050758,
+      "loss": 1.0859,
+      "step": 2624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.21973703426240196,
+      "learning_rate": 0.00019689252971063416,
+      "loss": 1.1679,
+      "step": 2625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3139171336564937,
+      "learning_rate": 0.00019688861549139607,
+      "loss": 1.2573,
+      "step": 2626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3076875510868648,
+      "learning_rate": 0.0001968846988474595,
+      "loss": 1.1237,
+      "step": 2627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28812414941928133,
+      "learning_rate": 0.00019688077977892245,
+      "loss": 1.0934,
+      "step": 2628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31182412232961415,
+      "learning_rate": 0.00019687685828588297,
+      "loss": 1.0891,
+      "step": 2629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2625520463054526,
+      "learning_rate": 0.00019687293436843926,
+      "loss": 1.1112,
+      "step": 2630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25837520805297015,
+      "learning_rate": 0.00019686900802668946,
+      "loss": 0.9673,
+      "step": 2631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.27460647995152365,
+      "learning_rate": 0.00019686507926073188,
+      "loss": 1.1495,
+      "step": 2632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26861650854795094,
+      "learning_rate": 0.00019686114807066478,
+      "loss": 1.0432,
+      "step": 2633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28385341495610805,
+      "learning_rate": 0.0001968572144565866,
+      "loss": 1.1461,
+      "step": 2634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.271570918418974,
+      "learning_rate": 0.00019685327841859572,
+      "loss": 1.185,
+      "step": 2635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2673269371926348,
+      "learning_rate": 0.00019684933995679074,
+      "loss": 1.1282,
+      "step": 2636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2403632191882647,
+      "learning_rate": 0.0001968453990712701,
+      "loss": 1.0608,
+      "step": 2637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2586849785633008,
+      "learning_rate": 0.00019684145576213252,
+      "loss": 1.1015,
+      "step": 2638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2649189546870879,
+      "learning_rate": 0.00019683751002947663,
+      "loss": 1.1428,
+      "step": 2639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2524600423422356,
+      "learning_rate": 0.0001968335618734012,
+      "loss": 1.0497,
+      "step": 2640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.25016994107064466,
+      "learning_rate": 0.00019682961129400503,
+      "loss": 1.1158,
+      "step": 2641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26421571146347217,
+      "learning_rate": 0.000196825658291387,
+      "loss": 1.0628,
+      "step": 2642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.282173839472887,
+      "learning_rate": 0.00019682170286564597,
+      "loss": 0.998,
+      "step": 2643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2323812037769926,
+      "learning_rate": 0.00019681774501688102,
+      "loss": 1.0137,
+      "step": 2644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2901381801505397,
+      "learning_rate": 0.0001968137847451911,
+      "loss": 1.0866,
+      "step": 2645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26576724197777407,
+      "learning_rate": 0.0001968098220506754,
+      "loss": 1.0927,
+      "step": 2646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2696230704622509,
+      "learning_rate": 0.0001968058569334331,
+      "loss": 1.1894,
+      "step": 2647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26017036741055427,
+      "learning_rate": 0.00019680188939356336,
+      "loss": 0.9647,
+      "step": 2648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2677987402758377,
+      "learning_rate": 0.0001967979194311655,
+      "loss": 1.0862,
+      "step": 2649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28711631837028534,
+      "learning_rate": 0.00019679394704633888,
+      "loss": 1.2582,
+      "step": 2650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26765916334917605,
+      "learning_rate": 0.00019678997223918288,
+      "loss": 1.0659,
+      "step": 2651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.26777059535455733,
+      "learning_rate": 0.000196785995009797,
+      "loss": 1.1978,
+      "step": 2652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.22518302167201712,
+      "learning_rate": 0.00019678201535828076,
+      "loss": 1.0836,
+      "step": 2653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2639486825941676,
+      "learning_rate": 0.00019677803328473377,
+      "loss": 1.1509,
+      "step": 2654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.23538041430872067,
+      "learning_rate": 0.00019677404878925566,
+      "loss": 0.9895,
+      "step": 2655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.295994436523556,
+      "learning_rate": 0.00019677006187194618,
+      "loss": 1.0979,
+      "step": 2656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.28307792530010506,
+      "learning_rate": 0.00019676607253290508,
+      "loss": 1.0768,
+      "step": 2657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2761519436955159,
+      "learning_rate": 0.0001967620807722322,
+      "loss": 1.0224,
+      "step": 2658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2879025567486724,
+      "learning_rate": 0.00019675808659002744,
+      "loss": 1.1083,
+      "step": 2659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.3004898697546929,
+      "learning_rate": 0.00019675408998639076,
+      "loss": 1.1356,
+      "step": 2660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.31015042588546826,
+      "learning_rate": 0.00019675009096142214,
+      "loss": 1.1486,
+      "step": 2661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2811558967690104,
+      "learning_rate": 0.0001967460895152217,
+      "loss": 1.0928,
+      "step": 2662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2633365965205071,
+      "learning_rate": 0.00019674208564788957,
+      "loss": 1.1153,
+      "step": 2663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.2860499750304183,
+      "learning_rate": 0.00019673807935952596,
+      "loss": 1.1082,
+      "step": 2664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.29758088608843136,
+      "learning_rate": 0.0001967340706502311,
+      "loss": 1.2912,
+      "step": 2665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2568418533290371,
+      "learning_rate": 0.00019673005952010534,
+      "loss": 1.1485,
+      "step": 2666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26096215810194917,
+      "learning_rate": 0.00019672604596924904,
+      "loss": 1.0748,
+      "step": 2667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28867342954233294,
+      "learning_rate": 0.00019672202999776266,
+      "loss": 1.0794,
+      "step": 2668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.21142091669081053,
+      "learning_rate": 0.0001967180116057467,
+      "loss": 1.0324,
+      "step": 2669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2592661813230898,
+      "learning_rate": 0.00019671399079330168,
+      "loss": 1.1463,
+      "step": 2670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.23960804899506707,
+      "learning_rate": 0.00019670996756052827,
+      "loss": 1.0908,
+      "step": 2671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24684475939341896,
+      "learning_rate": 0.00019670594190752713,
+      "loss": 1.036,
+      "step": 2672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28350271509951525,
+      "learning_rate": 0.00019670191383439907,
+      "loss": 1.0251,
+      "step": 2673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5535529462100653,
+      "learning_rate": 0.00019669788334124476,
+      "loss": 1.052,
+      "step": 2674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3077728064262205,
+      "learning_rate": 0.0001966938504281652,
+      "loss": 1.116,
+      "step": 2675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2656350026116821,
+      "learning_rate": 0.00019668981509526128,
+      "loss": 1.2018,
+      "step": 2676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2479927609636305,
+      "learning_rate": 0.00019668577734263394,
+      "loss": 1.104,
+      "step": 2677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2777215172332179,
+      "learning_rate": 0.00019668173717038426,
+      "loss": 1.1844,
+      "step": 2678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2829530059947927,
+      "learning_rate": 0.00019667769457861335,
+      "loss": 1.0842,
+      "step": 2679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2550816083760124,
+      "learning_rate": 0.00019667364956742236,
+      "loss": 1.1394,
+      "step": 2680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2753775575550518,
+      "learning_rate": 0.00019666960213691255,
+      "loss": 1.1293,
+      "step": 2681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2767717867669593,
+      "learning_rate": 0.0001966655522871852,
+      "loss": 1.0783,
+      "step": 2682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26993060617287373,
+      "learning_rate": 0.00019666150001834164,
+      "loss": 1.0235,
+      "step": 2683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2870195128020576,
+      "learning_rate": 0.00019665744533048328,
+      "loss": 1.0422,
+      "step": 2684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27640315301309837,
+      "learning_rate": 0.0001966533882237116,
+      "loss": 1.1165,
+      "step": 2685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2694095090611699,
+      "learning_rate": 0.00019664932869812814,
+      "loss": 1.0239,
+      "step": 2686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.32106995191546245,
+      "learning_rate": 0.0001966452667538345,
+      "loss": 1.1492,
+      "step": 2687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27424915699699154,
+      "learning_rate": 0.00019664120239093233,
+      "loss": 1.1685,
+      "step": 2688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29135122016294623,
+      "learning_rate": 0.0001966371356095233,
+      "loss": 1.1547,
+      "step": 2689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26527031382403177,
+      "learning_rate": 0.00019663306640970926,
+      "loss": 1.1219,
+      "step": 2690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27127353395064424,
+      "learning_rate": 0.00019662899479159197,
+      "loss": 0.9899,
+      "step": 2691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3033029254082761,
+      "learning_rate": 0.00019662492075527336,
+      "loss": 1.1249,
+      "step": 2692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26657961186508666,
+      "learning_rate": 0.00019662084430085538,
+      "loss": 1.1842,
+      "step": 2693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28599030101340955,
+      "learning_rate": 0.00019661676542844007,
+      "loss": 1.1839,
+      "step": 2694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30087762030890813,
+      "learning_rate": 0.00019661268413812946,
+      "loss": 1.0863,
+      "step": 2695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.325088281089176,
+      "learning_rate": 0.00019660860043002574,
+      "loss": 1.049,
+      "step": 2696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2828097114488758,
+      "learning_rate": 0.00019660451430423103,
+      "loss": 1.1857,
+      "step": 2697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2701302351874178,
+      "learning_rate": 0.00019660042576084767,
+      "loss": 1.1144,
+      "step": 2698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2835386842245424,
+      "learning_rate": 0.00019659633479997794,
+      "loss": 1.1066,
+      "step": 2699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.243379840423929,
+      "learning_rate": 0.00019659224142172424,
+      "loss": 1.0204,
+      "step": 2700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24266059096127535,
+      "learning_rate": 0.00019658814562618896,
+      "loss": 1.1088,
+      "step": 2701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2617930016670921,
+      "learning_rate": 0.00019658404741347462,
+      "loss": 1.1113,
+      "step": 2702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2547097198503651,
+      "learning_rate": 0.00019657994678368385,
+      "loss": 1.0444,
+      "step": 2703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2625250549016416,
+      "learning_rate": 0.00019657584373691917,
+      "loss": 1.1396,
+      "step": 2704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26689206957202594,
+      "learning_rate": 0.0001965717382732833,
+      "loss": 1.1596,
+      "step": 2705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2903595268276037,
+      "learning_rate": 0.000196567630392879,
+      "loss": 1.0474,
+      "step": 2706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2572495572058499,
+      "learning_rate": 0.00019656352009580908,
+      "loss": 1.0761,
+      "step": 2707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2470130498530773,
+      "learning_rate": 0.00019655940738217635,
+      "loss": 1.0546,
+      "step": 2708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2930008567798905,
+      "learning_rate": 0.00019655529225208378,
+      "loss": 1.1107,
+      "step": 2709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25989997031388734,
+      "learning_rate": 0.00019655117470563434,
+      "loss": 1.0839,
+      "step": 2710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.30098329989662703,
+      "learning_rate": 0.00019654705474293107,
+      "loss": 1.1056,
+      "step": 2711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2424595664908627,
+      "learning_rate": 0.00019654293236407707,
+      "loss": 1.0923,
+      "step": 2712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28519149075066125,
+      "learning_rate": 0.00019653880756917552,
+      "loss": 1.0625,
+      "step": 2713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31144165591888,
+      "learning_rate": 0.00019653468035832965,
+      "loss": 1.2098,
+      "step": 2714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2714581157840239,
+      "learning_rate": 0.0001965305507316427,
+      "loss": 1.092,
+      "step": 2715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28440142688999087,
+      "learning_rate": 0.0001965264186892181,
+      "loss": 1.0165,
+      "step": 2716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24262222631275346,
+      "learning_rate": 0.00019652228423115917,
+      "loss": 1.0364,
+      "step": 2717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2743500722654148,
+      "learning_rate": 0.00019651814735756942,
+      "loss": 1.0864,
+      "step": 2718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28526861621648486,
+      "learning_rate": 0.00019651400806855237,
+      "loss": 1.1124,
+      "step": 2719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31205779575841586,
+      "learning_rate": 0.00019650986636421164,
+      "loss": 1.102,
+      "step": 2720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2661120177899195,
+      "learning_rate": 0.00019650572224465084,
+      "loss": 1.1081,
+      "step": 2721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29878434646252183,
+      "learning_rate": 0.00019650157570997364,
+      "loss": 1.1004,
+      "step": 2722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2612778406780732,
+      "learning_rate": 0.00019649742676028394,
+      "loss": 1.2379,
+      "step": 2723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2933566914172725,
+      "learning_rate": 0.00019649327539568543,
+      "loss": 1.162,
+      "step": 2724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24094136996286655,
+      "learning_rate": 0.0001964891216162821,
+      "loss": 1.0807,
+      "step": 2725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2943459316325484,
+      "learning_rate": 0.00019648496542217783,
+      "loss": 0.998,
+      "step": 2726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.22961741180973028,
+      "learning_rate": 0.00019648080681347664,
+      "loss": 1.1305,
+      "step": 2727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2701859846638287,
+      "learning_rate": 0.00019647664579028267,
+      "loss": 1.0889,
+      "step": 2728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26857093567006685,
+      "learning_rate": 0.0001964724823527,
+      "loss": 1.2013,
+      "step": 2729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31949382152840944,
+      "learning_rate": 0.0001964683165008328,
+      "loss": 1.1718,
+      "step": 2730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28367090867227224,
+      "learning_rate": 0.00019646414823478535,
+      "loss": 1.1491,
+      "step": 2731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2504701281061607,
+      "learning_rate": 0.000196459977554662,
+      "loss": 1.1017,
+      "step": 2732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25221421853430986,
+      "learning_rate": 0.00019645580446056706,
+      "loss": 1.1185,
+      "step": 2733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28143888782957394,
+      "learning_rate": 0.000196451628952605,
+      "loss": 1.1759,
+      "step": 2734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2578191227270372,
+      "learning_rate": 0.00019644745103088033,
+      "loss": 1.1787,
+      "step": 2735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26056470647250435,
+      "learning_rate": 0.00019644327069549754,
+      "loss": 1.1649,
+      "step": 2736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27176880852893454,
+      "learning_rate": 0.00019643908794656135,
+      "loss": 1.1057,
+      "step": 2737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2882235242635738,
+      "learning_rate": 0.00019643490278417632,
+      "loss": 1.1081,
+      "step": 2738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27263458978440536,
+      "learning_rate": 0.00019643071520844725,
+      "loss": 1.0712,
+      "step": 2739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25866965428198324,
+      "learning_rate": 0.00019642652521947894,
+      "loss": 1.1159,
+      "step": 2740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2630902327301495,
+      "learning_rate": 0.00019642233281737625,
+      "loss": 1.0603,
+      "step": 2741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25820714973330133,
+      "learning_rate": 0.00019641813800224406,
+      "loss": 0.9922,
+      "step": 2742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2724772043679332,
+      "learning_rate": 0.00019641394077418736,
+      "loss": 1.0461,
+      "step": 2743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27088818855216434,
+      "learning_rate": 0.00019640974113331123,
+      "loss": 1.1959,
+      "step": 2744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2733607733632018,
+      "learning_rate": 0.00019640553907972072,
+      "loss": 1.1886,
+      "step": 2745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24676149904011196,
+      "learning_rate": 0.000196401334613521,
+      "loss": 1.0969,
+      "step": 2746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27834584403561474,
+      "learning_rate": 0.00019639712773481728,
+      "loss": 1.1597,
+      "step": 2747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25755736145056124,
+      "learning_rate": 0.0001963929184437149,
+      "loss": 0.9978,
+      "step": 2748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.31079398376305944,
+      "learning_rate": 0.00019638870674031913,
+      "loss": 1.1268,
+      "step": 2749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2774704968716273,
+      "learning_rate": 0.0001963844926247354,
+      "loss": 1.1641,
+      "step": 2750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2781649290866802,
+      "learning_rate": 0.00019638027609706916,
+      "loss": 1.0707,
+      "step": 2751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2552137884918176,
+      "learning_rate": 0.00019637605715742593,
+      "loss": 1.2287,
+      "step": 2752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2599068780978502,
+      "learning_rate": 0.00019637183580591133,
+      "loss": 1.0331,
+      "step": 2753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.27220701541938597,
+      "learning_rate": 0.00019636761204263093,
+      "loss": 1.0395,
+      "step": 2754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.24397827889832657,
+      "learning_rate": 0.0001963633858676905,
+      "loss": 1.1166,
+      "step": 2755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2497024411579182,
+      "learning_rate": 0.00019635915728119575,
+      "loss": 1.0847,
+      "step": 2756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2589117765444782,
+      "learning_rate": 0.00019635492628325256,
+      "loss": 1.0086,
+      "step": 2757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29313802184742693,
+      "learning_rate": 0.00019635069287396678,
+      "loss": 1.1229,
+      "step": 2758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2618259520734273,
+      "learning_rate": 0.00019634645705344435,
+      "loss": 1.1459,
+      "step": 2759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2454773691105509,
+      "learning_rate": 0.0001963422188217913,
+      "loss": 0.968,
+      "step": 2760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.28372490824875446,
+      "learning_rate": 0.00019633797817911365,
+      "loss": 1.1127,
+      "step": 2761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.3273752692424284,
+      "learning_rate": 0.00019633373512551754,
+      "loss": 1.1669,
+      "step": 2762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2366079234217832,
+      "learning_rate": 0.0001963294896611092,
+      "loss": 1.0848,
+      "step": 2763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5261995752384835,
+      "learning_rate": 0.00019632524178599483,
+      "loss": 1.1333,
+      "step": 2764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.26665251541131973,
+      "learning_rate": 0.00019632099150028074,
+      "loss": 1.0651,
+      "step": 2765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2584540431624957,
+      "learning_rate": 0.0001963167388040733,
+      "loss": 1.107,
+      "step": 2766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.2634462477386324,
+      "learning_rate": 0.00019631248369747893,
+      "loss": 1.124,
+      "step": 2767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.29816863160086016,
+      "learning_rate": 0.00019630822618060413,
+      "loss": 1.1173,
+      "step": 2768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.25963330208582736,
+      "learning_rate": 0.00019630396625355546,
+      "loss": 1.0285,
+      "step": 2769
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2887065947096526,
+      "learning_rate": 0.00019629970391643947,
+      "loss": 1.1987,
+      "step": 2770
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27628742653188076,
+      "learning_rate": 0.0001962954391693629,
+      "loss": 1.067,
+      "step": 2771
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2707072531440452,
+      "learning_rate": 0.00019629117201243242,
+      "loss": 0.9484,
+      "step": 2772
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2973518749444581,
+      "learning_rate": 0.0001962869024457549,
+      "loss": 1.0383,
+      "step": 2773
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26365187418455077,
+      "learning_rate": 0.0001962826304694371,
+      "loss": 1.0947,
+      "step": 2774
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2591589967435732,
+      "learning_rate": 0.00019627835608358596,
+      "loss": 1.07,
+      "step": 2775
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26492792550338157,
+      "learning_rate": 0.00019627407928830842,
+      "loss": 1.1614,
+      "step": 2776
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25376976035220183,
+      "learning_rate": 0.00019626980008371158,
+      "loss": 1.0263,
+      "step": 2777
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23447258204438867,
+      "learning_rate": 0.0001962655184699025,
+      "loss": 1.1122,
+      "step": 2778
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2562093782282185,
+      "learning_rate": 0.00019626123444698828,
+      "loss": 1.0457,
+      "step": 2779
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2723192435370688,
+      "learning_rate": 0.00019625694801507618,
+      "loss": 1.0636,
+      "step": 2780
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29084143851467165,
+      "learning_rate": 0.00019625265917427346,
+      "loss": 1.1492,
+      "step": 2781
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2759691568719512,
+      "learning_rate": 0.00019624836792468746,
+      "loss": 1.0947,
+      "step": 2782
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2847571657532783,
+      "learning_rate": 0.00019624407426642557,
+      "loss": 1.0998,
+      "step": 2783
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2821857453688257,
+      "learning_rate": 0.00019623977819959522,
+      "loss": 1.0525,
+      "step": 2784
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2710683595054374,
+      "learning_rate": 0.00019623547972430394,
+      "loss": 1.1189,
+      "step": 2785
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2650861466646489,
+      "learning_rate": 0.00019623117884065932,
+      "loss": 1.0532,
+      "step": 2786
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28632313517563235,
+      "learning_rate": 0.00019622687554876893,
+      "loss": 1.0432,
+      "step": 2787
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2927818066954993,
+      "learning_rate": 0.00019622256984874053,
+      "loss": 1.0847,
+      "step": 2788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2631443348972877,
+      "learning_rate": 0.00019621826174068185,
+      "loss": 1.1038,
+      "step": 2789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26694650397049624,
+      "learning_rate": 0.00019621395122470066,
+      "loss": 0.9954,
+      "step": 2790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27909721930965364,
+      "learning_rate": 0.00019620963830090492,
+      "loss": 0.9486,
+      "step": 2791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24079022570344805,
+      "learning_rate": 0.0001962053229694025,
+      "loss": 0.9676,
+      "step": 2792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2559668456069455,
+      "learning_rate": 0.0001962010052303014,
+      "loss": 1.1104,
+      "step": 2793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25016354937216784,
+      "learning_rate": 0.0001961966850837097,
+      "loss": 1.1744,
+      "step": 2794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25412464453761213,
+      "learning_rate": 0.0001961923625297355,
+      "loss": 1.1563,
+      "step": 2795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25700418834836664,
+      "learning_rate": 0.00019618803756848695,
+      "loss": 1.155,
+      "step": 2796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29832287781586436,
+      "learning_rate": 0.0001961837102000723,
+      "loss": 1.0929,
+      "step": 2797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24969396689469361,
+      "learning_rate": 0.00019617938042459988,
+      "loss": 1.0692,
+      "step": 2798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27702787637233217,
+      "learning_rate": 0.00019617504824217803,
+      "loss": 1.0812,
+      "step": 2799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2900431406784246,
+      "learning_rate": 0.00019617071365291512,
+      "loss": 1.1213,
+      "step": 2800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2650773606437207,
+      "learning_rate": 0.0001961663766569197,
+      "loss": 1.0438,
+      "step": 2801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23346920952561745,
+      "learning_rate": 0.00019616203725430023,
+      "loss": 0.9972,
+      "step": 2802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2739900839005771,
+      "learning_rate": 0.00019615769544516532,
+      "loss": 1.0435,
+      "step": 2803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3190423287534823,
+      "learning_rate": 0.00019615335122962372,
+      "loss": 1.1342,
+      "step": 2804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2830709781599123,
+      "learning_rate": 0.00019614900460778403,
+      "loss": 1.0853,
+      "step": 2805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26452593616904946,
+      "learning_rate": 0.00019614465557975507,
+      "loss": 1.1323,
+      "step": 2806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2592995682321865,
+      "learning_rate": 0.00019614030414564568,
+      "loss": 1.068,
+      "step": 2807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2827474115219191,
+      "learning_rate": 0.00019613595030556477,
+      "loss": 1.1488,
+      "step": 2808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2649963078693095,
+      "learning_rate": 0.0001961315940596213,
+      "loss": 0.996,
+      "step": 2809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26346285838705263,
+      "learning_rate": 0.00019612723540792426,
+      "loss": 1.1175,
+      "step": 2810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28954499176521176,
+      "learning_rate": 0.00019612287435058273,
+      "loss": 1.1733,
+      "step": 2811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2761795119898941,
+      "learning_rate": 0.00019611851088770585,
+      "loss": 1.1321,
+      "step": 2812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.277378324643611,
+      "learning_rate": 0.00019611414501940284,
+      "loss": 1.291,
+      "step": 2813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29898305586569235,
+      "learning_rate": 0.00019610977674578296,
+      "loss": 1.1056,
+      "step": 2814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2728332232087226,
+      "learning_rate": 0.00019610540606695547,
+      "loss": 1.1051,
+      "step": 2815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3020216145302391,
+      "learning_rate": 0.0001961010329830298,
+      "loss": 0.9909,
+      "step": 2816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2614179979610706,
+      "learning_rate": 0.00019609665749411543,
+      "loss": 1.0927,
+      "step": 2817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2486123416932773,
+      "learning_rate": 0.00019609227960032177,
+      "loss": 1.1096,
+      "step": 2818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31135834548170077,
+      "learning_rate": 0.00019608789930175845,
+      "loss": 1.1178,
+      "step": 2819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.277755259399668,
+      "learning_rate": 0.00019608351659853503,
+      "loss": 1.1473,
+      "step": 2820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2584745367908558,
+      "learning_rate": 0.00019607913149076125,
+      "loss": 1.0122,
+      "step": 2821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2619607556881495,
+      "learning_rate": 0.0001960747439785468,
+      "loss": 1.128,
+      "step": 2822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29050519695473265,
+      "learning_rate": 0.00019607035406200152,
+      "loss": 1.0491,
+      "step": 2823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.21812110204389035,
+      "learning_rate": 0.00019606596174123525,
+      "loss": 1.1075,
+      "step": 2824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2839251249015717,
+      "learning_rate": 0.00019606156701635792,
+      "loss": 1.1837,
+      "step": 2825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2958308386052112,
+      "learning_rate": 0.0001960571698874795,
+      "loss": 1.0984,
+      "step": 2826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2671536081657146,
+      "learning_rate": 0.00019605277035470998,
+      "loss": 1.0137,
+      "step": 2827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.260771384058964,
+      "learning_rate": 0.00019604836841815958,
+      "loss": 1.0675,
+      "step": 2828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2615419961091573,
+      "learning_rate": 0.00019604396407793835,
+      "loss": 1.055,
+      "step": 2829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.27494047722891274,
+      "learning_rate": 0.0001960395573341566,
+      "loss": 1.1479,
+      "step": 2830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3121477873782516,
+      "learning_rate": 0.00019603514818692454,
+      "loss": 1.0033,
+      "step": 2831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26737638033087213,
+      "learning_rate": 0.00019603073663635256,
+      "loss": 1.1077,
+      "step": 2832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24631307751232157,
+      "learning_rate": 0.00019602632268255103,
+      "loss": 1.0545,
+      "step": 2833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3013621704790624,
+      "learning_rate": 0.00019602190632563043,
+      "loss": 1.0969,
+      "step": 2834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2783637530154318,
+      "learning_rate": 0.00019601748756570126,
+      "loss": 1.0622,
+      "step": 2835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2574957740041804,
+      "learning_rate": 0.00019601306640287415,
+      "loss": 1.101,
+      "step": 2836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26247517047321034,
+      "learning_rate": 0.00019600864283725967,
+      "loss": 1.0651,
+      "step": 2837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25876279601882096,
+      "learning_rate": 0.0001960042168689686,
+      "loss": 1.062,
+      "step": 2838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25523429862063185,
+      "learning_rate": 0.00019599978849811164,
+      "loss": 1.1267,
+      "step": 2839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23537001704505256,
+      "learning_rate": 0.00019599535772479968,
+      "loss": 1.0823,
+      "step": 2840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2462574908056951,
+      "learning_rate": 0.00019599092454914351,
+      "loss": 1.0301,
+      "step": 2841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2721311709805547,
+      "learning_rate": 0.00019598648897125416,
+      "loss": 1.067,
+      "step": 2842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3069559448268241,
+      "learning_rate": 0.0001959820509912426,
+      "loss": 1.013,
+      "step": 2843
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2702912452683172,
+      "learning_rate": 0.00019597761060921985,
+      "loss": 1.1137,
+      "step": 2844
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2797213146710235,
+      "learning_rate": 0.00019597316782529715,
+      "loss": 1.0939,
+      "step": 2845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2787776466517135,
+      "learning_rate": 0.00019596872263958552,
+      "loss": 1.1058,
+      "step": 2846
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26116233626175234,
+      "learning_rate": 0.00019596427505219635,
+      "loss": 1.0094,
+      "step": 2847
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2693281423942913,
+      "learning_rate": 0.0001959598250632409,
+      "loss": 1.1635,
+      "step": 2848
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2506743951314745,
+      "learning_rate": 0.00019595537267283047,
+      "loss": 1.1075,
+      "step": 2849
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24106080467021557,
+      "learning_rate": 0.00019595091788107656,
+      "loss": 0.9761,
+      "step": 2850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2860007338075784,
+      "learning_rate": 0.0001959464606880906,
+      "loss": 1.0474,
+      "step": 2851
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2359433024977078,
+      "learning_rate": 0.00019594200109398417,
+      "loss": 1.0804,
+      "step": 2852
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24715768704519478,
+      "learning_rate": 0.0001959375390988689,
+      "loss": 1.1157,
+      "step": 2853
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2586351713329031,
+      "learning_rate": 0.0001959330747028564,
+      "loss": 1.0448,
+      "step": 2854
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25499580779707726,
+      "learning_rate": 0.00019592860790605842,
+      "loss": 1.1657,
+      "step": 2855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.29734674634840746,
+      "learning_rate": 0.0001959241387085867,
+      "loss": 1.1446,
+      "step": 2856
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2647465723279589,
+      "learning_rate": 0.00019591966711055315,
+      "loss": 1.1668,
+      "step": 2857
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26368966136330935,
+      "learning_rate": 0.00019591519311206964,
+      "loss": 1.1992,
+      "step": 2858
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3126849078518487,
+      "learning_rate": 0.00019591071671324817,
+      "loss": 1.141,
+      "step": 2859
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25583665345229606,
+      "learning_rate": 0.00019590623791420071,
+      "loss": 1.1441,
+      "step": 2860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23927224607942188,
+      "learning_rate": 0.00019590175671503938,
+      "loss": 1.1212,
+      "step": 2861
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28570558886340036,
+      "learning_rate": 0.00019589727311587632,
+      "loss": 1.0623,
+      "step": 2862
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.23141130803687024,
+      "learning_rate": 0.00019589278711682373,
+      "loss": 1.1051,
+      "step": 2863
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.31126189932366843,
+      "learning_rate": 0.00019588829871799388,
+      "loss": 1.098,
+      "step": 2864
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2368691173217574,
+      "learning_rate": 0.00019588380791949906,
+      "loss": 0.937,
+      "step": 2865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2905469407114906,
+      "learning_rate": 0.0001958793147214517,
+      "loss": 1.1837,
+      "step": 2866
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.25681129260211033,
+      "learning_rate": 0.00019587481912396426,
+      "loss": 1.0659,
+      "step": 2867
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.3115969168588661,
+      "learning_rate": 0.0001958703211271492,
+      "loss": 1.1245,
+      "step": 2868
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2629500967052983,
+      "learning_rate": 0.0001958658207311191,
+      "loss": 1.0451,
+      "step": 2869
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.26897588144338785,
+      "learning_rate": 0.0001958613179359866,
+      "loss": 1.0232,
+      "step": 2870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2745821063034373,
+      "learning_rate": 0.00019585681274186434,
+      "loss": 1.1058,
+      "step": 2871
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.28134411874328125,
+      "learning_rate": 0.00019585230514886513,
+      "loss": 1.0646,
+      "step": 2872
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.2739086818209478,
+      "learning_rate": 0.0001958477951571017,
+      "loss": 1.0774,
+      "step": 2873
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.24039032053442347,
+      "learning_rate": 0.000195843282766687,
+      "loss": 1.1367,
+      "step": 2874
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2729224048445481,
+      "learning_rate": 0.00019583876797773391,
+      "loss": 1.0894,
+      "step": 2875
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2548563216514712,
+      "learning_rate": 0.0001958342507903554,
+      "loss": 1.1571,
+      "step": 2876
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2611894266017868,
+      "learning_rate": 0.00019582973120466454,
+      "loss": 1.1219,
+      "step": 2877
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2517135134023402,
+      "learning_rate": 0.00019582520922077444,
+      "loss": 1.1457,
+      "step": 2878
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25685092446562596,
+      "learning_rate": 0.00019582068483879822,
+      "loss": 1.0947,
+      "step": 2879
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2531800123875422,
+      "learning_rate": 0.00019581615805884918,
+      "loss": 1.0824,
+      "step": 2880
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29746440693794585,
+      "learning_rate": 0.00019581162888104056,
+      "loss": 1.127,
+      "step": 2881
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2951013215212664,
+      "learning_rate": 0.0001958070973054857,
+      "loss": 0.9708,
+      "step": 2882
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26703604465532904,
+      "learning_rate": 0.00019580256333229804,
+      "loss": 1.0996,
+      "step": 2883
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3054149559628321,
+      "learning_rate": 0.00019579802696159098,
+      "loss": 1.1972,
+      "step": 2884
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2554844575267877,
+      "learning_rate": 0.00019579348819347814,
+      "loss": 1.1815,
+      "step": 2885
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27725302476518116,
+      "learning_rate": 0.00019578894702807303,
+      "loss": 1.1779,
+      "step": 2886
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2812121223614431,
+      "learning_rate": 0.0001957844034654893,
+      "loss": 1.0931,
+      "step": 2887
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28657879183155893,
+      "learning_rate": 0.0001957798575058407,
+      "loss": 1.209,
+      "step": 2888
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32607385402952277,
+      "learning_rate": 0.00019577530914924096,
+      "loss": 1.083,
+      "step": 2889
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2622053283912726,
+      "learning_rate": 0.00019577075839580395,
+      "loss": 1.1331,
+      "step": 2890
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27330452773103864,
+      "learning_rate": 0.00019576620524564347,
+      "loss": 1.1223,
+      "step": 2891
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28722568397660336,
+      "learning_rate": 0.00019576164969887353,
+      "loss": 1.0848,
+      "step": 2892
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25111702949361947,
+      "learning_rate": 0.00019575709175560815,
+      "loss": 1.1015,
+      "step": 2893
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27784956399299426,
+      "learning_rate": 0.00019575253141596136,
+      "loss": 1.0712,
+      "step": 2894
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.29027995612195606,
+      "learning_rate": 0.00019574796868004728,
+      "loss": 1.0522,
+      "step": 2895
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2662180849874586,
+      "learning_rate": 0.00019574340354798012,
+      "loss": 1.0711,
+      "step": 2896
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.272996240282476,
+      "learning_rate": 0.00019573883601987409,
+      "loss": 1.1081,
+      "step": 2897
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2500926481674787,
+      "learning_rate": 0.00019573426609584353,
+      "loss": 1.0818,
+      "step": 2898
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2690599907288768,
+      "learning_rate": 0.00019572969377600278,
+      "loss": 1.1512,
+      "step": 2899
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26895706721452967,
+      "learning_rate": 0.00019572511906046632,
+      "loss": 1.106,
+      "step": 2900
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28870985265507426,
+      "learning_rate": 0.00019572054194934855,
+      "loss": 1.0406,
+      "step": 2901
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2569564435809099,
+      "learning_rate": 0.00019571596244276408,
+      "loss": 1.1162,
+      "step": 2902
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26799107456705956,
+      "learning_rate": 0.0001957113805408275,
+      "loss": 1.0301,
+      "step": 2903
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2476133822775531,
+      "learning_rate": 0.00019570679624365348,
+      "loss": 1.1889,
+      "step": 2904
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28688265263367885,
+      "learning_rate": 0.00019570220955135673,
+      "loss": 1.0879,
+      "step": 2905
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.279647668676218,
+      "learning_rate": 0.000195697620464052,
+      "loss": 1.1741,
+      "step": 2906
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26565186704575483,
+      "learning_rate": 0.0001956930289818542,
+      "loss": 1.0533,
+      "step": 2907
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27848478850209946,
+      "learning_rate": 0.00019568843510487822,
+      "loss": 1.0685,
+      "step": 2908
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27824963257277385,
+      "learning_rate": 0.00019568383883323902,
+      "loss": 1.154,
+      "step": 2909
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2889721158688345,
+      "learning_rate": 0.0001956792401670516,
+      "loss": 1.0229,
+      "step": 2910
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2433224132508536,
+      "learning_rate": 0.00019567463910643106,
+      "loss": 1.0934,
+      "step": 2911
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2558852060337151,
+      "learning_rate": 0.00019567003565149256,
+      "loss": 1.035,
+      "step": 2912
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25619363373049336,
+      "learning_rate": 0.0001956654298023513,
+      "loss": 1.1415,
+      "step": 2913
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2913829589793719,
+      "learning_rate": 0.0001956608215591225,
+      "loss": 1.0902,
+      "step": 2914
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25636509940637525,
+      "learning_rate": 0.00019565621092192156,
+      "loss": 1.0804,
+      "step": 2915
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2911863913856473,
+      "learning_rate": 0.00019565159789086377,
+      "loss": 1.0234,
+      "step": 2916
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3098448701698118,
+      "learning_rate": 0.00019564698246606467,
+      "loss": 1.018,
+      "step": 2917
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28894296908926365,
+      "learning_rate": 0.00019564236464763971,
+      "loss": 1.0444,
+      "step": 2918
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2969290728405071,
+      "learning_rate": 0.00019563774443570448,
+      "loss": 1.0826,
+      "step": 2919
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26737166147445435,
+      "learning_rate": 0.00019563312183037458,
+      "loss": 1.1668,
+      "step": 2920
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28206692002567496,
+      "learning_rate": 0.0001956284968317657,
+      "loss": 1.0922,
+      "step": 2921
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.24812211567721631,
+      "learning_rate": 0.0001956238694399936,
+      "loss": 1.1432,
+      "step": 2922
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2599840736277116,
+      "learning_rate": 0.00019561923965517405,
+      "loss": 1.0521,
+      "step": 2923
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2543940271199818,
+      "learning_rate": 0.00019561460747742295,
+      "loss": 1.1435,
+      "step": 2924
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27672619084132033,
+      "learning_rate": 0.0001956099729068562,
+      "loss": 1.0804,
+      "step": 2925
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.257274251896411,
+      "learning_rate": 0.0001956053359435898,
+      "loss": 1.1605,
+      "step": 2926
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2700086089277536,
+      "learning_rate": 0.00019560069658773976,
+      "loss": 1.1006,
+      "step": 2927
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2729066392922699,
+      "learning_rate": 0.00019559605483942223,
+      "loss": 0.9848,
+      "step": 2928
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2819343853836847,
+      "learning_rate": 0.0001955914106987533,
+      "loss": 1.0987,
+      "step": 2929
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28151027604586293,
+      "learning_rate": 0.00019558676416584929,
+      "loss": 1.1282,
+      "step": 2930
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26677614820796297,
+      "learning_rate": 0.0001955821152408264,
+      "loss": 1.1034,
+      "step": 2931
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3059666303412851,
+      "learning_rate": 0.00019557746392380104,
+      "loss": 1.1612,
+      "step": 2932
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2819684123621462,
+      "learning_rate": 0.00019557281021488957,
+      "loss": 1.0681,
+      "step": 2933
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.23809460047261669,
+      "learning_rate": 0.00019556815411420842,
+      "loss": 1.0539,
+      "step": 2934
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26195198259872626,
+      "learning_rate": 0.0001955634956218742,
+      "loss": 1.0663,
+      "step": 2935
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2776958285854396,
+      "learning_rate": 0.00019555883473800344,
+      "loss": 1.2042,
+      "step": 2936
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2776878056257854,
+      "learning_rate": 0.00019555417146271275,
+      "loss": 1.0723,
+      "step": 2937
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.22289840395374008,
+      "learning_rate": 0.00019554950579611888,
+      "loss": 1.101,
+      "step": 2938
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2765820638955678,
+      "learning_rate": 0.00019554483773833855,
+      "loss": 1.1081,
+      "step": 2939
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.313800298123316,
+      "learning_rate": 0.00019554016728948865,
+      "loss": 1.1322,
+      "step": 2940
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25205600427022146,
+      "learning_rate": 0.00019553549444968602,
+      "loss": 1.0419,
+      "step": 2941
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.271795514200843,
+      "learning_rate": 0.00019553081921904757,
+      "loss": 1.1375,
+      "step": 2942
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28101236125294443,
+      "learning_rate": 0.00019552614159769034,
+      "loss": 1.0403,
+      "step": 2943
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25207664771282795,
+      "learning_rate": 0.0001955214615857314,
+      "loss": 1.0785,
+      "step": 2944
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27380543511172994,
+      "learning_rate": 0.00019551677918328784,
+      "loss": 1.1187,
+      "step": 2945
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2732262776480482,
+      "learning_rate": 0.00019551209439047683,
+      "loss": 1.1426,
+      "step": 2946
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25773244127240097,
+      "learning_rate": 0.00019550740720741564,
+      "loss": 1.019,
+      "step": 2947
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2649249368780316,
+      "learning_rate": 0.0001955027176342216,
+      "loss": 1.0553,
+      "step": 2948
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2778435224973774,
+      "learning_rate": 0.00019549802567101198,
+      "loss": 1.1484,
+      "step": 2949
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2835832433902629,
+      "learning_rate": 0.00019549333131790427,
+      "loss": 1.1626,
+      "step": 2950
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2684901729578864,
+      "learning_rate": 0.00019548863457501592,
+      "loss": 1.0469,
+      "step": 2951
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2917738619766324,
+      "learning_rate": 0.0001954839354424645,
+      "loss": 1.0706,
+      "step": 2952
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2719180117243129,
+      "learning_rate": 0.00019547923392036756,
+      "loss": 1.069,
+      "step": 2953
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.32365658454747653,
+      "learning_rate": 0.00019547453000884278,
+      "loss": 1.2248,
+      "step": 2954
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.22919570649820376,
+      "learning_rate": 0.0001954698237080079,
+      "loss": 0.9762,
+      "step": 2955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.27049554069580856,
+      "learning_rate": 0.00019546511501798068,
+      "loss": 1.0445,
+      "step": 2956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2321876330280462,
+      "learning_rate": 0.00019546040393887896,
+      "loss": 0.9582,
+      "step": 2957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2866826620321833,
+      "learning_rate": 0.00019545569047082063,
+      "loss": 1.0803,
+      "step": 2958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2513540583580118,
+      "learning_rate": 0.00019545097461392364,
+      "loss": 1.083,
+      "step": 2959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26654262390528605,
+      "learning_rate": 0.00019544625636830606,
+      "loss": 1.0319,
+      "step": 2960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2413056145935159,
+      "learning_rate": 0.00019544153573408592,
+      "loss": 1.158,
+      "step": 2961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26314675374807356,
+      "learning_rate": 0.00019543681271138135,
+      "loss": 1.1581,
+      "step": 2962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.302120035744808,
+      "learning_rate": 0.00019543208730031056,
+      "loss": 1.0621,
+      "step": 2963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2784231038369151,
+      "learning_rate": 0.0001954273595009918,
+      "loss": 1.1521,
+      "step": 2964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3301489966327534,
+      "learning_rate": 0.00019542262931354342,
+      "loss": 1.1991,
+      "step": 2965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2625920950709,
+      "learning_rate": 0.00019541789673808378,
+      "loss": 1.1439,
+      "step": 2966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2799876119710994,
+      "learning_rate": 0.00019541316177473127,
+      "loss": 1.2343,
+      "step": 2967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.25861702387999425,
+      "learning_rate": 0.00019540842442360444,
+      "loss": 1.0334,
+      "step": 2968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26861453164120885,
+      "learning_rate": 0.00019540368468482183,
+      "loss": 1.0876,
+      "step": 2969
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2790820406297911,
+      "learning_rate": 0.00019539894255850203,
+      "loss": 1.2192,
+      "step": 2970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26958726146743006,
+      "learning_rate": 0.00019539419804476377,
+      "loss": 1.071,
+      "step": 2971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2677639222097805,
+      "learning_rate": 0.00019538945114372573,
+      "loss": 1.2223,
+      "step": 2972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.26245494649962503,
+      "learning_rate": 0.00019538470185550674,
+      "loss": 1.105,
+      "step": 2973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2693830042475195,
+      "learning_rate": 0.00019537995018022563,
+      "loss": 1.1118,
+      "step": 2974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.24631907921671092,
+      "learning_rate": 0.0001953751961180013,
+      "loss": 1.0035,
+      "step": 2975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.30613365197670445,
+      "learning_rate": 0.00019537043966895277,
+      "loss": 1.0775,
+      "step": 2976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.2776873480898374,
+      "learning_rate": 0.00019536568083319903,
+      "loss": 1.0197,
+      "step": 2977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.28885110017090304,
+      "learning_rate": 0.00019536091961085922,
+      "loss": 1.1091,
+      "step": 2978
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2567410212869977,
+      "learning_rate": 0.00019535615600205247,
+      "loss": 1.0443,
+      "step": 2979
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28945345981693216,
+      "learning_rate": 0.00019535139000689795,
+      "loss": 1.0625,
+      "step": 2980
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2774312492094651,
+      "learning_rate": 0.000195346621625515,
+      "loss": 1.1077,
+      "step": 2981
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26395699634012787,
+      "learning_rate": 0.00019534185085802293,
+      "loss": 1.0201,
+      "step": 2982
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28496537247736714,
+      "learning_rate": 0.0001953370777045411,
+      "loss": 1.1266,
+      "step": 2983
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3243637973540252,
+      "learning_rate": 0.00019533230216518897,
+      "loss": 1.1888,
+      "step": 2984
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29081576092276984,
+      "learning_rate": 0.00019532752424008607,
+      "loss": 1.0315,
+      "step": 2985
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2850948882845963,
+      "learning_rate": 0.00019532274392935198,
+      "loss": 1.0013,
+      "step": 2986
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2923430922930697,
+      "learning_rate": 0.0001953179612331063,
+      "loss": 0.9897,
+      "step": 2987
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29321038881690914,
+      "learning_rate": 0.00019531317615146873,
+      "loss": 1.1548,
+      "step": 2988
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28328990204826776,
+      "learning_rate": 0.00019530838868455906,
+      "loss": 1.0857,
+      "step": 2989
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25769647168620946,
+      "learning_rate": 0.00019530359883249701,
+      "loss": 1.025,
+      "step": 2990
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2796837930311345,
+      "learning_rate": 0.00019529880659540256,
+      "loss": 1.1879,
+      "step": 2991
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2787744820467792,
+      "learning_rate": 0.00019529401197339557,
+      "loss": 1.1248,
+      "step": 2992
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.269889282961234,
+      "learning_rate": 0.00019528921496659603,
+      "loss": 1.0331,
+      "step": 2993
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28222433704753624,
+      "learning_rate": 0.00019528441557512398,
+      "loss": 1.0523,
+      "step": 2994
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27863870289129955,
+      "learning_rate": 0.00019527961379909957,
+      "loss": 1.1745,
+      "step": 2995
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31883034906636215,
+      "learning_rate": 0.00019527480963864294,
+      "loss": 1.0541,
+      "step": 2996
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27111306738362173,
+      "learning_rate": 0.0001952700030938743,
+      "loss": 1.1796,
+      "step": 2997
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27408759182730913,
+      "learning_rate": 0.00019526519416491401,
+      "loss": 1.0041,
+      "step": 2998
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24522235103299234,
+      "learning_rate": 0.0001952603828518823,
+      "loss": 0.8971,
+      "step": 2999
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2583032592016034,
+      "learning_rate": 0.00019525556915489967,
+      "loss": 0.9294,
+      "step": 3000
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25624359117015355,
+      "learning_rate": 0.00019525075307408655,
+      "loss": 0.9701,
+      "step": 3001
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28720231190552714,
+      "learning_rate": 0.0001952459346095635,
+      "loss": 1.0356,
+      "step": 3002
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29696988241107114,
+      "learning_rate": 0.00019524111376145105,
+      "loss": 1.0428,
+      "step": 3003
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26750283784091977,
+      "learning_rate": 0.00019523629052986988,
+      "loss": 1.0176,
+      "step": 3004
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28397737724131106,
+      "learning_rate": 0.00019523146491494067,
+      "loss": 1.0977,
+      "step": 3005
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30294167237041875,
+      "learning_rate": 0.0001952266369167842,
+      "loss": 1.1875,
+      "step": 3006
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2580269760608724,
+      "learning_rate": 0.00019522180653552132,
+      "loss": 1.0923,
+      "step": 3007
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2788977172196587,
+      "learning_rate": 0.00019521697377127285,
+      "loss": 1.1057,
+      "step": 3008
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26291500633347537,
+      "learning_rate": 0.00019521213862415979,
+      "loss": 1.1551,
+      "step": 3009
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2760993256298351,
+      "learning_rate": 0.00019520730109430314,
+      "loss": 1.0357,
+      "step": 3010
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2851223724038372,
+      "learning_rate": 0.0001952024611818239,
+      "loss": 1.1088,
+      "step": 3011
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2613411081969399,
+      "learning_rate": 0.00019519761888684326,
+      "loss": 1.0058,
+      "step": 3012
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2374352044099212,
+      "learning_rate": 0.0001951927742094824,
+      "loss": 1.0545,
+      "step": 3013
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27080671804410106,
+      "learning_rate": 0.00019518792714986254,
+      "loss": 1.1475,
+      "step": 3014
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2993708318879525,
+      "learning_rate": 0.00019518307770810496,
+      "loss": 1.0931,
+      "step": 3015
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29400732138156965,
+      "learning_rate": 0.00019517822588433102,
+      "loss": 1.0799,
+      "step": 3016
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25464534887846263,
+      "learning_rate": 0.0001951733716786622,
+      "loss": 0.993,
+      "step": 3017
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2652448178700676,
+      "learning_rate": 0.0001951685150912199,
+      "loss": 1.1835,
+      "step": 3018
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3077147987981465,
+      "learning_rate": 0.00019516365612212572,
+      "loss": 1.0706,
+      "step": 3019
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30685934393160413,
+      "learning_rate": 0.00019515879477150123,
+      "loss": 1.0244,
+      "step": 3020
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2856291230251649,
+      "learning_rate": 0.00019515393103946812,
+      "loss": 1.0963,
+      "step": 3021
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2767228605351276,
+      "learning_rate": 0.00019514906492614805,
+      "loss": 0.9945,
+      "step": 3022
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27035783848571304,
+      "learning_rate": 0.00019514419643166283,
+      "loss": 1.1075,
+      "step": 3023
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2990769279659998,
+      "learning_rate": 0.0001951393255561343,
+      "loss": 1.1556,
+      "step": 3024
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2844357764929915,
+      "learning_rate": 0.00019513445229968438,
+      "loss": 0.9933,
+      "step": 3025
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3070197133609208,
+      "learning_rate": 0.000195129576662435,
+      "loss": 1.0885,
+      "step": 3026
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2820008198156176,
+      "learning_rate": 0.0001951246986445082,
+      "loss": 1.0819,
+      "step": 3027
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27229352040640303,
+      "learning_rate": 0.00019511981824602598,
+      "loss": 1.046,
+      "step": 3028
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2739544259171004,
+      "learning_rate": 0.00019511493546711054,
+      "loss": 1.0647,
+      "step": 3029
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2620879572449313,
+      "learning_rate": 0.00019511005030788407,
+      "loss": 1.1027,
+      "step": 3030
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.251134914749705,
+      "learning_rate": 0.00019510516276846884,
+      "loss": 1.0464,
+      "step": 3031
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.23480951173895892,
+      "learning_rate": 0.0001951002728489871,
+      "loss": 1.0258,
+      "step": 3032
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2637238517781191,
+      "learning_rate": 0.0001950953805495613,
+      "loss": 1.0609,
+      "step": 3033
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.23941970843056096,
+      "learning_rate": 0.0001950904858703138,
+      "loss": 0.9995,
+      "step": 3034
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2674317195338645,
+      "learning_rate": 0.00019508558881136716,
+      "loss": 1.1166,
+      "step": 3035
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2671056081371556,
+      "learning_rate": 0.0001950806893728439,
+      "loss": 1.1134,
+      "step": 3036
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25637597126512307,
+      "learning_rate": 0.0001950757875548666,
+      "loss": 1.0847,
+      "step": 3037
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26941442259152115,
+      "learning_rate": 0.000195070883357558,
+      "loss": 1.1351,
+      "step": 3038
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26139367123922513,
+      "learning_rate": 0.00019506597678104078,
+      "loss": 1.1819,
+      "step": 3039
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2730769424356869,
+      "learning_rate": 0.00019506106782543774,
+      "loss": 1.0862,
+      "step": 3040
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2564183571077773,
+      "learning_rate": 0.00019505615649087173,
+      "loss": 1.057,
+      "step": 3041
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27496154581521765,
+      "learning_rate": 0.00019505124277746568,
+      "loss": 1.0365,
+      "step": 3042
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2906578614460428,
+      "learning_rate": 0.00019504632668534253,
+      "loss": 1.0765,
+      "step": 3043
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24889624819261374,
+      "learning_rate": 0.00019504140821462534,
+      "loss": 1.0847,
+      "step": 3044
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26592584153440635,
+      "learning_rate": 0.00019503648736543715,
+      "loss": 1.0803,
+      "step": 3045
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2944881481822344,
+      "learning_rate": 0.00019503156413790113,
+      "loss": 1.0591,
+      "step": 3046
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2918642575968384,
+      "learning_rate": 0.00019502663853214052,
+      "loss": 1.0976,
+      "step": 3047
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26006791100294435,
+      "learning_rate": 0.00019502171054827856,
+      "loss": 1.1608,
+      "step": 3048
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3177927171422205,
+      "learning_rate": 0.00019501678018643854,
+      "loss": 1.1429,
+      "step": 3049
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26535744168889774,
+      "learning_rate": 0.0001950118474467439,
+      "loss": 1.0564,
+      "step": 3050
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30516176813295376,
+      "learning_rate": 0.00019500691232931806,
+      "loss": 1.085,
+      "step": 3051
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3062397007408206,
+      "learning_rate": 0.00019500197483428454,
+      "loss": 1.202,
+      "step": 3052
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.30600809859823863,
+      "learning_rate": 0.0001949970349617669,
+      "loss": 1.0306,
+      "step": 3053
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26553989672712847,
+      "learning_rate": 0.00019499209271188874,
+      "loss": 1.1253,
+      "step": 3054
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3115655362592651,
+      "learning_rate": 0.00019498714808477375,
+      "loss": 1.0844,
+      "step": 3055
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2733439826555348,
+      "learning_rate": 0.00019498220108054573,
+      "loss": 1.0594,
+      "step": 3056
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27824607858191935,
+      "learning_rate": 0.00019497725169932839,
+      "loss": 1.1842,
+      "step": 3057
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29361249395554595,
+      "learning_rate": 0.00019497229994124563,
+      "loss": 1.156,
+      "step": 3058
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2909096666085275,
+      "learning_rate": 0.00019496734580642139,
+      "loss": 1.0713,
+      "step": 3059
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3008406748557583,
+      "learning_rate": 0.00019496238929497968,
+      "loss": 0.9974,
+      "step": 3060
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2557208115014124,
+      "learning_rate": 0.00019495743040704445,
+      "loss": 1.1056,
+      "step": 3061
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2619314513826558,
+      "learning_rate": 0.00019495246914273985,
+      "loss": 1.121,
+      "step": 3062
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2777726951083541,
+      "learning_rate": 0.00019494750550219,
+      "loss": 1.1167,
+      "step": 3063
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27592070412234426,
+      "learning_rate": 0.00019494253948551922,
+      "loss": 1.1754,
+      "step": 3064
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.27823194441497656,
+      "learning_rate": 0.0001949375710928517,
+      "loss": 1.0839,
+      "step": 3065
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2982548750310373,
+      "learning_rate": 0.00019493260032431176,
+      "loss": 1.1597,
+      "step": 3066
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2859522245926003,
+      "learning_rate": 0.00019492762718002386,
+      "loss": 1.1175,
+      "step": 3067
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25703939560705324,
+      "learning_rate": 0.00019492265166011244,
+      "loss": 1.0775,
+      "step": 3068
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28439233978147976,
+      "learning_rate": 0.000194917673764702,
+      "loss": 1.1447,
+      "step": 3069
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2794544690967338,
+      "learning_rate": 0.00019491269349391712,
+      "loss": 1.1776,
+      "step": 3070
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2438574837351192,
+      "learning_rate": 0.00019490771084788242,
+      "loss": 1.1019,
+      "step": 3071
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.24023569609710485,
+      "learning_rate": 0.00019490272582672262,
+      "loss": 1.1135,
+      "step": 3072
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29972615964686367,
+      "learning_rate": 0.00019489773843056244,
+      "loss": 1.069,
+      "step": 3073
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2589384956974427,
+      "learning_rate": 0.00019489274865952676,
+      "loss": 1.2025,
+      "step": 3074
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.2597338399690944,
+      "learning_rate": 0.00019488775651374038,
+      "loss": 1.0932,
+      "step": 3075
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.31065520260111457,
+      "learning_rate": 0.00019488276199332825,
+      "loss": 1.2195,
+      "step": 3076
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.25453729205708414,
+      "learning_rate": 0.0001948777650984154,
+      "loss": 1.107,
+      "step": 3077
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.29862291936090024,
+      "learning_rate": 0.00019487276582912683,
+      "loss": 1.1301,
+      "step": 3078
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3221818146284146,
+      "learning_rate": 0.00019486776418558766,
+      "loss": 1.1191,
+      "step": 3079
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3782095978448149,
+      "learning_rate": 0.0001948627601679231,
+      "loss": 1.0482,
+      "step": 3080
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.28378395622830654,
+      "learning_rate": 0.0001948577537762583,
+      "loss": 1.1376,
+      "step": 3081
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.3319042029552207,
+      "learning_rate": 0.00019485274501071864,
+      "loss": 1.0665,
+      "step": 3082
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.26347510714336875,
+      "learning_rate": 0.00019484773387142942,
+      "loss": 1.1515,
+      "step": 3083
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3101715271811368,
+      "learning_rate": 0.000194842720358516,
+      "loss": 1.1255,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25293432322802967,
+      "learning_rate": 0.00019483770447210397,
+      "loss": 1.0296,
+      "step": 3085
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28621886274041924,
+      "learning_rate": 0.00019483268621231875,
+      "loss": 0.9487,
+      "step": 3086
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2747791927139919,
+      "learning_rate": 0.00019482766557928592,
+      "loss": 1.0543,
+      "step": 3087
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26223503943348814,
+      "learning_rate": 0.00019482264257313122,
+      "loss": 1.0122,
+      "step": 3088
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27709490802994835,
+      "learning_rate": 0.00019481761719398027,
+      "loss": 1.119,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26196843932694963,
+      "learning_rate": 0.00019481258944195886,
+      "loss": 1.0707,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2782018365147699,
+      "learning_rate": 0.00019480755931719281,
+      "loss": 1.038,
+      "step": 3091
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24361390545083103,
+      "learning_rate": 0.00019480252681980802,
+      "loss": 1.056,
+      "step": 3092
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2674864730406895,
+      "learning_rate": 0.0001947974919499304,
+      "loss": 1.1546,
+      "step": 3093
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25038494984812143,
+      "learning_rate": 0.00019479245470768595,
+      "loss": 1.0509,
+      "step": 3094
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27600843103065575,
+      "learning_rate": 0.00019478741509320076,
+      "loss": 1.1192,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2736297017538167,
+      "learning_rate": 0.00019478237310660093,
+      "loss": 1.1505,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2545876917992688,
+      "learning_rate": 0.00019477732874801265,
+      "loss": 1.0992,
+      "step": 3097
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2539192301177348,
+      "learning_rate": 0.0001947722820175622,
+      "loss": 1.044,
+      "step": 3098
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25814263130227105,
+      "learning_rate": 0.00019476723291537575,
+      "loss": 1.1827,
+      "step": 3099
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2976356116278242,
+      "learning_rate": 0.0001947621814415798,
+      "loss": 0.9778,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2625055643669641,
+      "learning_rate": 0.00019475712759630068,
+      "loss": 1.0887,
+      "step": 3101
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26312176570889856,
+      "learning_rate": 0.00019475207137966487,
+      "loss": 1.0807,
+      "step": 3102
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2618322867889844,
+      "learning_rate": 0.00019474701279179895,
+      "loss": 1.2045,
+      "step": 3103
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2891550174279668,
+      "learning_rate": 0.00019474195183282947,
+      "loss": 1.0771,
+      "step": 3104
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3202983567379544,
+      "learning_rate": 0.00019473688850288312,
+      "loss": 1.1852,
+      "step": 3105
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25021772062444586,
+      "learning_rate": 0.0001947318228020866,
+      "loss": 1.1832,
+      "step": 3106
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2930599815597174,
+      "learning_rate": 0.00019472675473056666,
+      "loss": 1.0763,
+      "step": 3107
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30676853669698495,
+      "learning_rate": 0.00019472168428845014,
+      "loss": 1.0405,
+      "step": 3108
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2648409443563144,
+      "learning_rate": 0.00019471661147586395,
+      "loss": 1.1125,
+      "step": 3109
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26584314244912965,
+      "learning_rate": 0.00019471153629293503,
+      "loss": 1.0697,
+      "step": 3110
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37140190776674137,
+      "learning_rate": 0.0001947064587397904,
+      "loss": 1.0939,
+      "step": 3111
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.259017588096064,
+      "learning_rate": 0.00019470137881655712,
+      "loss": 1.0809,
+      "step": 3112
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26998271747435276,
+      "learning_rate": 0.00019469629652336232,
+      "loss": 1.0425,
+      "step": 3113
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25718343306878455,
+      "learning_rate": 0.0001946912118603332,
+      "loss": 1.1163,
+      "step": 3114
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2568678014483129,
+      "learning_rate": 0.00019468612482759695,
+      "loss": 0.9441,
+      "step": 3115
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2765085464241682,
+      "learning_rate": 0.00019468103542528094,
+      "loss": 1.0876,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23905715647271397,
+      "learning_rate": 0.0001946759436535125,
+      "loss": 1.1495,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23414722392686665,
+      "learning_rate": 0.00019467084951241907,
+      "loss": 1.0045,
+      "step": 3118
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2825675728276391,
+      "learning_rate": 0.00019466575300212816,
+      "loss": 1.0469,
+      "step": 3119
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27456256708811555,
+      "learning_rate": 0.00019466065412276727,
+      "loss": 0.952,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23902984443073147,
+      "learning_rate": 0.00019465555287446402,
+      "loss": 1.1261,
+      "step": 3121
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2860878827274362,
+      "learning_rate": 0.00019465044925734605,
+      "loss": 0.9592,
+      "step": 3122
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27511877405249713,
+      "learning_rate": 0.00019464534327154112,
+      "loss": 1.0913,
+      "step": 3123
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.25798267940905334,
+      "learning_rate": 0.000194640234917177,
+      "loss": 1.0157,
+      "step": 3124
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.284100415728136,
+      "learning_rate": 0.00019463512419438153,
+      "loss": 1.1027,
+      "step": 3125
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2382259658744238,
+      "learning_rate": 0.00019463001110328257,
+      "loss": 1.1828,
+      "step": 3126
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31222857704200785,
+      "learning_rate": 0.0001946248956440081,
+      "loss": 1.1124,
+      "step": 3127
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2625066692419136,
+      "learning_rate": 0.00019461977781668618,
+      "loss": 1.0737,
+      "step": 3128
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2793854323707662,
+      "learning_rate": 0.00019461465762144487,
+      "loss": 1.1363,
+      "step": 3129
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27108476180470265,
+      "learning_rate": 0.00019460953505841223,
+      "loss": 1.1485,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26903383341011894,
+      "learning_rate": 0.0001946044101277166,
+      "loss": 1.0214,
+      "step": 3131
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30317173630025673,
+      "learning_rate": 0.00019459928282948607,
+      "loss": 1.0941,
+      "step": 3132
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2833482336806812,
+      "learning_rate": 0.00019459415316384906,
+      "loss": 1.1549,
+      "step": 3133
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2652521067523786,
+      "learning_rate": 0.00019458902113093395,
+      "loss": 1.0997,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2796165666023849,
+      "learning_rate": 0.0001945838867308691,
+      "loss": 1.1974,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2795419726491581,
+      "learning_rate": 0.00019457874996378304,
+      "loss": 1.0421,
+      "step": 3136
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27461888378734867,
+      "learning_rate": 0.00019457361082980432,
+      "loss": 1.0375,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2751564840738016,
+      "learning_rate": 0.00019456846932906156,
+      "loss": 0.9755,
+      "step": 3138
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27520201040938214,
+      "learning_rate": 0.00019456332546168343,
+      "loss": 0.9982,
+      "step": 3139
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28743296092030146,
+      "learning_rate": 0.00019455817922779868,
+      "loss": 0.8786,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2791638788893473,
+      "learning_rate": 0.000194553030627536,
+      "loss": 1.1424,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2605528838298395,
+      "learning_rate": 0.00019454787966102435,
+      "loss": 1.0785,
+      "step": 3142
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2652776678480821,
+      "learning_rate": 0.00019454272632839255,
+      "loss": 1.0047,
+      "step": 3143
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2770395767937961,
+      "learning_rate": 0.00019453757062976964,
+      "loss": 1.1224,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2774545055078794,
+      "learning_rate": 0.00019453241256528462,
+      "loss": 1.0218,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29292827749120715,
+      "learning_rate": 0.00019452725213506654,
+      "loss": 1.1559,
+      "step": 3146
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2769766053696637,
+      "learning_rate": 0.00019452208933924459,
+      "loss": 1.0685,
+      "step": 3147
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.24168206829639038,
+      "learning_rate": 0.00019451692417794792,
+      "loss": 1.092,
+      "step": 3148
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3243097289376712,
+      "learning_rate": 0.00019451175665130584,
+      "loss": 1.1109,
+      "step": 3149
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2851118834219872,
+      "learning_rate": 0.00019450658675944764,
+      "loss": 1.0859,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26646697928523183,
+      "learning_rate": 0.00019450141450250272,
+      "loss": 1.092,
+      "step": 3151
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31149135380894666,
+      "learning_rate": 0.0001944962398806005,
+      "loss": 1.0533,
+      "step": 3152
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31641860365957436,
+      "learning_rate": 0.00019449106289387048,
+      "loss": 1.0906,
+      "step": 3153
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2861114243921771,
+      "learning_rate": 0.00019448588354244227,
+      "loss": 1.1436,
+      "step": 3154
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2678359187873902,
+      "learning_rate": 0.0001944807018264454,
+      "loss": 1.0751,
+      "step": 3155
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2795381844318716,
+      "learning_rate": 0.00019447551774600958,
+      "loss": 1.0243,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2646096339194622,
+      "learning_rate": 0.00019447033130126458,
+      "loss": 1.0279,
+      "step": 3157
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.295589829682441,
+      "learning_rate": 0.00019446514249234017,
+      "loss": 1.0735,
+      "step": 3158
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2980645819323912,
+      "learning_rate": 0.0001944599513193662,
+      "loss": 1.145,
+      "step": 3159
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3115102825868531,
+      "learning_rate": 0.00019445475778247256,
+      "loss": 1.1983,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2813320240238785,
+      "learning_rate": 0.00019444956188178927,
+      "loss": 1.1677,
+      "step": 3161
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.23562367846120347,
+      "learning_rate": 0.00019444436361744632,
+      "loss": 1.1973,
+      "step": 3162
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2928636477767414,
+      "learning_rate": 0.0001944391629895738,
+      "loss": 1.1904,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2938628446043182,
+      "learning_rate": 0.0001944339599983019,
+      "loss": 1.1338,
+      "step": 3164
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2698493025290344,
+      "learning_rate": 0.00019442875464376077,
+      "loss": 1.085,
+      "step": 3165
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30254516206760285,
+      "learning_rate": 0.00019442354692608075,
+      "loss": 1.21,
+      "step": 3166
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.26210762133120497,
+      "learning_rate": 0.0001944183368453921,
+      "loss": 1.0347,
+      "step": 3167
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2307520308777141,
+      "learning_rate": 0.00019441312440182524,
+      "loss": 1.0734,
+      "step": 3168
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.256456644213979,
+      "learning_rate": 0.0001944079095955106,
+      "loss": 1.1074,
+      "step": 3169
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27238535013463366,
+      "learning_rate": 0.00019440269242657868,
+      "loss": 1.1605,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27536976098604626,
+      "learning_rate": 0.00019439747289516009,
+      "loss": 1.2405,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2537359030538808,
+      "learning_rate": 0.00019439225100138536,
+      "loss": 1.0907,
+      "step": 3172
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27291761429286016,
+      "learning_rate": 0.00019438702674538525,
+      "loss": 1.0893,
+      "step": 3173
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2650287732828565,
+      "learning_rate": 0.00019438180012729047,
+      "loss": 1.0594,
+      "step": 3174
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2566176933125399,
+      "learning_rate": 0.00019437657114723184,
+      "loss": 1.0371,
+      "step": 3175
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.27936305636263997,
+      "learning_rate": 0.0001943713398053402,
+      "loss": 1.1312,
+      "step": 3176
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.29167361844632245,
+      "learning_rate": 0.00019436610610174646,
+      "loss": 1.116,
+      "step": 3177
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.28843098771577746,
+      "learning_rate": 0.00019436087003658163,
+      "loss": 1.0541,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.32516725974523086,
+      "learning_rate": 0.0001943556316099767,
+      "loss": 1.1834,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2541028278026109,
+      "learning_rate": 0.0001943503908220628,
+      "loss": 1.1216,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2834034653405795,
+      "learning_rate": 0.00019434514767297108,
+      "loss": 1.2544,
+      "step": 3181
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.30931710915213423,
+      "learning_rate": 0.00019433990216283274,
+      "loss": 1.0865,
+      "step": 3182
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2777726505717253,
+      "learning_rate": 0.00019433465429177904,
+      "loss": 1.0133,
+      "step": 3183
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2893309338829881,
+      "learning_rate": 0.00019432940405994135,
+      "loss": 1.1005,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2718000624288272,
+      "learning_rate": 0.00019432415146745103,
+      "loss": 0.9954,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2963971328585791,
+      "learning_rate": 0.00019431889651443953,
+      "loss": 1.0576,
+      "step": 3186
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.2767798358909039,
+      "learning_rate": 0.00019431363920103837,
+      "loss": 1.1268,
+      "step": 3187
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.31700853184697214,
+      "learning_rate": 0.00019430837952737914,
+      "loss": 1.061,
+      "step": 3188
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25763423615478015,
+      "learning_rate": 0.0001943031174935934,
+      "loss": 1.0927,
+      "step": 3189
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2610139408028157,
+      "learning_rate": 0.00019429785309981292,
+      "loss": 1.0666,
+      "step": 3190
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2884216831748455,
+      "learning_rate": 0.00019429258634616941,
+      "loss": 1.064,
+      "step": 3191
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2502879838444048,
+      "learning_rate": 0.00019428731723279463,
+      "loss": 1.1431,
+      "step": 3192
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25880669047725313,
+      "learning_rate": 0.0001942820457598205,
+      "loss": 1.0648,
+      "step": 3193
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.251763418883559,
+      "learning_rate": 0.0001942767719273789,
+      "loss": 1.095,
+      "step": 3194
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28454723290611905,
+      "learning_rate": 0.00019427149573560183,
+      "loss": 1.0639,
+      "step": 3195
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2740113209524548,
+      "learning_rate": 0.00019426621718462137,
+      "loss": 1.0383,
+      "step": 3196
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2894651427378075,
+      "learning_rate": 0.00019426093627456954,
+      "loss": 1.0393,
+      "step": 3197
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.272356851821048,
+      "learning_rate": 0.00019425565300557857,
+      "loss": 1.0492,
+      "step": 3198
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2834745514772499,
+      "learning_rate": 0.00019425036737778063,
+      "loss": 1.1115,
+      "step": 3199
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.290735855354118,
+      "learning_rate": 0.00019424507939130802,
+      "loss": 1.1519,
+      "step": 3200
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26309247215034526,
+      "learning_rate": 0.00019423978904629303,
+      "loss": 1.1589,
+      "step": 3201
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3034788416160877,
+      "learning_rate": 0.00019423449634286812,
+      "loss": 1.0927,
+      "step": 3202
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24953378202378695,
+      "learning_rate": 0.00019422920128116573,
+      "loss": 1.0734,
+      "step": 3203
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30391403166399206,
+      "learning_rate": 0.00019422390386131835,
+      "loss": 1.1223,
+      "step": 3204
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26768113646614333,
+      "learning_rate": 0.00019421860408345856,
+      "loss": 1.074,
+      "step": 3205
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26968925174933783,
+      "learning_rate": 0.000194213301947719,
+      "loss": 1.1126,
+      "step": 3206
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25338870451527706,
+      "learning_rate": 0.0001942079974542323,
+      "loss": 1.2239,
+      "step": 3207
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27878381417056575,
+      "learning_rate": 0.0001942026906031313,
+      "loss": 1.184,
+      "step": 3208
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25893716244148623,
+      "learning_rate": 0.00019419738139454874,
+      "loss": 1.1045,
+      "step": 3209
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2657479758737288,
+      "learning_rate": 0.0001941920698286175,
+      "loss": 1.0898,
+      "step": 3210
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.274528853814968,
+      "learning_rate": 0.00019418675590547054,
+      "loss": 1.2649,
+      "step": 3211
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28015720519786214,
+      "learning_rate": 0.00019418143962524084,
+      "loss": 1.1167,
+      "step": 3212
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26974290585106064,
+      "learning_rate": 0.00019417612098806137,
+      "loss": 1.0439,
+      "step": 3213
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26730225585717254,
+      "learning_rate": 0.00019417079999406532,
+      "loss": 1.1091,
+      "step": 3214
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24578940286887485,
+      "learning_rate": 0.0001941654766433858,
+      "loss": 1.1115,
+      "step": 3215
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24916795650749093,
+      "learning_rate": 0.00019416015093615604,
+      "loss": 0.9763,
+      "step": 3216
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3032785764886797,
+      "learning_rate": 0.00019415482287250935,
+      "loss": 1.0748,
+      "step": 3217
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2840877489578721,
+      "learning_rate": 0.00019414949245257903,
+      "loss": 1.0943,
+      "step": 3218
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2721658275584308,
+      "learning_rate": 0.0001941441596764985,
+      "loss": 1.0798,
+      "step": 3219
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27711342122729105,
+      "learning_rate": 0.00019413882454440118,
+      "loss": 1.0857,
+      "step": 3220
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.264240156198682,
+      "learning_rate": 0.00019413348705642065,
+      "loss": 1.1476,
+      "step": 3221
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2672728525407875,
+      "learning_rate": 0.00019412814721269042,
+      "loss": 1.1006,
+      "step": 3222
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2790227898377365,
+      "learning_rate": 0.00019412280501334418,
+      "loss": 1.0214,
+      "step": 3223
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26686766116618904,
+      "learning_rate": 0.00019411746045851553,
+      "loss": 1.0939,
+      "step": 3224
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26689843862224927,
+      "learning_rate": 0.00019411211354833832,
+      "loss": 1.1118,
+      "step": 3225
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2645315881506633,
+      "learning_rate": 0.00019410676428294633,
+      "loss": 1.064,
+      "step": 3226
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27750130260616634,
+      "learning_rate": 0.00019410141266247338,
+      "loss": 1.1626,
+      "step": 3227
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2739213945232073,
+      "learning_rate": 0.0001940960586870535,
+      "loss": 1.0854,
+      "step": 3228
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28287550537217365,
+      "learning_rate": 0.00019409070235682055,
+      "loss": 1.0474,
+      "step": 3229
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2731751302387,
+      "learning_rate": 0.0001940853436719087,
+      "loss": 0.9977,
+      "step": 3230
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2950771929234802,
+      "learning_rate": 0.00019407998263245194,
+      "loss": 1.1031,
+      "step": 3231
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.274205168506372,
+      "learning_rate": 0.0001940746192385845,
+      "loss": 1.0148,
+      "step": 3232
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24694847878806808,
+      "learning_rate": 0.0001940692534904406,
+      "loss": 1.0189,
+      "step": 3233
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2884458695976643,
+      "learning_rate": 0.00019406388538815454,
+      "loss": 1.0534,
+      "step": 3234
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25614325988090403,
+      "learning_rate": 0.0001940585149318606,
+      "loss": 1.1938,
+      "step": 3235
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28634252057649984,
+      "learning_rate": 0.0001940531421216932,
+      "loss": 1.1215,
+      "step": 3236
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26353357452427006,
+      "learning_rate": 0.00019404776695778684,
+      "loss": 1.0671,
+      "step": 3237
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.252301189841136,
+      "learning_rate": 0.00019404238944027596,
+      "loss": 1.0318,
+      "step": 3238
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2921004823006027,
+      "learning_rate": 0.0001940370095692952,
+      "loss": 1.1743,
+      "step": 3239
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3010241530589875,
+      "learning_rate": 0.0001940316273449792,
+      "loss": 1.1167,
+      "step": 3240
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26470231728020255,
+      "learning_rate": 0.00019402624276746263,
+      "loss": 1.1322,
+      "step": 3241
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26518892722531195,
+      "learning_rate": 0.00019402085583688022,
+      "loss": 1.043,
+      "step": 3242
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28690580948021155,
+      "learning_rate": 0.0001940154665533668,
+      "loss": 1.0667,
+      "step": 3243
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2810395515843624,
+      "learning_rate": 0.00019401007491705725,
+      "loss": 0.9801,
+      "step": 3244
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2572499963401515,
+      "learning_rate": 0.00019400468092808647,
+      "loss": 1.159,
+      "step": 3245
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2644492737063085,
+      "learning_rate": 0.00019399928458658952,
+      "loss": 1.119,
+      "step": 3246
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2693774572143553,
+      "learning_rate": 0.00019399388589270134,
+      "loss": 1.1763,
+      "step": 3247
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3074129041443745,
+      "learning_rate": 0.00019398848484655714,
+      "loss": 1.1109,
+      "step": 3248
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.22885608780661518,
+      "learning_rate": 0.00019398308144829202,
+      "loss": 1.1484,
+      "step": 3249
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2431406554688736,
+      "learning_rate": 0.0001939776756980412,
+      "loss": 1.1024,
+      "step": 3250
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2508813769274511,
+      "learning_rate": 0.00019397226759594003,
+      "loss": 1.1161,
+      "step": 3251
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.259498174018731,
+      "learning_rate": 0.00019396685714212378,
+      "loss": 1.1121,
+      "step": 3252
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25690080144023086,
+      "learning_rate": 0.00019396144433672787,
+      "loss": 1.1951,
+      "step": 3253
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2703955152553523,
+      "learning_rate": 0.00019395602917988774,
+      "loss": 1.0848,
+      "step": 3254
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3248488191794343,
+      "learning_rate": 0.00019395061167173895,
+      "loss": 1.1507,
+      "step": 3255
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28991024777070024,
+      "learning_rate": 0.00019394519181241705,
+      "loss": 1.1015,
+      "step": 3256
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2764459982278797,
+      "learning_rate": 0.00019393976960205772,
+      "loss": 0.9972,
+      "step": 3257
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26862234853504874,
+      "learning_rate": 0.00019393434504079657,
+      "loss": 1.1453,
+      "step": 3258
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27971969886015396,
+      "learning_rate": 0.00019392891812876944,
+      "loss": 1.0825,
+      "step": 3259
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2770468588755897,
+      "learning_rate": 0.00019392348886611207,
+      "loss": 1.0536,
+      "step": 3260
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.271609510880968,
+      "learning_rate": 0.00019391805725296038,
+      "loss": 1.0481,
+      "step": 3261
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24387465904724714,
+      "learning_rate": 0.00019391262328945027,
+      "loss": 1.0953,
+      "step": 3262
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.306543464049848,
+      "learning_rate": 0.00019390718697571776,
+      "loss": 1.1486,
+      "step": 3263
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30439534530685314,
+      "learning_rate": 0.00019390174831189887,
+      "loss": 1.0251,
+      "step": 3264
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2702567545507524,
+      "learning_rate": 0.0001938963072981297,
+      "loss": 1.0219,
+      "step": 3265
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24990979733983004,
+      "learning_rate": 0.00019389086393454644,
+      "loss": 0.9841,
+      "step": 3266
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.30401365328102387,
+      "learning_rate": 0.0001938854182212853,
+      "loss": 1.0608,
+      "step": 3267
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3051517353842027,
+      "learning_rate": 0.00019387997015848254,
+      "loss": 1.0624,
+      "step": 3268
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28665306616118164,
+      "learning_rate": 0.00019387451974627455,
+      "loss": 1.0742,
+      "step": 3269
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3231516980435964,
+      "learning_rate": 0.0001938690669847977,
+      "loss": 1.1872,
+      "step": 3270
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2956462625808931,
+      "learning_rate": 0.00019386361187418848,
+      "loss": 1.1729,
+      "step": 3271
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2657542153528369,
+      "learning_rate": 0.00019385815441458335,
+      "loss": 1.0359,
+      "step": 3272
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3012539639681833,
+      "learning_rate": 0.0001938526946061189,
+      "loss": 1.224,
+      "step": 3273
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.23858077659378998,
+      "learning_rate": 0.00019384723244893182,
+      "loss": 0.9866,
+      "step": 3274
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.28237967640012435,
+      "learning_rate": 0.00019384176794315876,
+      "loss": 1.1095,
+      "step": 3275
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2599923150922761,
+      "learning_rate": 0.0001938363010889365,
+      "loss": 1.1204,
+      "step": 3276
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2869091154710219,
+      "learning_rate": 0.00019383083188640178,
+      "loss": 1.0668,
+      "step": 3277
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.31710794291071975,
+      "learning_rate": 0.00019382536033569155,
+      "loss": 1.0612,
+      "step": 3278
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.29559893780311114,
+      "learning_rate": 0.0001938198864369427,
+      "loss": 1.077,
+      "step": 3279
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2898590799989514,
+      "learning_rate": 0.00019381441019029224,
+      "loss": 0.9822,
+      "step": 3280
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2653384099576511,
+      "learning_rate": 0.00019380893159587722,
+      "loss": 1.0328,
+      "step": 3281
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3059551368813422,
+      "learning_rate": 0.00019380345065383468,
+      "loss": 1.1349,
+      "step": 3282
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.24886086837418994,
+      "learning_rate": 0.0001937979673643019,
+      "loss": 1.068,
+      "step": 3283
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.23919413973427292,
+      "learning_rate": 0.000193792481727416,
+      "loss": 1.1181,
+      "step": 3284
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.25404095588714376,
+      "learning_rate": 0.0001937869937433143,
+      "loss": 1.045,
+      "step": 3285
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.223637661799907,
+      "learning_rate": 0.00019378150341213416,
+      "loss": 1.0104,
+      "step": 3286
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2779252243873626,
+      "learning_rate": 0.00019377601073401293,
+      "loss": 1.1353,
+      "step": 3287
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.27508596502440386,
+      "learning_rate": 0.0001937705157090881,
+      "loss": 1.0028,
+      "step": 3288
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2829961966551457,
+      "learning_rate": 0.0001937650183374972,
+      "loss": 1.1252,
+      "step": 3289
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2547718896278451,
+      "learning_rate": 0.00019375951861937775,
+      "loss": 1.1001,
+      "step": 3290
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.2386176095295127,
+      "learning_rate": 0.00019375401655486745,
+      "loss": 0.9711,
+      "step": 3291
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.26234578344158516,
+      "learning_rate": 0.00019374851214410397,
+      "loss": 1.0339,
+      "step": 3292
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2919799926949464,
+      "learning_rate": 0.00019374300538722503,
+      "loss": 1.0436,
+      "step": 3293
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26522986184686614,
+      "learning_rate": 0.00019373749628436848,
+      "loss": 1.1032,
+      "step": 3294
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27747724141686647,
+      "learning_rate": 0.00019373198483567215,
+      "loss": 1.1199,
+      "step": 3295
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.251558053545879,
+      "learning_rate": 0.00019372647104127401,
+      "loss": 1.0908,
+      "step": 3296
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2866203569622034,
+      "learning_rate": 0.00019372095490131206,
+      "loss": 1.1136,
+      "step": 3297
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26519901166294035,
+      "learning_rate": 0.00019371543641592427,
+      "loss": 1.0744,
+      "step": 3298
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2670922090903857,
+      "learning_rate": 0.0001937099155852488,
+      "loss": 1.0004,
+      "step": 3299
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2582252671213613,
+      "learning_rate": 0.0001937043924094238,
+      "loss": 1.0158,
+      "step": 3300
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2780671456060699,
+      "learning_rate": 0.00019369886688858746,
+      "loss": 1.0033,
+      "step": 3301
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26115496930707605,
+      "learning_rate": 0.00019369333902287812,
+      "loss": 1.1021,
+      "step": 3302
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2806226279994436,
+      "learning_rate": 0.00019368780881243408,
+      "loss": 1.0112,
+      "step": 3303
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2544906876251208,
+      "learning_rate": 0.00019368227625739376,
+      "loss": 1.1054,
+      "step": 3304
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2856721128498456,
+      "learning_rate": 0.00019367674135789559,
+      "loss": 1.1403,
+      "step": 3305
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28275774405141135,
+      "learning_rate": 0.00019367120411407807,
+      "loss": 1.1926,
+      "step": 3306
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30945016209372866,
+      "learning_rate": 0.00019366566452607984,
+      "loss": 1.0632,
+      "step": 3307
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28757906013236484,
+      "learning_rate": 0.00019366012259403945,
+      "loss": 1.0334,
+      "step": 3308
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28623289799348706,
+      "learning_rate": 0.00019365457831809564,
+      "loss": 1.0065,
+      "step": 3309
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31283779658466215,
+      "learning_rate": 0.00019364903169838714,
+      "loss": 1.1444,
+      "step": 3310
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25353872879290523,
+      "learning_rate": 0.0001936434827350528,
+      "loss": 1.1551,
+      "step": 3311
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28420385173740026,
+      "learning_rate": 0.00019363793142823142,
+      "loss": 1.0866,
+      "step": 3312
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2756468967799023,
+      "learning_rate": 0.00019363237777806193,
+      "loss": 1.1621,
+      "step": 3313
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2725051208626741,
+      "learning_rate": 0.0001936268217846834,
+      "loss": 1.0359,
+      "step": 3314
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2785478077945499,
+      "learning_rate": 0.0001936212634482348,
+      "loss": 1.1558,
+      "step": 3315
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2506394562927538,
+      "learning_rate": 0.00019361570276885522,
+      "loss": 1.1897,
+      "step": 3316
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2606058706252523,
+      "learning_rate": 0.00019361013974668385,
+      "loss": 1.1776,
+      "step": 3317
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2496048710743572,
+      "learning_rate": 0.0001936045743818599,
+      "loss": 0.9721,
+      "step": 3318
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26993436748357164,
+      "learning_rate": 0.00019359900667452264,
+      "loss": 1.104,
+      "step": 3319
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.255141548047229,
+      "learning_rate": 0.0001935934366248114,
+      "loss": 1.0802,
+      "step": 3320
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27647316322378807,
+      "learning_rate": 0.00019358786423286564,
+      "loss": 1.0241,
+      "step": 3321
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2923601566113196,
+      "learning_rate": 0.00019358228949882474,
+      "loss": 1.0406,
+      "step": 3322
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24263042595969897,
+      "learning_rate": 0.00019357671242282821,
+      "loss": 0.9864,
+      "step": 3323
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27774846850692353,
+      "learning_rate": 0.00019357113300501566,
+      "loss": 1.0937,
+      "step": 3324
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27923854319931557,
+      "learning_rate": 0.0001935655512455267,
+      "loss": 1.0601,
+      "step": 3325
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3061443475173794,
+      "learning_rate": 0.000193559967144501,
+      "loss": 1.2078,
+      "step": 3326
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2846307312850083,
+      "learning_rate": 0.00019355438070207834,
+      "loss": 1.0301,
+      "step": 3327
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27792518753199547,
+      "learning_rate": 0.0001935487919183985,
+      "loss": 1.0656,
+      "step": 3328
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2845450991292036,
+      "learning_rate": 0.00019354320079360132,
+      "loss": 1.0507,
+      "step": 3329
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.31509060671206557,
+      "learning_rate": 0.0001935376073278268,
+      "loss": 1.1854,
+      "step": 3330
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27538788262531033,
+      "learning_rate": 0.00019353201152121484,
+      "loss": 1.0761,
+      "step": 3331
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3085113970962017,
+      "learning_rate": 0.00019352641337390552,
+      "loss": 1.0149,
+      "step": 3332
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2649187417930221,
+      "learning_rate": 0.00019352081288603895,
+      "loss": 0.9846,
+      "step": 3333
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2702951480669255,
+      "learning_rate": 0.0001935152100577552,
+      "loss": 1.1289,
+      "step": 3334
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2587123130961769,
+      "learning_rate": 0.00019350960488919458,
+      "loss": 1.1603,
+      "step": 3335
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26093692625675025,
+      "learning_rate": 0.00019350399738049735,
+      "loss": 1.189,
+      "step": 3336
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2571900526712086,
+      "learning_rate": 0.0001934983875318038,
+      "loss": 1.1219,
+      "step": 3337
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2823069020729405,
+      "learning_rate": 0.0001934927753432543,
+      "loss": 1.0871,
+      "step": 3338
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29397830877053815,
+      "learning_rate": 0.00019348716081498942,
+      "loss": 1.0944,
+      "step": 3339
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29592501153858697,
+      "learning_rate": 0.00019348154394714952,
+      "loss": 1.0847,
+      "step": 3340
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27462052813299775,
+      "learning_rate": 0.00019347592473987528,
+      "loss": 1.1752,
+      "step": 3341
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24616995149088777,
+      "learning_rate": 0.00019347030319330727,
+      "loss": 1.1025,
+      "step": 3342
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27129435980716926,
+      "learning_rate": 0.00019346467930758614,
+      "loss": 1.1367,
+      "step": 3343
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27336188709510006,
+      "learning_rate": 0.0001934590530828527,
+      "loss": 1.1193,
+      "step": 3344
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27826604163520996,
+      "learning_rate": 0.0001934534245192477,
+      "loss": 1.1751,
+      "step": 3345
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2998027219687268,
+      "learning_rate": 0.00019344779361691203,
+      "loss": 1.1368,
+      "step": 3346
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28056799541807836,
+      "learning_rate": 0.0001934421603759866,
+      "loss": 1.0334,
+      "step": 3347
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24723826046732345,
+      "learning_rate": 0.00019343652479661237,
+      "loss": 1.0881,
+      "step": 3348
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3324987905537479,
+      "learning_rate": 0.00019343088687893037,
+      "loss": 1.0454,
+      "step": 3349
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2612587061661008,
+      "learning_rate": 0.00019342524662308174,
+      "loss": 1.1285,
+      "step": 3350
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3175396328178754,
+      "learning_rate": 0.0001934196040292076,
+      "loss": 1.1002,
+      "step": 3351
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2982419959527945,
+      "learning_rate": 0.00019341395909744914,
+      "loss": 1.1169,
+      "step": 3352
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2665630466583381,
+      "learning_rate": 0.00019340831182794763,
+      "loss": 1.0946,
+      "step": 3353
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2747715419393994,
+      "learning_rate": 0.00019340266222084445,
+      "loss": 1.0806,
+      "step": 3354
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26678051390620816,
+      "learning_rate": 0.00019339701027628093,
+      "loss": 1.0541,
+      "step": 3355
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2598902398444097,
+      "learning_rate": 0.00019339135599439852,
+      "loss": 1.0699,
+      "step": 3356
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.254933790104563,
+      "learning_rate": 0.00019338569937533872,
+      "loss": 1.174,
+      "step": 3357
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3128276440616595,
+      "learning_rate": 0.00019338004041924314,
+      "loss": 1.1016,
+      "step": 3358
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.296430797427781,
+      "learning_rate": 0.00019337437912625332,
+      "loss": 1.108,
+      "step": 3359
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2970936872836418,
+      "learning_rate": 0.00019336871549651102,
+      "loss": 1.0936,
+      "step": 3360
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.30049312775342935,
+      "learning_rate": 0.0001933630495301579,
+      "loss": 1.1077,
+      "step": 3361
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25869029149150324,
+      "learning_rate": 0.0001933573812273358,
+      "loss": 1.1524,
+      "step": 3362
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.25180878439069226,
+      "learning_rate": 0.00019335171058818657,
+      "loss": 0.9934,
+      "step": 3363
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2582499973976482,
+      "learning_rate": 0.0001933460376128521,
+      "loss": 1.0993,
+      "step": 3364
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27195751975656457,
+      "learning_rate": 0.0001933403623014744,
+      "loss": 1.0789,
+      "step": 3365
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2725473211575861,
+      "learning_rate": 0.00019333468465419545,
+      "loss": 1.1782,
+      "step": 3366
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2696499000977069,
+      "learning_rate": 0.00019332900467115735,
+      "loss": 1.079,
+      "step": 3367
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26468748026222183,
+      "learning_rate": 0.00019332332235250227,
+      "loss": 1.0032,
+      "step": 3368
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27136388855246457,
+      "learning_rate": 0.00019331763769837239,
+      "loss": 1.1859,
+      "step": 3369
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3147503114706967,
+      "learning_rate": 0.00019331195070890997,
+      "loss": 0.9721,
+      "step": 3370
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.26623718570346866,
+      "learning_rate": 0.00019330626138425733,
+      "loss": 1.1342,
+      "step": 3371
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.238352475784514,
+      "learning_rate": 0.0001933005697245569,
+      "loss": 1.0439,
+      "step": 3372
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28164020535606377,
+      "learning_rate": 0.00019329487572995104,
+      "loss": 1.0531,
+      "step": 3373
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2798281524042237,
+      "learning_rate": 0.0001932891794005823,
+      "loss": 1.0947,
+      "step": 3374
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2593240400068639,
+      "learning_rate": 0.0001932834807365932,
+      "loss": 1.3053,
+      "step": 3375
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2956714093140169,
+      "learning_rate": 0.0001932777797381264,
+      "loss": 1.1146,
+      "step": 3376
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24933340729423376,
+      "learning_rate": 0.0001932720764053245,
+      "loss": 1.0169,
+      "step": 3377
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2746804152121392,
+      "learning_rate": 0.0001932663707383303,
+      "loss": 1.2153,
+      "step": 3378
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2438039762627365,
+      "learning_rate": 0.00019326066273728652,
+      "loss": 1.1174,
+      "step": 3379
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2667417249557865,
+      "learning_rate": 0.00019325495240233608,
+      "loss": 1.1442,
+      "step": 3380
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.282105093224216,
+      "learning_rate": 0.0001932492397336218,
+      "loss": 1.0328,
+      "step": 3381
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2700271902826537,
+      "learning_rate": 0.00019324352473128675,
+      "loss": 1.1098,
+      "step": 3382
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2810010796311137,
+      "learning_rate": 0.00019323780739547382,
+      "loss": 1.1228,
+      "step": 3383
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27543586143634324,
+      "learning_rate": 0.00019323208772632623,
+      "loss": 1.0899,
+      "step": 3384
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28148476530118366,
+      "learning_rate": 0.00019322636572398705,
+      "loss": 1.0954,
+      "step": 3385
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24691162930671992,
+      "learning_rate": 0.00019322064138859943,
+      "loss": 1.0638,
+      "step": 3386
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2766771037796505,
+      "learning_rate": 0.0001932149147203067,
+      "loss": 1.0988,
+      "step": 3387
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.27879746565463825,
+      "learning_rate": 0.00019320918571925214,
+      "loss": 1.1285,
+      "step": 3388
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.22878400180565586,
+      "learning_rate": 0.00019320345438557913,
+      "loss": 1.0721,
+      "step": 3389
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.28209510083017475,
+      "learning_rate": 0.0001931977207194311,
+      "loss": 1.0092,
+      "step": 3390
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2718261063085623,
+      "learning_rate": 0.00019319198472095154,
+      "loss": 1.118,
+      "step": 3391
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29842791503503757,
+      "learning_rate": 0.00019318624639028397,
+      "loss": 1.0923,
+      "step": 3392
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24518361819708703,
+      "learning_rate": 0.00019318050572757206,
+      "loss": 1.0292,
+      "step": 3393
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.24792628965630142,
+      "learning_rate": 0.00019317476273295937,
+      "loss": 1.0833,
+      "step": 3394
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.29078999381758364,
+      "learning_rate": 0.00019316901740658974,
+      "loss": 1.1031,
+      "step": 3395
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2567096843477097,
+      "learning_rate": 0.00019316326974860688,
+      "loss": 1.0013,
+      "step": 3396
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.2554741348697831,
+      "learning_rate": 0.00019315751975915464,
+      "loss": 1.0173,
+      "step": 3397
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28242380704250564,
+      "learning_rate": 0.00019315176743837692,
+      "loss": 1.1378,
+      "step": 3398
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31001384373991414,
+      "learning_rate": 0.00019314601278641767,
+      "loss": 1.1999,
+      "step": 3399
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2833937019206038,
+      "learning_rate": 0.0001931402558034209,
+      "loss": 1.1491,
+      "step": 3400
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3026457744535161,
+      "learning_rate": 0.00019313449648953075,
+      "loss": 1.2268,
+      "step": 3401
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26414264616917044,
+      "learning_rate": 0.00019312873484489122,
+      "loss": 1.0955,
+      "step": 3402
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24532958588409323,
+      "learning_rate": 0.0001931229708696466,
+      "loss": 1.0478,
+      "step": 3403
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30331695478379483,
+      "learning_rate": 0.00019311720456394115,
+      "loss": 1.0953,
+      "step": 3404
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28074250597379036,
+      "learning_rate": 0.00019311143592791908,
+      "loss": 1.1166,
+      "step": 3405
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2409279290445585,
+      "learning_rate": 0.00019310566496172482,
+      "loss": 1.1022,
+      "step": 3406
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27337622876374895,
+      "learning_rate": 0.00019309989166550276,
+      "loss": 1.0369,
+      "step": 3407
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27239368467190694,
+      "learning_rate": 0.00019309411603939746,
+      "loss": 1.0825,
+      "step": 3408
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2510203565258273,
+      "learning_rate": 0.00019308833808355335,
+      "loss": 1.1297,
+      "step": 3409
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3019608047256465,
+      "learning_rate": 0.0001930825577981151,
+      "loss": 1.0537,
+      "step": 3410
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27782305027611853,
+      "learning_rate": 0.00019307677518322732,
+      "loss": 1.0568,
+      "step": 3411
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27292941976306373,
+      "learning_rate": 0.00019307099023903475,
+      "loss": 1.1049,
+      "step": 3412
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31644950630512886,
+      "learning_rate": 0.00019306520296568213,
+      "loss": 0.9,
+      "step": 3413
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25546779805105374,
+      "learning_rate": 0.00019305941336331437,
+      "loss": 1.1242,
+      "step": 3414
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2792316426100012,
+      "learning_rate": 0.00019305362143207629,
+      "loss": 1.0101,
+      "step": 3415
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28856119813600223,
+      "learning_rate": 0.00019304782717211282,
+      "loss": 1.0683,
+      "step": 3416
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2538864469655074,
+      "learning_rate": 0.00019304203058356903,
+      "loss": 1.0736,
+      "step": 3417
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27697511534173397,
+      "learning_rate": 0.00019303623166658994,
+      "loss": 1.0237,
+      "step": 3418
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2816777125591762,
+      "learning_rate": 0.00019303043042132067,
+      "loss": 0.9735,
+      "step": 3419
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.270336650155053,
+      "learning_rate": 0.00019302462684790643,
+      "loss": 1.0452,
+      "step": 3420
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3171677773770914,
+      "learning_rate": 0.00019301882094649244,
+      "loss": 1.0691,
+      "step": 3421
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30382729652830914,
+      "learning_rate": 0.00019301301271722397,
+      "loss": 1.1365,
+      "step": 3422
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2811965071217192,
+      "learning_rate": 0.00019300720216024642,
+      "loss": 1.0391,
+      "step": 3423
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2525300639055227,
+      "learning_rate": 0.00019300138927570517,
+      "loss": 1.1168,
+      "step": 3424
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2566849695783432,
+      "learning_rate": 0.00019299557406374574,
+      "loss": 1.0444,
+      "step": 3425
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29423746734781697,
+      "learning_rate": 0.00019298975652451357,
+      "loss": 1.1018,
+      "step": 3426
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27988383125452326,
+      "learning_rate": 0.00019298393665815434,
+      "loss": 1.0843,
+      "step": 3427
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2914047802879343,
+      "learning_rate": 0.00019297811446481364,
+      "loss": 1.1238,
+      "step": 3428
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26554271237108157,
+      "learning_rate": 0.0001929722899446372,
+      "loss": 1.0654,
+      "step": 3429
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2626759072606683,
+      "learning_rate": 0.00019296646309777078,
+      "loss": 1.0299,
+      "step": 3430
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24310810286273438,
+      "learning_rate": 0.00019296063392436016,
+      "loss": 1.1926,
+      "step": 3431
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2681389299412346,
+      "learning_rate": 0.0001929548024245513,
+      "loss": 1.0535,
+      "step": 3432
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2705554100674985,
+      "learning_rate": 0.00019294896859849007,
+      "loss": 1.1006,
+      "step": 3433
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29956584728704405,
+      "learning_rate": 0.00019294313244632246,
+      "loss": 1.1618,
+      "step": 3434
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25979972995620976,
+      "learning_rate": 0.00019293729396819455,
+      "loss": 1.1341,
+      "step": 3435
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.31840805453945864,
+      "learning_rate": 0.0001929314531642525,
+      "loss": 1.0505,
+      "step": 3436
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.30601505649557453,
+      "learning_rate": 0.0001929256100346424,
+      "loss": 1.1247,
+      "step": 3437
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.255162069191932,
+      "learning_rate": 0.0001929197645795105,
+      "loss": 1.0914,
+      "step": 3438
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27220081852340894,
+      "learning_rate": 0.00019291391679900308,
+      "loss": 1.1204,
+      "step": 3439
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2680103064629152,
+      "learning_rate": 0.00019290806669326651,
+      "loss": 1.1374,
+      "step": 3440
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26907374179829274,
+      "learning_rate": 0.0001929022142624472,
+      "loss": 1.104,
+      "step": 3441
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25948281052343475,
+      "learning_rate": 0.00019289635950669158,
+      "loss": 1.1315,
+      "step": 3442
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25383699970065504,
+      "learning_rate": 0.00019289050242614616,
+      "loss": 1.0563,
+      "step": 3443
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28630705685350616,
+      "learning_rate": 0.00019288464302095757,
+      "loss": 1.0699,
+      "step": 3444
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2622608246535195,
+      "learning_rate": 0.00019287878129127238,
+      "loss": 1.0172,
+      "step": 3445
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27001089784818566,
+      "learning_rate": 0.00019287291723723735,
+      "loss": 0.9808,
+      "step": 3446
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3040076339208099,
+      "learning_rate": 0.00019286705085899916,
+      "loss": 1.2147,
+      "step": 3447
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2996187859781569,
+      "learning_rate": 0.00019286118215670471,
+      "loss": 1.1517,
+      "step": 3448
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.33317490146125284,
+      "learning_rate": 0.00019285531113050075,
+      "loss": 1.0603,
+      "step": 3449
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2900875599565296,
+      "learning_rate": 0.00019284943778053433,
+      "loss": 1.1299,
+      "step": 3450
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28251347088813633,
+      "learning_rate": 0.00019284356210695234,
+      "loss": 1.1064,
+      "step": 3451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28341004906992046,
+      "learning_rate": 0.00019283768410990185,
+      "loss": 1.0892,
+      "step": 3452
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2883977918976394,
+      "learning_rate": 0.00019283180378953,
+      "loss": 1.1733,
+      "step": 3453
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.260390677587717,
+      "learning_rate": 0.0001928259211459839,
+      "loss": 0.9542,
+      "step": 3454
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29136611867774026,
+      "learning_rate": 0.0001928200361794108,
+      "loss": 1.0044,
+      "step": 3455
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26446944983636694,
+      "learning_rate": 0.00019281414888995795,
+      "loss": 1.083,
+      "step": 3456
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2483864576667101,
+      "learning_rate": 0.0001928082592777727,
+      "loss": 1.0629,
+      "step": 3457
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24787693382866996,
+      "learning_rate": 0.00019280236734300243,
+      "loss": 1.0367,
+      "step": 3458
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3038463165988747,
+      "learning_rate": 0.00019279647308579457,
+      "loss": 1.0523,
+      "step": 3459
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.278602231033513,
+      "learning_rate": 0.00019279057650629667,
+      "loss": 1.168,
+      "step": 3460
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2488098065303038,
+      "learning_rate": 0.0001927846776046563,
+      "loss": 1.0209,
+      "step": 3461
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2758374803041767,
+      "learning_rate": 0.00019277877638102103,
+      "loss": 1.2443,
+      "step": 3462
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26309237485251713,
+      "learning_rate": 0.00019277287283553856,
+      "loss": 0.9914,
+      "step": 3463
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.23795915412482324,
+      "learning_rate": 0.00019276696696835668,
+      "loss": 1.1086,
+      "step": 3464
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3224730190795291,
+      "learning_rate": 0.0001927610587796231,
+      "loss": 1.2036,
+      "step": 3465
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.282804393013691,
+      "learning_rate": 0.00019275514826948577,
+      "loss": 1.137,
+      "step": 3466
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26066639065374003,
+      "learning_rate": 0.00019274923543809253,
+      "loss": 1.127,
+      "step": 3467
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24788428251360412,
+      "learning_rate": 0.00019274332028559142,
+      "loss": 1.0519,
+      "step": 3468
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.290526276121983,
+      "learning_rate": 0.0001927374028121304,
+      "loss": 1.1116,
+      "step": 3469
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2673170033387275,
+      "learning_rate": 0.00019273148301785759,
+      "loss": 1.1649,
+      "step": 3470
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.264616160691623,
+      "learning_rate": 0.00019272556090292115,
+      "loss": 1.1035,
+      "step": 3471
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.22386779170530668,
+      "learning_rate": 0.00019271963646746927,
+      "loss": 1.0443,
+      "step": 3472
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29588656285029913,
+      "learning_rate": 0.00019271370971165022,
+      "loss": 1.1571,
+      "step": 3473
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26869130366990424,
+      "learning_rate": 0.00019270778063561233,
+      "loss": 0.8483,
+      "step": 3474
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2442917412236581,
+      "learning_rate": 0.00019270184923950395,
+      "loss": 1.1119,
+      "step": 3475
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27434823182121076,
+      "learning_rate": 0.00019269591552347352,
+      "loss": 1.0498,
+      "step": 3476
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26614974020743387,
+      "learning_rate": 0.00019268997948766956,
+      "loss": 1.0791,
+      "step": 3477
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2959411757891701,
+      "learning_rate": 0.00019268404113224059,
+      "loss": 1.1704,
+      "step": 3478
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26576214735143877,
+      "learning_rate": 0.00019267810045733527,
+      "loss": 1.0451,
+      "step": 3479
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26797549883523514,
+      "learning_rate": 0.00019267215746310222,
+      "loss": 1.1066,
+      "step": 3480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26415504530857153,
+      "learning_rate": 0.0001926662121496902,
+      "loss": 0.9906,
+      "step": 3481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2811191242637275,
+      "learning_rate": 0.000192660264517248,
+      "loss": 1.2062,
+      "step": 3482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26967275386109973,
+      "learning_rate": 0.0001926543145659244,
+      "loss": 1.1001,
+      "step": 3483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24658725442432267,
+      "learning_rate": 0.00019264836229586837,
+      "loss": 1.0698,
+      "step": 3484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2716734887510684,
+      "learning_rate": 0.00019264240770722885,
+      "loss": 1.0129,
+      "step": 3485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.24634587706914243,
+      "learning_rate": 0.00019263645080015485,
+      "loss": 0.9294,
+      "step": 3486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26665979653082733,
+      "learning_rate": 0.00019263049157479544,
+      "loss": 1.0361,
+      "step": 3487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.262098888665953,
+      "learning_rate": 0.0001926245300312998,
+      "loss": 1.0617,
+      "step": 3488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.27380325184122906,
+      "learning_rate": 0.00019261856616981703,
+      "loss": 1.0235,
+      "step": 3489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25372809019534937,
+      "learning_rate": 0.00019261259999049646,
+      "loss": 1.0733,
+      "step": 3490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2703271495420422,
+      "learning_rate": 0.00019260663149348736,
+      "loss": 1.1621,
+      "step": 3491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.285631485877341,
+      "learning_rate": 0.00019260066067893915,
+      "loss": 1.119,
+      "step": 3492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28067090620266755,
+      "learning_rate": 0.00019259468754700114,
+      "loss": 1.1283,
+      "step": 3493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.25704313202950074,
+      "learning_rate": 0.00019258871209782292,
+      "loss": 1.1286,
+      "step": 3494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.29208835332994176,
+      "learning_rate": 0.00019258273433155399,
+      "loss": 1.1172,
+      "step": 3495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26926728161587604,
+      "learning_rate": 0.00019257675424834395,
+      "loss": 1.0684,
+      "step": 3496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.2732931154751818,
+      "learning_rate": 0.00019257077184834244,
+      "loss": 1.1041,
+      "step": 3497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26420765126059936,
+      "learning_rate": 0.00019256478713169917,
+      "loss": 1.0886,
+      "step": 3498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26082579867661154,
+      "learning_rate": 0.00019255880009856396,
+      "loss": 1.0275,
+      "step": 3499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.28137431081329883,
+      "learning_rate": 0.0001925528107490866,
+      "loss": 1.0764,
+      "step": 3500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.26907207440559944,
+      "learning_rate": 0.00019254681908341696,
+      "loss": 1.0703,
+      "step": 3501
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31906228924539476,
+      "learning_rate": 0.00019254082510170503,
+      "loss": 1.0448,
+      "step": 3502
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27170761831755913,
+      "learning_rate": 0.0001925348288041008,
+      "loss": 1.1028,
+      "step": 3503
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2855022956911831,
+      "learning_rate": 0.00019252883019075433,
+      "loss": 1.0736,
+      "step": 3504
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2857537178954334,
+      "learning_rate": 0.0001925228292618157,
+      "loss": 1.0172,
+      "step": 3505
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26288096894393326,
+      "learning_rate": 0.0001925168260174351,
+      "loss": 1.0262,
+      "step": 3506
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2682150517518947,
+      "learning_rate": 0.00019251082045776283,
+      "loss": 1.0227,
+      "step": 3507
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28175497164571117,
+      "learning_rate": 0.00019250481258294911,
+      "loss": 1.0507,
+      "step": 3508
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28227600468816066,
+      "learning_rate": 0.00019249880239314435,
+      "loss": 1.1972,
+      "step": 3509
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2933560845393136,
+      "learning_rate": 0.0001924927898884989,
+      "loss": 1.0141,
+      "step": 3510
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2659979972720397,
+      "learning_rate": 0.0001924867750691633,
+      "loss": 1.1605,
+      "step": 3511
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28458190404309464,
+      "learning_rate": 0.00019248075793528794,
+      "loss": 1.1147,
+      "step": 3512
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30482251375794783,
+      "learning_rate": 0.00019247473848702358,
+      "loss": 1.2108,
+      "step": 3513
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30589628132765245,
+      "learning_rate": 0.00019246871672452072,
+      "loss": 1.1377,
+      "step": 3514
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2515859551958094,
+      "learning_rate": 0.00019246269264793013,
+      "loss": 1.0792,
+      "step": 3515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2822974779571469,
+      "learning_rate": 0.00019245666625740252,
+      "loss": 1.1263,
+      "step": 3516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26320433675688415,
+      "learning_rate": 0.00019245063755308873,
+      "loss": 1.0395,
+      "step": 3517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27747602120555126,
+      "learning_rate": 0.00019244460653513966,
+      "loss": 1.0819,
+      "step": 3518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27754855559046226,
+      "learning_rate": 0.00019243857320370622,
+      "loss": 1.1293,
+      "step": 3519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2718497648773705,
+      "learning_rate": 0.00019243253755893934,
+      "loss": 1.0904,
+      "step": 3520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24497317555871573,
+      "learning_rate": 0.00019242649960099018,
+      "loss": 1.0817,
+      "step": 3521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2832235216362736,
+      "learning_rate": 0.00019242045933000974,
+      "loss": 1.1896,
+      "step": 3522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2804650893498078,
+      "learning_rate": 0.00019241441674614925,
+      "loss": 1.0998,
+      "step": 3523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2746480898368063,
+      "learning_rate": 0.00019240837184955986,
+      "loss": 1.1329,
+      "step": 3524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27811869745054746,
+      "learning_rate": 0.0001924023246403929,
+      "loss": 1.2432,
+      "step": 3525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2438034625107248,
+      "learning_rate": 0.0001923962751187997,
+      "loss": 1.0454,
+      "step": 3526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29872712215291863,
+      "learning_rate": 0.00019239022328493166,
+      "loss": 1.0986,
+      "step": 3527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28352993745178234,
+      "learning_rate": 0.00019238416913894022,
+      "loss": 1.0937,
+      "step": 3528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2573940016691036,
+      "learning_rate": 0.00019237811268097685,
+      "loss": 1.061,
+      "step": 3529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26547291167111237,
+      "learning_rate": 0.00019237205391119317,
+      "loss": 1.1062,
+      "step": 3530
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2434569652472095,
+      "learning_rate": 0.0001923659928297408,
+      "loss": 0.994,
+      "step": 3531
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24604489384038644,
+      "learning_rate": 0.00019235992943677138,
+      "loss": 1.0407,
+      "step": 3532
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2566897641461854,
+      "learning_rate": 0.0001923538637324367,
+      "loss": 1.0814,
+      "step": 3533
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2833118574755522,
+      "learning_rate": 0.00019234779571688856,
+      "loss": 1.0133,
+      "step": 3534
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2641589772291687,
+      "learning_rate": 0.00019234172539027875,
+      "loss": 1.0779,
+      "step": 3535
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26348102788720923,
+      "learning_rate": 0.00019233565275275926,
+      "loss": 1.0349,
+      "step": 3536
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26442584979280237,
+      "learning_rate": 0.00019232957780448203,
+      "loss": 1.0908,
+      "step": 3537
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.21497138036257077,
+      "learning_rate": 0.00019232350054559908,
+      "loss": 1.0206,
+      "step": 3538
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28584115929317455,
+      "learning_rate": 0.00019231742097626248,
+      "loss": 1.0948,
+      "step": 3539
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28169323923794704,
+      "learning_rate": 0.00019231133909662442,
+      "loss": 1.1113,
+      "step": 3540
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27099757848452277,
+      "learning_rate": 0.0001923052549068371,
+      "loss": 1.1069,
+      "step": 3541
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2576793427814868,
+      "learning_rate": 0.00019229916840705276,
+      "loss": 0.9917,
+      "step": 3542
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.261636637446303,
+      "learning_rate": 0.0001922930795974237,
+      "loss": 1.0584,
+      "step": 3543
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27616031671725766,
+      "learning_rate": 0.0001922869884781023,
+      "loss": 1.0674,
+      "step": 3544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2603326493271169,
+      "learning_rate": 0.0001922808950492411,
+      "loss": 0.9925,
+      "step": 3545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2611098162609211,
+      "learning_rate": 0.00019227479931099243,
+      "loss": 1.1144,
+      "step": 3546
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28213224868364634,
+      "learning_rate": 0.00019226870126350893,
+      "loss": 0.9999,
+      "step": 3547
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26909330897869277,
+      "learning_rate": 0.00019226260090694322,
+      "loss": 1.1827,
+      "step": 3548
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2204883513538407,
+      "learning_rate": 0.00019225649824144788,
+      "loss": 1.0633,
+      "step": 3549
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2800738248678148,
+      "learning_rate": 0.00019225039326717575,
+      "loss": 1.0618,
+      "step": 3550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3026024017055852,
+      "learning_rate": 0.0001922442859842795,
+      "loss": 1.0372,
+      "step": 3551
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25052538631469556,
+      "learning_rate": 0.00019223817639291206,
+      "loss": 1.1017,
+      "step": 3552
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2534940085440102,
+      "learning_rate": 0.00019223206449322627,
+      "loss": 1.063,
+      "step": 3553
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2829769188702781,
+      "learning_rate": 0.0001922259502853751,
+      "loss": 1.0918,
+      "step": 3554
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2515647647546932,
+      "learning_rate": 0.0001922198337695116,
+      "loss": 1.1084,
+      "step": 3555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2716176574803204,
+      "learning_rate": 0.00019221371494578874,
+      "loss": 1.1048,
+      "step": 3556
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2544586203933426,
+      "learning_rate": 0.00019220759381435976,
+      "loss": 0.9334,
+      "step": 3557
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25686564637963666,
+      "learning_rate": 0.00019220147037537775,
+      "loss": 1.1342,
+      "step": 3558
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2856418475083659,
+      "learning_rate": 0.00019219534462899603,
+      "loss": 1.0993,
+      "step": 3559
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25892037090159264,
+      "learning_rate": 0.00019218921657536785,
+      "loss": 1.1242,
+      "step": 3560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22637468596243374,
+      "learning_rate": 0.00019218308621464657,
+      "loss": 0.9695,
+      "step": 3561
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24205716240527825,
+      "learning_rate": 0.00019217695354698566,
+      "loss": 0.9977,
+      "step": 3562
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28419771649107817,
+      "learning_rate": 0.00019217081857253855,
+      "loss": 1.1305,
+      "step": 3563
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2997025561180733,
+      "learning_rate": 0.00019216468129145878,
+      "loss": 1.1392,
+      "step": 3564
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2575883782828923,
+      "learning_rate": 0.00019215854170389992,
+      "loss": 1.1146,
+      "step": 3565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2378156355843756,
+      "learning_rate": 0.00019215239981001565,
+      "loss": 1.1623,
+      "step": 3566
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.22766346459594225,
+      "learning_rate": 0.00019214625560995963,
+      "loss": 1.1813,
+      "step": 3567
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2815971957249664,
+      "learning_rate": 0.0001921401091038857,
+      "loss": 1.0946,
+      "step": 3568
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24854801853444775,
+      "learning_rate": 0.0001921339602919476,
+      "loss": 1.1996,
+      "step": 3569
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28777374446942766,
+      "learning_rate": 0.00019212780917429923,
+      "loss": 1.2163,
+      "step": 3570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2755310826907456,
+      "learning_rate": 0.00019212165575109452,
+      "loss": 1.2163,
+      "step": 3571
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.23202223527427898,
+      "learning_rate": 0.00019211550002248755,
+      "loss": 1.0259,
+      "step": 3572
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25126040011653317,
+      "learning_rate": 0.00019210934198863225,
+      "loss": 1.1719,
+      "step": 3573
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.31182851203253326,
+      "learning_rate": 0.00019210318164968276,
+      "loss": 0.9726,
+      "step": 3574
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30037163533010824,
+      "learning_rate": 0.00019209701900579332,
+      "loss": 1.1423,
+      "step": 3575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2769715350587386,
+      "learning_rate": 0.00019209085405711806,
+      "loss": 1.023,
+      "step": 3576
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2737139517166542,
+      "learning_rate": 0.0001920846868038113,
+      "loss": 1.1156,
+      "step": 3577
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27215926972723625,
+      "learning_rate": 0.00019207851724602738,
+      "loss": 1.2292,
+      "step": 3578
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25589677103292885,
+      "learning_rate": 0.0001920723453839207,
+      "loss": 1.066,
+      "step": 3579
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27063989588024306,
+      "learning_rate": 0.00019206617121764573,
+      "loss": 1.1828,
+      "step": 3580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2621987437844738,
+      "learning_rate": 0.00019205999474735695,
+      "loss": 1.113,
+      "step": 3581
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27664343721303936,
+      "learning_rate": 0.00019205381597320895,
+      "loss": 1.0935,
+      "step": 3582
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2521436526457091,
+      "learning_rate": 0.00019204763489535633,
+      "loss": 1.0474,
+      "step": 3583
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29893749568707256,
+      "learning_rate": 0.00019204145151395383,
+      "loss": 1.0565,
+      "step": 3584
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2865188185320406,
+      "learning_rate": 0.00019203526582915615,
+      "loss": 1.0099,
+      "step": 3585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.30840096394075645,
+      "learning_rate": 0.0001920290778411181,
+      "loss": 1.0744,
+      "step": 3586
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27719489340614456,
+      "learning_rate": 0.00019202288754999454,
+      "loss": 1.1818,
+      "step": 3587
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26038266306240404,
+      "learning_rate": 0.00019201669495594036,
+      "loss": 1.1681,
+      "step": 3588
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2535864656959433,
+      "learning_rate": 0.00019201050005911057,
+      "loss": 1.0594,
+      "step": 3589
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26996214011774267,
+      "learning_rate": 0.0001920043028596602,
+      "loss": 1.0319,
+      "step": 3590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.293644543486573,
+      "learning_rate": 0.00019199810335774432,
+      "loss": 1.0956,
+      "step": 3591
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24802640655532923,
+      "learning_rate": 0.0001919919015535181,
+      "loss": 1.028,
+      "step": 3592
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.288494054514753,
+      "learning_rate": 0.0001919856974471367,
+      "loss": 1.2163,
+      "step": 3593
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28835626422546046,
+      "learning_rate": 0.00019197949103875542,
+      "loss": 1.1595,
+      "step": 3594
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2834452761231032,
+      "learning_rate": 0.00019197328232852957,
+      "loss": 1.2283,
+      "step": 3595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2712391659872608,
+      "learning_rate": 0.00019196707131661456,
+      "loss": 0.9865,
+      "step": 3596
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.28960322854085857,
+      "learning_rate": 0.00019196085800316577,
+      "loss": 1.0389,
+      "step": 3597
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.26819478685827985,
+      "learning_rate": 0.00019195464238833872,
+      "loss": 1.0676,
+      "step": 3598
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.27909258767693773,
+      "learning_rate": 0.00019194842447228894,
+      "loss": 1.1414,
+      "step": 3599
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.29782335526237375,
+      "learning_rate": 0.00019194220425517203,
+      "loss": 1.0944,
+      "step": 3600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2990681773116163,
+      "learning_rate": 0.00019193598173714368,
+      "loss": 1.2351,
+      "step": 3601
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.25309601233066126,
+      "learning_rate": 0.00019192975691835967,
+      "loss": 0.9833,
+      "step": 3602
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2530498249236964,
+      "learning_rate": 0.00019192352979897564,
+      "loss": 1.1116,
+      "step": 3603
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.261372825515641,
+      "learning_rate": 0.00019191730037914755,
+      "loss": 1.1451,
+      "step": 3604
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.2224404725086543,
+      "learning_rate": 0.00019191106865903125,
+      "loss": 1.1346,
+      "step": 3605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.24745135431170612,
+      "learning_rate": 0.00019190483463878266,
+      "loss": 1.0474,
+      "step": 3606
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2680064496452726,
+      "learning_rate": 0.00019189859831855786,
+      "loss": 1.0726,
+      "step": 3607
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2728343814388248,
+      "learning_rate": 0.00019189235969851285,
+      "loss": 1.143,
+      "step": 3608
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2758236497986754,
+      "learning_rate": 0.0001918861187788038,
+      "loss": 1.1073,
+      "step": 3609
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24891564132409086,
+      "learning_rate": 0.00019187987555958688,
+      "loss": 0.9501,
+      "step": 3610
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28175698358764867,
+      "learning_rate": 0.00019187363004101834,
+      "loss": 1.119,
+      "step": 3611
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29465454747771924,
+      "learning_rate": 0.00019186738222325446,
+      "loss": 1.1071,
+      "step": 3612
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28563894659906613,
+      "learning_rate": 0.00019186113210645158,
+      "loss": 0.9567,
+      "step": 3613
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2665261670344539,
+      "learning_rate": 0.00019185487969076618,
+      "loss": 1.0447,
+      "step": 3614
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29057112065403007,
+      "learning_rate": 0.00019184862497635466,
+      "loss": 1.094,
+      "step": 3615
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24917921323756612,
+      "learning_rate": 0.0001918423679633736,
+      "loss": 1.0332,
+      "step": 3616
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25724912043463755,
+      "learning_rate": 0.0001918361086519795,
+      "loss": 1.1526,
+      "step": 3617
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30458714457974434,
+      "learning_rate": 0.00019182984704232912,
+      "loss": 1.1059,
+      "step": 3618
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.233003803525938,
+      "learning_rate": 0.00019182358313457907,
+      "loss": 1.0059,
+      "step": 3619
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3095013736710616,
+      "learning_rate": 0.00019181731692888615,
+      "loss": 1.2226,
+      "step": 3620
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3112580093485402,
+      "learning_rate": 0.0001918110484254072,
+      "loss": 1.0111,
+      "step": 3621
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2615583637323915,
+      "learning_rate": 0.00019180477762429905,
+      "loss": 1.0417,
+      "step": 3622
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25331226852742855,
+      "learning_rate": 0.00019179850452571864,
+      "loss": 1.1169,
+      "step": 3623
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2653623351810319,
+      "learning_rate": 0.00019179222912982295,
+      "loss": 1.0001,
+      "step": 3624
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3027616299847975,
+      "learning_rate": 0.00019178595143676903,
+      "loss": 1.0122,
+      "step": 3625
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2596463436944122,
+      "learning_rate": 0.00019177967144671403,
+      "loss": 1.0603,
+      "step": 3626
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26322578302738187,
+      "learning_rate": 0.00019177338915981503,
+      "loss": 1.1179,
+      "step": 3627
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26304450867502877,
+      "learning_rate": 0.00019176710457622932,
+      "loss": 1.1117,
+      "step": 3628
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26025672008441436,
+      "learning_rate": 0.00019176081769611413,
+      "loss": 1.0988,
+      "step": 3629
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.264676786106681,
+      "learning_rate": 0.00019175452851962678,
+      "loss": 1.1449,
+      "step": 3630
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28753505819216246,
+      "learning_rate": 0.00019174823704692473,
+      "loss": 0.9785,
+      "step": 3631
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27286886659302656,
+      "learning_rate": 0.00019174194327816534,
+      "loss": 1.0712,
+      "step": 3632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2851788136509597,
+      "learning_rate": 0.0001917356472135062,
+      "loss": 1.1066,
+      "step": 3633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2596390703916512,
+      "learning_rate": 0.00019172934885310484,
+      "loss": 1.03,
+      "step": 3634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27793648005875066,
+      "learning_rate": 0.00019172304819711886,
+      "loss": 1.0643,
+      "step": 3635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2785181884043944,
+      "learning_rate": 0.00019171674524570595,
+      "loss": 1.1464,
+      "step": 3636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30776612944819687,
+      "learning_rate": 0.00019171043999902386,
+      "loss": 1.0927,
+      "step": 3637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2477006043242119,
+      "learning_rate": 0.00019170413245723036,
+      "loss": 1.0933,
+      "step": 3638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28543848973925623,
+      "learning_rate": 0.00019169782262048332,
+      "loss": 1.0738,
+      "step": 3639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23369616317344152,
+      "learning_rate": 0.00019169151048894066,
+      "loss": 1.0539,
+      "step": 3640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26607663383639485,
+      "learning_rate": 0.00019168519606276027,
+      "loss": 1.0068,
+      "step": 3641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27489969296500305,
+      "learning_rate": 0.00019167887934210028,
+      "loss": 1.1252,
+      "step": 3642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26853165269525164,
+      "learning_rate": 0.00019167256032711868,
+      "loss": 1.1099,
+      "step": 3643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26605252525945594,
+      "learning_rate": 0.00019166623901797366,
+      "loss": 1.063,
+      "step": 3644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2857813746118553,
+      "learning_rate": 0.0001916599154148234,
+      "loss": 1.1668,
+      "step": 3645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3004731926385603,
+      "learning_rate": 0.00019165358951782612,
+      "loss": 1.1223,
+      "step": 3646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30149091959544416,
+      "learning_rate": 0.0001916472613271402,
+      "loss": 1.0124,
+      "step": 3647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24733679580665724,
+      "learning_rate": 0.00019164093084292393,
+      "loss": 1.1462,
+      "step": 3648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.21990278220565118,
+      "learning_rate": 0.00019163459806533582,
+      "loss": 1.0147,
+      "step": 3649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26246347077998894,
+      "learning_rate": 0.00019162826299453427,
+      "loss": 1.0042,
+      "step": 3650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2552863623874917,
+      "learning_rate": 0.00019162192563067785,
+      "loss": 1.0509,
+      "step": 3651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26792837644378786,
+      "learning_rate": 0.00019161558597392516,
+      "loss": 1.0793,
+      "step": 3652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26602863063633964,
+      "learning_rate": 0.00019160924402443488,
+      "loss": 1.0786,
+      "step": 3653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2748453626108522,
+      "learning_rate": 0.00019160289978236567,
+      "loss": 1.1711,
+      "step": 3654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2892103795799046,
+      "learning_rate": 0.00019159655324787634,
+      "loss": 1.0855,
+      "step": 3655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29356881676510077,
+      "learning_rate": 0.00019159020442112567,
+      "loss": 1.0765,
+      "step": 3656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25797959221770866,
+      "learning_rate": 0.0001915838533022726,
+      "loss": 1.0762,
+      "step": 3657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24086139476910498,
+      "learning_rate": 0.00019157749989147602,
+      "loss": 1.0265,
+      "step": 3658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29161814627285754,
+      "learning_rate": 0.00019157114418889498,
+      "loss": 1.0909,
+      "step": 3659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2899831061438026,
+      "learning_rate": 0.0001915647861946885,
+      "loss": 1.0793,
+      "step": 3660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2444771404216185,
+      "learning_rate": 0.00019155842590901564,
+      "loss": 1.0686,
+      "step": 3661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2757863183788381,
+      "learning_rate": 0.0001915520633320357,
+      "loss": 1.1698,
+      "step": 3662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2850756152744979,
+      "learning_rate": 0.00019154569846390783,
+      "loss": 1.2098,
+      "step": 3663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2229292059572456,
+      "learning_rate": 0.00019153933130479128,
+      "loss": 1.1241,
+      "step": 3664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2763336437474556,
+      "learning_rate": 0.00019153296185484545,
+      "loss": 1.0689,
+      "step": 3665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26355818754319077,
+      "learning_rate": 0.00019152659011422975,
+      "loss": 1.1286,
+      "step": 3666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2834332018069187,
+      "learning_rate": 0.0001915202160831036,
+      "loss": 1.1524,
+      "step": 3667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2829528091832065,
+      "learning_rate": 0.0001915138397616265,
+      "loss": 1.1356,
+      "step": 3668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2549124521154805,
+      "learning_rate": 0.00019150746114995809,
+      "loss": 1.0091,
+      "step": 3669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2787455502134115,
+      "learning_rate": 0.00019150108024825792,
+      "loss": 1.1973,
+      "step": 3670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3045237462568005,
+      "learning_rate": 0.00019149469705668574,
+      "loss": 1.0021,
+      "step": 3671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2563873686787024,
+      "learning_rate": 0.00019148831157540122,
+      "loss": 1.0652,
+      "step": 3672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26639470855690445,
+      "learning_rate": 0.0001914819238045642,
+      "loss": 1.0402,
+      "step": 3673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24615772856913087,
+      "learning_rate": 0.00019147553374433457,
+      "loss": 1.1036,
+      "step": 3674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26050952114504417,
+      "learning_rate": 0.00019146914139487223,
+      "loss": 1.082,
+      "step": 3675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2711202214111707,
+      "learning_rate": 0.00019146274675633712,
+      "loss": 1.022,
+      "step": 3676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.27642386341556874,
+      "learning_rate": 0.0001914563498288893,
+      "loss": 1.1196,
+      "step": 3677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26017602194644773,
+      "learning_rate": 0.00019144995061268886,
+      "loss": 1.0743,
+      "step": 3678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2802611591626233,
+      "learning_rate": 0.00019144354910789592,
+      "loss": 1.1542,
+      "step": 3679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30109865674446373,
+      "learning_rate": 0.00019143714531467067,
+      "loss": 1.0761,
+      "step": 3680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29232637101985814,
+      "learning_rate": 0.00019143073923317343,
+      "loss": 1.1051,
+      "step": 3681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26726588819259994,
+      "learning_rate": 0.00019142433086356445,
+      "loss": 1.1052,
+      "step": 3682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3170937940560179,
+      "learning_rate": 0.00019141792020600416,
+      "loss": 1.0991,
+      "step": 3683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.26117278170269886,
+      "learning_rate": 0.00019141150726065292,
+      "loss": 1.1537,
+      "step": 3684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23933035001939557,
+      "learning_rate": 0.00019140509202767128,
+      "loss": 1.0034,
+      "step": 3685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.261470828936781,
+      "learning_rate": 0.00019139867450721978,
+      "loss": 1.0356,
+      "step": 3686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28669585869759606,
+      "learning_rate": 0.000191392254699459,
+      "loss": 1.1138,
+      "step": 3687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28522961501456195,
+      "learning_rate": 0.00019138583260454962,
+      "loss": 1.1261,
+      "step": 3688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.24515845894099997,
+      "learning_rate": 0.00019137940822265234,
+      "loss": 1.0393,
+      "step": 3689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30181790008831666,
+      "learning_rate": 0.00019137298155392794,
+      "loss": 1.1297,
+      "step": 3690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23025123557623256,
+      "learning_rate": 0.0001913665525985372,
+      "loss": 0.9475,
+      "step": 3691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2916644640665306,
+      "learning_rate": 0.0001913601213566411,
+      "loss": 1.0052,
+      "step": 3692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2914800928299035,
+      "learning_rate": 0.00019135368782840058,
+      "loss": 1.0656,
+      "step": 3693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2638508110242151,
+      "learning_rate": 0.00019134725201397655,
+      "loss": 1.1046,
+      "step": 3694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2862216325349102,
+      "learning_rate": 0.00019134081391353018,
+      "loss": 1.1441,
+      "step": 3695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.31225707110177847,
+      "learning_rate": 0.00019133437352722253,
+      "loss": 1.143,
+      "step": 3696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.30593521073577173,
+      "learning_rate": 0.00019132793085521477,
+      "loss": 1.1625,
+      "step": 3697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.29258514429430604,
+      "learning_rate": 0.00019132148589766812,
+      "loss": 1.1549,
+      "step": 3698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.25691353350560925,
+      "learning_rate": 0.00019131503865474388,
+      "loss": 1.086,
+      "step": 3699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28413967582202954,
+      "learning_rate": 0.00019130858912660346,
+      "loss": 1.0516,
+      "step": 3700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2486682279993249,
+      "learning_rate": 0.0001913021373134082,
+      "loss": 1.144,
+      "step": 3701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2799664249634028,
+      "learning_rate": 0.00019129568321531957,
+      "loss": 1.203,
+      "step": 3702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2360258518505738,
+      "learning_rate": 0.00019128922683249905,
+      "loss": 1.0592,
+      "step": 3703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.28969445762878065,
+      "learning_rate": 0.0001912827681651083,
+      "loss": 1.2039,
+      "step": 3704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2925284783203515,
+      "learning_rate": 0.0001912763072133089,
+      "loss": 1.0849,
+      "step": 3705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33511999769210443,
+      "learning_rate": 0.00019126984397726252,
+      "loss": 1.0263,
+      "step": 3706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2898231101162137,
+      "learning_rate": 0.00019126337845713098,
+      "loss": 1.042,
+      "step": 3707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2787807763701908,
+      "learning_rate": 0.000191256910653076,
+      "loss": 0.9801,
+      "step": 3708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.2655169550180212,
+      "learning_rate": 0.0001912504405652595,
+      "loss": 1.024,
+      "step": 3709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.23484914814404617,
+      "learning_rate": 0.00019124396819384336,
+      "loss": 1.083,
+      "step": 3710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2399809754737731,
+      "learning_rate": 0.00019123749353898957,
+      "loss": 0.8806,
+      "step": 3711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25071481306083293,
+      "learning_rate": 0.00019123101660086018,
+      "loss": 1.0832,
+      "step": 3712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26019964031949383,
+      "learning_rate": 0.00019122453737961724,
+      "loss": 1.105,
+      "step": 3713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27184915954233857,
+      "learning_rate": 0.00019121805587542294,
+      "loss": 1.1295,
+      "step": 3714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3232217248521111,
+      "learning_rate": 0.00019121157208843947,
+      "loss": 1.0362,
+      "step": 3715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2484485517327626,
+      "learning_rate": 0.00019120508601882906,
+      "loss": 1.0417,
+      "step": 3716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28882594941774437,
+      "learning_rate": 0.00019119859766675407,
+      "loss": 1.0759,
+      "step": 3717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26746773610204805,
+      "learning_rate": 0.00019119210703237685,
+      "loss": 0.9896,
+      "step": 3718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3030903175749709,
+      "learning_rate": 0.00019118561411585986,
+      "loss": 1.0477,
+      "step": 3719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2647502528125401,
+      "learning_rate": 0.00019117911891736552,
+      "loss": 1.0581,
+      "step": 3720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31861248281328974,
+      "learning_rate": 0.00019117262143705647,
+      "loss": 1.144,
+      "step": 3721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25783949986971005,
+      "learning_rate": 0.00019116612167509526,
+      "loss": 1.2301,
+      "step": 3722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24441169216486786,
+      "learning_rate": 0.00019115961963164454,
+      "loss": 1.118,
+      "step": 3723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2659364840495592,
+      "learning_rate": 0.00019115311530686706,
+      "loss": 1.0523,
+      "step": 3724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2522530337190297,
+      "learning_rate": 0.00019114660870092558,
+      "loss": 0.9005,
+      "step": 3725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2647152274884048,
+      "learning_rate": 0.00019114009981398294,
+      "loss": 1.0931,
+      "step": 3726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2702016934536367,
+      "learning_rate": 0.000191133588646202,
+      "loss": 1.0832,
+      "step": 3727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24668859082298314,
+      "learning_rate": 0.00019112707519774576,
+      "loss": 0.9758,
+      "step": 3728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2672446623933551,
+      "learning_rate": 0.0001911205594687772,
+      "loss": 1.149,
+      "step": 3729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2600484573038422,
+      "learning_rate": 0.00019111404145945933,
+      "loss": 1.057,
+      "step": 3730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31076021763085576,
+      "learning_rate": 0.00019110752116995535,
+      "loss": 1.1161,
+      "step": 3731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2626503992229002,
+      "learning_rate": 0.00019110099860042835,
+      "loss": 1.1429,
+      "step": 3732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2550015497297485,
+      "learning_rate": 0.00019109447375104165,
+      "loss": 1.207,
+      "step": 3733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29381829699592593,
+      "learning_rate": 0.00019108794662195847,
+      "loss": 1.071,
+      "step": 3734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2582343525213059,
+      "learning_rate": 0.00019108141721334217,
+      "loss": 1.0302,
+      "step": 3735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3033664358520454,
+      "learning_rate": 0.00019107488552535617,
+      "loss": 1.0559,
+      "step": 3736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2804498013242838,
+      "learning_rate": 0.00019106835155816395,
+      "loss": 1.1039,
+      "step": 3737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2804541814983532,
+      "learning_rate": 0.00019106181531192894,
+      "loss": 1.1296,
+      "step": 3738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.30659613447065276,
+      "learning_rate": 0.0001910552767868148,
+      "loss": 1.1612,
+      "step": 3739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2891921914653932,
+      "learning_rate": 0.00019104873598298517,
+      "loss": 1.0669,
+      "step": 3740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28558383598746206,
+      "learning_rate": 0.00019104219290060366,
+      "loss": 1.2211,
+      "step": 3741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29494252698624446,
+      "learning_rate": 0.00019103564753983405,
+      "loss": 0.9863,
+      "step": 3742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2556255730742099,
+      "learning_rate": 0.00019102909990084018,
+      "loss": 1.0618,
+      "step": 3743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2506965201053014,
+      "learning_rate": 0.00019102254998378584,
+      "loss": 0.9738,
+      "step": 3744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29913681320353497,
+      "learning_rate": 0.00019101599778883498,
+      "loss": 1.0876,
+      "step": 3745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2773312706953958,
+      "learning_rate": 0.0001910094433161516,
+      "loss": 1.0741,
+      "step": 3746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2917033659607459,
+      "learning_rate": 0.0001910028865658997,
+      "loss": 1.1748,
+      "step": 3747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2296887419120594,
+      "learning_rate": 0.00019099632753824335,
+      "loss": 0.9938,
+      "step": 3748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2704584908164929,
+      "learning_rate": 0.00019098976623334673,
+      "loss": 1.0166,
+      "step": 3749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.21917669304416104,
+      "learning_rate": 0.00019098320265137402,
+      "loss": 1.0756,
+      "step": 3750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26280707594546593,
+      "learning_rate": 0.0001909766367924895,
+      "loss": 1.2615,
+      "step": 3751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29377384953319624,
+      "learning_rate": 0.00019097006865685743,
+      "loss": 1.0886,
+      "step": 3752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.23301855169248875,
+      "learning_rate": 0.00019096349824464226,
+      "loss": 0.9808,
+      "step": 3753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.253520755622324,
+      "learning_rate": 0.00019095692555600832,
+      "loss": 1.1587,
+      "step": 3754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2754796183392769,
+      "learning_rate": 0.00019095035059112024,
+      "loss": 1.1064,
+      "step": 3755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2747255493731151,
+      "learning_rate": 0.00019094377335014242,
+      "loss": 1.055,
+      "step": 3756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28773064135087933,
+      "learning_rate": 0.00019093719383323952,
+      "loss": 1.1881,
+      "step": 3757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2806691047427897,
+      "learning_rate": 0.0001909306120405762,
+      "loss": 1.1801,
+      "step": 3758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3132693357401042,
+      "learning_rate": 0.00019092402797231715,
+      "loss": 1.0493,
+      "step": 3759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25209467298228705,
+      "learning_rate": 0.00019091744162862717,
+      "loss": 1.03,
+      "step": 3760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31060570827727507,
+      "learning_rate": 0.0001909108530096711,
+      "loss": 1.21,
+      "step": 3761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25830756717657666,
+      "learning_rate": 0.00019090426211561376,
+      "loss": 1.0167,
+      "step": 3762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2535702002191736,
+      "learning_rate": 0.00019089766894662014,
+      "loss": 1.0515,
+      "step": 3763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.278969068298096,
+      "learning_rate": 0.00019089107350285522,
+      "loss": 1.1331,
+      "step": 3764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2807324463593699,
+      "learning_rate": 0.00019088447578448407,
+      "loss": 1.0715,
+      "step": 3765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28955522038757464,
+      "learning_rate": 0.0001908778757916718,
+      "loss": 1.0668,
+      "step": 3766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2595700503045936,
+      "learning_rate": 0.00019087127352458358,
+      "loss": 1.1446,
+      "step": 3767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28444433757950754,
+      "learning_rate": 0.00019086466898338462,
+      "loss": 1.0789,
+      "step": 3768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2787597254903445,
+      "learning_rate": 0.00019085806216824017,
+      "loss": 0.9789,
+      "step": 3769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2941116656712117,
+      "learning_rate": 0.0001908514530793157,
+      "loss": 1.0751,
+      "step": 3770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.22941662757128714,
+      "learning_rate": 0.00019084484171677646,
+      "loss": 1.1191,
+      "step": 3771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24999713030985823,
+      "learning_rate": 0.00019083822808078798,
+      "loss": 1.1661,
+      "step": 3772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28806317138126275,
+      "learning_rate": 0.00019083161217151574,
+      "loss": 1.138,
+      "step": 3773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26401411481347825,
+      "learning_rate": 0.00019082499398912533,
+      "loss": 1.0933,
+      "step": 3774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27780628237715865,
+      "learning_rate": 0.00019081837353378237,
+      "loss": 1.1479,
+      "step": 3775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2567041443182185,
+      "learning_rate": 0.00019081175080565253,
+      "loss": 0.9848,
+      "step": 3776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.33134929043752975,
+      "learning_rate": 0.00019080512580490154,
+      "loss": 1.005,
+      "step": 3777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.25790224327167643,
+      "learning_rate": 0.0001907984985316952,
+      "loss": 1.063,
+      "step": 3778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2542673197030738,
+      "learning_rate": 0.0001907918689861994,
+      "loss": 1.1054,
+      "step": 3779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29841436435986246,
+      "learning_rate": 0.00019078523716858,
+      "loss": 1.0715,
+      "step": 3780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2607650355473665,
+      "learning_rate": 0.000190778603079003,
+      "loss": 1.0562,
+      "step": 3781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27320429258836315,
+      "learning_rate": 0.00019077196671763436,
+      "loss": 1.0277,
+      "step": 3782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2548164654391468,
+      "learning_rate": 0.00019076532808464026,
+      "loss": 1.0822,
+      "step": 3783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3085428982977869,
+      "learning_rate": 0.00019075868718018677,
+      "loss": 1.1456,
+      "step": 3784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.31897081017855045,
+      "learning_rate": 0.0001907520440044401,
+      "loss": 1.0907,
+      "step": 3785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2491436372725657,
+      "learning_rate": 0.00019074539855756646,
+      "loss": 1.0872,
+      "step": 3786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2844533490165419,
+      "learning_rate": 0.00019073875083973222,
+      "loss": 1.0052,
+      "step": 3787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3018741634697304,
+      "learning_rate": 0.0001907321008511037,
+      "loss": 0.9965,
+      "step": 3788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2423698565672897,
+      "learning_rate": 0.0001907254485918473,
+      "loss": 1.0812,
+      "step": 3789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2975822751648389,
+      "learning_rate": 0.0001907187940621296,
+      "loss": 0.9894,
+      "step": 3790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29200730112195994,
+      "learning_rate": 0.000190712137262117,
+      "loss": 1.0714,
+      "step": 3791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28370419210343656,
+      "learning_rate": 0.0001907054781919762,
+      "loss": 1.077,
+      "step": 3792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2722830323949,
+      "learning_rate": 0.0001906988168518738,
+      "loss": 1.0713,
+      "step": 3793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24874541049005577,
+      "learning_rate": 0.00019069215324197646,
+      "loss": 1.0738,
+      "step": 3794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28546406332135904,
+      "learning_rate": 0.00019068548736245102,
+      "loss": 1.1608,
+      "step": 3795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.29760842933931714,
+      "learning_rate": 0.00019067881921346427,
+      "loss": 1.0152,
+      "step": 3796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2656364460449629,
+      "learning_rate": 0.00019067214879518306,
+      "loss": 1.1331,
+      "step": 3797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2807658825912433,
+      "learning_rate": 0.00019066547610777437,
+      "loss": 1.0905,
+      "step": 3798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2821813716967829,
+      "learning_rate": 0.00019065880115140513,
+      "loss": 1.0497,
+      "step": 3799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26730901537371593,
+      "learning_rate": 0.00019065212392624243,
+      "loss": 1.0252,
+      "step": 3800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26067574808177124,
+      "learning_rate": 0.00019064544443245335,
+      "loss": 0.9627,
+      "step": 3801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26226327026433305,
+      "learning_rate": 0.00019063876267020507,
+      "loss": 1.0757,
+      "step": 3802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2804595303992103,
+      "learning_rate": 0.00019063207863966478,
+      "loss": 1.1859,
+      "step": 3803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.265318805788959,
+      "learning_rate": 0.00019062539234099973,
+      "loss": 1.0262,
+      "step": 3804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2530307087999173,
+      "learning_rate": 0.00019061870377437733,
+      "loss": 1.0954,
+      "step": 3805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2734697145712888,
+      "learning_rate": 0.00019061201293996488,
+      "loss": 1.0704,
+      "step": 3806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.24462035828995993,
+      "learning_rate": 0.00019060531983792987,
+      "loss": 1.1139,
+      "step": 3807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2894517260409347,
+      "learning_rate": 0.00019059862446843982,
+      "loss": 1.0643,
+      "step": 3808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.28221499051147075,
+      "learning_rate": 0.00019059192683166222,
+      "loss": 1.0379,
+      "step": 3809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2596597115877193,
+      "learning_rate": 0.00019058522692776473,
+      "loss": 1.1256,
+      "step": 3810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.26056578947550624,
+      "learning_rate": 0.00019057852475691498,
+      "loss": 1.002,
+      "step": 3811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27913230085219015,
+      "learning_rate": 0.00019057182031928074,
+      "loss": 1.1348,
+      "step": 3812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.27929429057142885,
+      "learning_rate": 0.00019056511361502975,
+      "loss": 1.1987,
+      "step": 3813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.235530597825584,
+      "learning_rate": 0.0001905584046443299,
+      "loss": 1.1361,
+      "step": 3814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.2522023187837088,
+      "learning_rate": 0.00019055169340734908,
+      "loss": 1.0713,
+      "step": 3815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.280427344983831,
+      "learning_rate": 0.0001905449799042552,
+      "loss": 1.0557,
+      "step": 3816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2969054368145568,
+      "learning_rate": 0.0001905382641352163,
+      "loss": 1.1357,
+      "step": 3817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2440950945037354,
+      "learning_rate": 0.00019053154610040044,
+      "loss": 0.9862,
+      "step": 3818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26826804773960794,
+      "learning_rate": 0.0001905248257999757,
+      "loss": 1.1239,
+      "step": 3819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2505950276337422,
+      "learning_rate": 0.00019051810323411034,
+      "loss": 1.0886,
+      "step": 3820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27676795581151226,
+      "learning_rate": 0.00019051137840297256,
+      "loss": 1.1414,
+      "step": 3821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2571438975717034,
+      "learning_rate": 0.00019050465130673067,
+      "loss": 1.0289,
+      "step": 3822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25111218618944997,
+      "learning_rate": 0.00019049792194555294,
+      "loss": 0.9651,
+      "step": 3823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2610987856579739,
+      "learning_rate": 0.00019049119031960788,
+      "loss": 1.1635,
+      "step": 3824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2432839279077218,
+      "learning_rate": 0.00019048445642906388,
+      "loss": 1.1106,
+      "step": 3825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25411198588402173,
+      "learning_rate": 0.00019047772027408954,
+      "loss": 1.0766,
+      "step": 3826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2611685254036077,
+      "learning_rate": 0.00019047098185485335,
+      "loss": 1.0616,
+      "step": 3827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2486133581840728,
+      "learning_rate": 0.00019046424117152402,
+      "loss": 0.9964,
+      "step": 3828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3298032342837613,
+      "learning_rate": 0.00019045749822427016,
+      "loss": 0.9625,
+      "step": 3829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29292144210717325,
+      "learning_rate": 0.00019045075301326057,
+      "loss": 1.0938,
+      "step": 3830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2906727228238908,
+      "learning_rate": 0.00019044400553866405,
+      "loss": 1.0653,
+      "step": 3831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25691594404144225,
+      "learning_rate": 0.00019043725580064939,
+      "loss": 1.0817,
+      "step": 3832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2787203095759583,
+      "learning_rate": 0.00019043050379938565,
+      "loss": 1.0643,
+      "step": 3833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26242429857133936,
+      "learning_rate": 0.00019042374953504165,
+      "loss": 1.0383,
+      "step": 3834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2750170989226645,
+      "learning_rate": 0.00019041699300778654,
+      "loss": 1.0113,
+      "step": 3835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24689367063544376,
+      "learning_rate": 0.00019041023421778933,
+      "loss": 1.0589,
+      "step": 3836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3101949979537771,
+      "learning_rate": 0.0001904034731652192,
+      "loss": 1.0844,
+      "step": 3837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3026108279779279,
+      "learning_rate": 0.00019039670985024533,
+      "loss": 0.9748,
+      "step": 3838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28436273819827806,
+      "learning_rate": 0.00019038994427303697,
+      "loss": 1.13,
+      "step": 3839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26672874828160903,
+      "learning_rate": 0.00019038317643376346,
+      "loss": 0.9425,
+      "step": 3840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2551937537263912,
+      "learning_rate": 0.00019037640633259417,
+      "loss": 1.1193,
+      "step": 3841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2421616493910976,
+      "learning_rate": 0.00019036963396969848,
+      "loss": 1.1925,
+      "step": 3842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27403481599431856,
+      "learning_rate": 0.00019036285934524594,
+      "loss": 0.9328,
+      "step": 3843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2736343566787731,
+      "learning_rate": 0.00019035608245940603,
+      "loss": 1.1229,
+      "step": 3844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2984644122142435,
+      "learning_rate": 0.00019034930331234836,
+      "loss": 1.0706,
+      "step": 3845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25600173944542287,
+      "learning_rate": 0.00019034252190424264,
+      "loss": 1.1189,
+      "step": 3846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27750211746463055,
+      "learning_rate": 0.0001903357382352585,
+      "loss": 1.2051,
+      "step": 3847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.308006852697485,
+      "learning_rate": 0.00019032895230556573,
+      "loss": 1.1935,
+      "step": 3848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2606236441816338,
+      "learning_rate": 0.00019032216411533415,
+      "loss": 1.1713,
+      "step": 3849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28436108514736663,
+      "learning_rate": 0.00019031537366473369,
+      "loss": 1.0714,
+      "step": 3850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3140913304294229,
+      "learning_rate": 0.00019030858095393422,
+      "loss": 1.109,
+      "step": 3851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27820186741488906,
+      "learning_rate": 0.00019030178598310573,
+      "loss": 1.1456,
+      "step": 3852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25313171010488306,
+      "learning_rate": 0.00019029498875241832,
+      "loss": 1.0996,
+      "step": 3853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26632647868200215,
+      "learning_rate": 0.00019028818926204207,
+      "loss": 1.0288,
+      "step": 3854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2802151300023979,
+      "learning_rate": 0.00019028138751214714,
+      "loss": 1.1178,
+      "step": 3855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28551319793457147,
+      "learning_rate": 0.00019027458350290375,
+      "loss": 1.1742,
+      "step": 3856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29354399099059836,
+      "learning_rate": 0.00019026777723448214,
+      "loss": 1.1434,
+      "step": 3857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28766227685882445,
+      "learning_rate": 0.00019026096870705274,
+      "loss": 1.1227,
+      "step": 3858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25123810673486535,
+      "learning_rate": 0.0001902541579207858,
+      "loss": 1.0514,
+      "step": 3859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3035800237589507,
+      "learning_rate": 0.00019024734487585186,
+      "loss": 1.0958,
+      "step": 3860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24205745440797985,
+      "learning_rate": 0.0001902405295724214,
+      "loss": 1.1692,
+      "step": 3861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27865447448135033,
+      "learning_rate": 0.00019023371201066497,
+      "loss": 1.0144,
+      "step": 3862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2660377758295032,
+      "learning_rate": 0.0001902268921907532,
+      "loss": 0.9894,
+      "step": 3863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2726012616002911,
+      "learning_rate": 0.00019022007011285674,
+      "loss": 1.0454,
+      "step": 3864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2726846785840144,
+      "learning_rate": 0.0001902132457771463,
+      "loss": 1.0598,
+      "step": 3865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25701844846319727,
+      "learning_rate": 0.0001902064191837927,
+      "loss": 1.1679,
+      "step": 3866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2921405168078913,
+      "learning_rate": 0.00019019959033296678,
+      "loss": 1.0524,
+      "step": 3867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27129522822886254,
+      "learning_rate": 0.00019019275922483943,
+      "loss": 1.105,
+      "step": 3868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2718161044722209,
+      "learning_rate": 0.0001901859258595816,
+      "loss": 1.0613,
+      "step": 3869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2536896991586868,
+      "learning_rate": 0.00019017909023736428,
+      "loss": 0.9996,
+      "step": 3870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29175405932195664,
+      "learning_rate": 0.00019017225235835853,
+      "loss": 1.1386,
+      "step": 3871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.30144877670952286,
+      "learning_rate": 0.00019016541222273553,
+      "loss": 1.1468,
+      "step": 3872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2931426576366692,
+      "learning_rate": 0.00019015856983066644,
+      "loss": 1.1721,
+      "step": 3873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2916770600223048,
+      "learning_rate": 0.00019015172518232242,
+      "loss": 1.0945,
+      "step": 3874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27166976489578,
+      "learning_rate": 0.00019014487827787483,
+      "loss": 1.0891,
+      "step": 3875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2839705782430795,
+      "learning_rate": 0.00019013802911749505,
+      "loss": 1.1197,
+      "step": 3876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.298375581966575,
+      "learning_rate": 0.0001901311777013544,
+      "loss": 1.0689,
+      "step": 3877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3013484850412901,
+      "learning_rate": 0.0001901243240296244,
+      "loss": 1.0352,
+      "step": 3878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25624648167577585,
+      "learning_rate": 0.00019011746810247658,
+      "loss": 1.0749,
+      "step": 3879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25490116870005675,
+      "learning_rate": 0.00019011060992008244,
+      "loss": 1.1551,
+      "step": 3880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28069672688477176,
+      "learning_rate": 0.00019010374948261367,
+      "loss": 0.9975,
+      "step": 3881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2651997155968314,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 1.1731,
+      "step": 3882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3115313345415178,
+      "learning_rate": 0.00019009002184313897,
+      "loss": 1.1077,
+      "step": 3883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2775305592186655,
+      "learning_rate": 0.00019008315464147662,
+      "loss": 1.1736,
+      "step": 3884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28786278339401167,
+      "learning_rate": 0.0001900762851854267,
+      "loss": 1.0521,
+      "step": 3885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.23797474326903245,
+      "learning_rate": 0.0001900694134751611,
+      "loss": 1.1405,
+      "step": 3886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2605654562908968,
+      "learning_rate": 0.00019006253951085186,
+      "loss": 1.029,
+      "step": 3887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25694650291010046,
+      "learning_rate": 0.00019005566329267096,
+      "loss": 1.0665,
+      "step": 3888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26716631196903606,
+      "learning_rate": 0.0001900487848207905,
+      "loss": 1.1494,
+      "step": 3889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3116260242272765,
+      "learning_rate": 0.00019004190409538255,
+      "loss": 1.117,
+      "step": 3890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26774746733662874,
+      "learning_rate": 0.00019003502111661943,
+      "loss": 1.1987,
+      "step": 3891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.25363750120087714,
+      "learning_rate": 0.0001900281358846733,
+      "loss": 1.0831,
+      "step": 3892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28339718118220925,
+      "learning_rate": 0.00019002124839971647,
+      "loss": 1.1161,
+      "step": 3893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2541180184606548,
+      "learning_rate": 0.00019001435866192133,
+      "loss": 1.037,
+      "step": 3894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2662101976131836,
+      "learning_rate": 0.0001900074666714603,
+      "loss": 1.0495,
+      "step": 3895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24244962084116864,
+      "learning_rate": 0.00019000057242850584,
+      "loss": 1.143,
+      "step": 3896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2815830104446293,
+      "learning_rate": 0.00018999367593323048,
+      "loss": 1.1087,
+      "step": 3897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2661433168734344,
+      "learning_rate": 0.00018998677718580687,
+      "loss": 1.0699,
+      "step": 3898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.278452400141501,
+      "learning_rate": 0.00018997987618640756,
+      "loss": 0.9984,
+      "step": 3899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2950352883266951,
+      "learning_rate": 0.00018997297293520533,
+      "loss": 1.0725,
+      "step": 3900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24855309318205562,
+      "learning_rate": 0.00018996606743237288,
+      "loss": 1.1091,
+      "step": 3901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26157079774866104,
+      "learning_rate": 0.00018995915967808305,
+      "loss": 1.137,
+      "step": 3902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2817964268191354,
+      "learning_rate": 0.00018995224967250873,
+      "loss": 1.0887,
+      "step": 3903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.303055127418326,
+      "learning_rate": 0.00018994533741582283,
+      "loss": 1.0509,
+      "step": 3904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29319713632683003,
+      "learning_rate": 0.00018993842290819833,
+      "loss": 1.1527,
+      "step": 3905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.28460911802279176,
+      "learning_rate": 0.00018993150614980824,
+      "loss": 1.1663,
+      "step": 3906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.24323458888587937,
+      "learning_rate": 0.00018992458714082574,
+      "loss": 1.107,
+      "step": 3907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2507473651775092,
+      "learning_rate": 0.0001899176658814239,
+      "loss": 1.1291,
+      "step": 3908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26851442265034603,
+      "learning_rate": 0.00018991074237177595,
+      "loss": 1.0296,
+      "step": 3909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.27910401854051703,
+      "learning_rate": 0.0001899038166120552,
+      "loss": 1.0243,
+      "step": 3910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2643335688729103,
+      "learning_rate": 0.0001898968886024349,
+      "loss": 1.1176,
+      "step": 3911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2845122162938141,
+      "learning_rate": 0.0001898899583430885,
+      "loss": 1.1254,
+      "step": 3912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2631105012541486,
+      "learning_rate": 0.00018988302583418937,
+      "loss": 1.0436,
+      "step": 3913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26527130872751753,
+      "learning_rate": 0.00018987609107591104,
+      "loss": 1.0063,
+      "step": 3914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.26612171965346043,
+      "learning_rate": 0.00018986915406842708,
+      "loss": 1.1018,
+      "step": 3915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.29895893771415827,
+      "learning_rate": 0.000189862214811911,
+      "loss": 1.0477,
+      "step": 3916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.298508391495162,
+      "learning_rate": 0.00018985527330653653,
+      "loss": 1.1202,
+      "step": 3917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2862203670774759,
+      "learning_rate": 0.0001898483295524774,
+      "loss": 1.1676,
+      "step": 3918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7409275636159137,
+      "learning_rate": 0.00018984138354990736,
+      "loss": 1.4645,
+      "step": 3919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26519753642569927,
+      "learning_rate": 0.0001898344352990002,
+      "loss": 1.0743,
+      "step": 3920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29786591251006633,
+      "learning_rate": 0.00018982748479992988,
+      "loss": 1.1278,
+      "step": 3921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.278707089601725,
+      "learning_rate": 0.00018982053205287024,
+      "loss": 1.0982,
+      "step": 3922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.233468164464521,
+      "learning_rate": 0.00018981357705799538,
+      "loss": 1.0338,
+      "step": 3923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32242913731155076,
+      "learning_rate": 0.0001898066198154793,
+      "loss": 1.1282,
+      "step": 3924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2712636448427821,
+      "learning_rate": 0.00018979966032549612,
+      "loss": 1.0868,
+      "step": 3925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27812674663111897,
+      "learning_rate": 0.00018979269858822,
+      "loss": 0.9507,
+      "step": 3926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31036023474014135,
+      "learning_rate": 0.00018978573460382516,
+      "loss": 1.066,
+      "step": 3927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2741036928468329,
+      "learning_rate": 0.00018977876837248587,
+      "loss": 1.114,
+      "step": 3928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32960721822124844,
+      "learning_rate": 0.0001897717998943765,
+      "loss": 0.9916,
+      "step": 3929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25938355281972847,
+      "learning_rate": 0.0001897648291696714,
+      "loss": 1.0566,
+      "step": 3930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27023081057739146,
+      "learning_rate": 0.00018975785619854504,
+      "loss": 1.1365,
+      "step": 3931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30667049605255553,
+      "learning_rate": 0.00018975088098117194,
+      "loss": 1.0995,
+      "step": 3932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24394315516483825,
+      "learning_rate": 0.00018974390351772665,
+      "loss": 1.0412,
+      "step": 3933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2801905314214369,
+      "learning_rate": 0.00018973692380838371,
+      "loss": 1.0848,
+      "step": 3934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2381280727680932,
+      "learning_rate": 0.00018972994185331788,
+      "loss": 1.0596,
+      "step": 3935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.23812773003536764,
+      "learning_rate": 0.00018972295765270388,
+      "loss": 1.0666,
+      "step": 3936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32096602227564136,
+      "learning_rate": 0.00018971597120671647,
+      "loss": 1.0948,
+      "step": 3937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26277005671959386,
+      "learning_rate": 0.0001897089825155305,
+      "loss": 0.992,
+      "step": 3938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25895544694478345,
+      "learning_rate": 0.00018970199157932084,
+      "loss": 1.0121,
+      "step": 3939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.33276374338059417,
+      "learning_rate": 0.0001896949983982625,
+      "loss": 1.1811,
+      "step": 3940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2899952697366727,
+      "learning_rate": 0.00018968800297253043,
+      "loss": 1.0459,
+      "step": 3941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28332378815166603,
+      "learning_rate": 0.0001896810053022997,
+      "loss": 1.1406,
+      "step": 3942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2563768370833164,
+      "learning_rate": 0.00018967400538774548,
+      "loss": 1.0918,
+      "step": 3943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2527720587986432,
+      "learning_rate": 0.00018966700322904293,
+      "loss": 1.0922,
+      "step": 3944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25592462221706,
+      "learning_rate": 0.00018965999882636725,
+      "loss": 1.0781,
+      "step": 3945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3089279882540109,
+      "learning_rate": 0.00018965299217989375,
+      "loss": 1.1559,
+      "step": 3946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25525835490586896,
+      "learning_rate": 0.00018964598328979776,
+      "loss": 1.1349,
+      "step": 3947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2841608784955134,
+      "learning_rate": 0.00018963897215625472,
+      "loss": 0.984,
+      "step": 3948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.272016327051106,
+      "learning_rate": 0.00018963195877944007,
+      "loss": 1.058,
+      "step": 3949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3037390334949688,
+      "learning_rate": 0.0001896249431595293,
+      "loss": 1.0846,
+      "step": 3950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2497613489471199,
+      "learning_rate": 0.000189617925296698,
+      "loss": 1.0793,
+      "step": 3951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27852691608884916,
+      "learning_rate": 0.00018961090519112182,
+      "loss": 1.1262,
+      "step": 3952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2844330777819289,
+      "learning_rate": 0.0001896038828429764,
+      "loss": 1.0902,
+      "step": 3953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.30372667850264373,
+      "learning_rate": 0.0001895968582524375,
+      "loss": 0.9642,
+      "step": 3954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24532996718177777,
+      "learning_rate": 0.00018958983141968095,
+      "loss": 0.9768,
+      "step": 3955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27188695012581215,
+      "learning_rate": 0.0001895828023448825,
+      "loss": 1.0809,
+      "step": 3956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28620857758847607,
+      "learning_rate": 0.00018957577102821817,
+      "loss": 1.1338,
+      "step": 3957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2967269096451259,
+      "learning_rate": 0.00018956873746986386,
+      "loss": 1.0498,
+      "step": 3958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28197346617437485,
+      "learning_rate": 0.00018956170166999558,
+      "loss": 1.1001,
+      "step": 3959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32447316500456774,
+      "learning_rate": 0.00018955466362878943,
+      "loss": 1.1346,
+      "step": 3960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29206790559618895,
+      "learning_rate": 0.00018954762334642158,
+      "loss": 1.1531,
+      "step": 3961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.31059482758918283,
+      "learning_rate": 0.00018954058082306817,
+      "loss": 1.0824,
+      "step": 3962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2599356809697041,
+      "learning_rate": 0.0001895335360589054,
+      "loss": 1.0651,
+      "step": 3963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28568670159243,
+      "learning_rate": 0.00018952648905410966,
+      "loss": 1.111,
+      "step": 3964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2741580155840068,
+      "learning_rate": 0.0001895194398088573,
+      "loss": 1.0367,
+      "step": 3965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2460551214193779,
+      "learning_rate": 0.00018951238832332464,
+      "loss": 1.1306,
+      "step": 3966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28164939664440597,
+      "learning_rate": 0.00018950533459768823,
+      "loss": 1.0736,
+      "step": 3967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2935359891153058,
+      "learning_rate": 0.00018949827863212456,
+      "loss": 1.1287,
+      "step": 3968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27127546853360357,
+      "learning_rate": 0.00018949122042681023,
+      "loss": 0.9919,
+      "step": 3969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3153198801953517,
+      "learning_rate": 0.00018948415998192182,
+      "loss": 1.1561,
+      "step": 3970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2639733592331646,
+      "learning_rate": 0.0001894770972976361,
+      "loss": 1.0033,
+      "step": 3971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26313646288953435,
+      "learning_rate": 0.0001894700323741298,
+      "loss": 1.1798,
+      "step": 3972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2711515260554288,
+      "learning_rate": 0.0001894629652115797,
+      "loss": 1.1126,
+      "step": 3973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.272492854127571,
+      "learning_rate": 0.0001894558958101627,
+      "loss": 0.9806,
+      "step": 3974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27502865859424286,
+      "learning_rate": 0.00018944882417005565,
+      "loss": 1.088,
+      "step": 3975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3343809998581711,
+      "learning_rate": 0.00018944175029143558,
+      "loss": 1.1771,
+      "step": 3976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2894367965617339,
+      "learning_rate": 0.0001894346741744795,
+      "loss": 1.1597,
+      "step": 3977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25286298932737533,
+      "learning_rate": 0.00018942759581936446,
+      "loss": 1.0403,
+      "step": 3978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27494211784286493,
+      "learning_rate": 0.00018942051522626764,
+      "loss": 1.1405,
+      "step": 3979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28818353355011783,
+      "learning_rate": 0.00018941343239536624,
+      "loss": 1.0295,
+      "step": 3980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.25536310488024094,
+      "learning_rate": 0.0001894063473268375,
+      "loss": 1.0547,
+      "step": 3981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.32569914313859244,
+      "learning_rate": 0.00018939926002085872,
+      "loss": 1.2414,
+      "step": 3982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2869364080234295,
+      "learning_rate": 0.0001893921704776073,
+      "loss": 0.9819,
+      "step": 3983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2657350930953723,
+      "learning_rate": 0.0001893850786972606,
+      "loss": 1.0408,
+      "step": 3984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29775662133242436,
+      "learning_rate": 0.0001893779846799961,
+      "loss": 1.1624,
+      "step": 3985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27355863349278053,
+      "learning_rate": 0.00018937088842599142,
+      "loss": 1.2827,
+      "step": 3986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2732303129004138,
+      "learning_rate": 0.00018936378993542408,
+      "loss": 1.0976,
+      "step": 3987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3021647282609973,
+      "learning_rate": 0.00018935668920847171,
+      "loss": 1.0696,
+      "step": 3988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.272227355179039,
+      "learning_rate": 0.00018934958624531207,
+      "loss": 1.0169,
+      "step": 3989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2958029148879499,
+      "learning_rate": 0.00018934248104612283,
+      "loss": 1.1677,
+      "step": 3990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2961165738395803,
+      "learning_rate": 0.00018933537361108188,
+      "loss": 1.1413,
+      "step": 3991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3276060272777321,
+      "learning_rate": 0.00018932826394036707,
+      "loss": 1.1431,
+      "step": 3992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24414858635640996,
+      "learning_rate": 0.00018932115203415631,
+      "loss": 0.9642,
+      "step": 3993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.24286855215872527,
+      "learning_rate": 0.0001893140378926276,
+      "loss": 1.0458,
+      "step": 3994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2846207625112956,
+      "learning_rate": 0.0001893069215159589,
+      "loss": 1.1316,
+      "step": 3995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28248248879909615,
+      "learning_rate": 0.00018929980290432842,
+      "loss": 1.1581,
+      "step": 3996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27218202374232026,
+      "learning_rate": 0.00018929268205791422,
+      "loss": 1.2227,
+      "step": 3997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26852560065398273,
+      "learning_rate": 0.00018928555897689456,
+      "loss": 0.9302,
+      "step": 3998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2776665044902241,
+      "learning_rate": 0.00018927843366144765,
+      "loss": 1.0248,
+      "step": 3999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27797522607330477,
+      "learning_rate": 0.00018927130611175183,
+      "loss": 1.166,
+      "step": 4000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.23621425637102375,
+      "learning_rate": 0.00018926417632798547,
+      "loss": 1.1106,
+      "step": 4001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2926545007076596,
+      "learning_rate": 0.000189257044310327,
+      "loss": 1.1281,
+      "step": 4002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27712375321038807,
+      "learning_rate": 0.00018924991005895493,
+      "loss": 1.0626,
+      "step": 4003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.231003959100629,
+      "learning_rate": 0.0001892427735740477,
+      "loss": 0.9582,
+      "step": 4004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.29692646912091747,
+      "learning_rate": 0.00018923563485578405,
+      "loss": 1.0728,
+      "step": 4005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26325491425794023,
+      "learning_rate": 0.0001892284939043425,
+      "loss": 1.1163,
+      "step": 4006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27243932891523137,
+      "learning_rate": 0.00018922135071990185,
+      "loss": 1.0374,
+      "step": 4007
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28198690740674986,
+      "learning_rate": 0.0001892142053026408,
+      "loss": 1.1425,
+      "step": 4008
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28509269273338117,
+      "learning_rate": 0.00018920705765273818,
+      "loss": 1.3779,
+      "step": 4009
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28126037107058494,
+      "learning_rate": 0.0001891999077703729,
+      "loss": 1.1616,
+      "step": 4010
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3282196814898794,
+      "learning_rate": 0.00018919275565572387,
+      "loss": 1.1184,
+      "step": 4011
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2796592489259632,
+      "learning_rate": 0.00018918560130897006,
+      "loss": 1.1493,
+      "step": 4012
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35705040515515696,
+      "learning_rate": 0.00018917844473029054,
+      "loss": 1.035,
+      "step": 4013
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26954678750551575,
+      "learning_rate": 0.00018917128591986439,
+      "loss": 1.1217,
+      "step": 4014
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.268134645202258,
+      "learning_rate": 0.00018916412487787076,
+      "loss": 1.1415,
+      "step": 4015
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2869059780412037,
+      "learning_rate": 0.0001891569616044889,
+      "loss": 1.1142,
+      "step": 4016
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2866248425276922,
+      "learning_rate": 0.000189149796099898,
+      "loss": 0.968,
+      "step": 4017
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2854847382998254,
+      "learning_rate": 0.00018914262836427744,
+      "loss": 1.1804,
+      "step": 4018
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.28804702455957,
+      "learning_rate": 0.00018913545839780658,
+      "loss": 0.9394,
+      "step": 4019
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.26449776920625534,
+      "learning_rate": 0.00018912828620066486,
+      "loss": 1.1066,
+      "step": 4020
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2946737668152781,
+      "learning_rate": 0.00018912111177303177,
+      "loss": 1.2112,
+      "step": 4021
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.2901645389196644,
+      "learning_rate": 0.00018911393511508685,
+      "loss": 1.039,
+      "step": 4022
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27790893974884995,
+      "learning_rate": 0.00018910675622700967,
+      "loss": 1.058,
+      "step": 4023
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.27176601010684964,
+      "learning_rate": 0.00018909957510897992,
+      "loss": 1.1212,
+      "step": 4024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27542708140447214,
+      "learning_rate": 0.00018909239176117732,
+      "loss": 1.0686,
+      "step": 4025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27230675704373336,
+      "learning_rate": 0.0001890852061837816,
+      "loss": 1.1652,
+      "step": 4026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2557788458829725,
+      "learning_rate": 0.00018907801837697265,
+      "loss": 0.955,
+      "step": 4027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2601254941936526,
+      "learning_rate": 0.00018907082834093028,
+      "loss": 1.0526,
+      "step": 4028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.266059316761016,
+      "learning_rate": 0.00018906363607583445,
+      "loss": 1.0664,
+      "step": 4029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26037123171682397,
+      "learning_rate": 0.00018905644158186515,
+      "loss": 1.027,
+      "step": 4030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2563328237402112,
+      "learning_rate": 0.00018904924485920247,
+      "loss": 1.1243,
+      "step": 4031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28925250146928455,
+      "learning_rate": 0.0001890420459080264,
+      "loss": 1.1726,
+      "step": 4032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28975657876816185,
+      "learning_rate": 0.0001890348447285172,
+      "loss": 1.0362,
+      "step": 4033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.266527827382827,
+      "learning_rate": 0.00018902764132085507,
+      "loss": 1.1282,
+      "step": 4034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2819868839387067,
+      "learning_rate": 0.00018902043568522027,
+      "loss": 1.0481,
+      "step": 4035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26260787323629486,
+      "learning_rate": 0.0001890132278217931,
+      "loss": 1.0494,
+      "step": 4036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29231684987402756,
+      "learning_rate": 0.00018900601773075396,
+      "loss": 1.0668,
+      "step": 4037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2689538216980018,
+      "learning_rate": 0.00018899880541228332,
+      "loss": 1.0549,
+      "step": 4038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2748948697066311,
+      "learning_rate": 0.0001889915908665616,
+      "loss": 1.0691,
+      "step": 4039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26070699821281274,
+      "learning_rate": 0.00018898437409376942,
+      "loss": 1.0906,
+      "step": 4040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29273143034067534,
+      "learning_rate": 0.00018897715509408734,
+      "loss": 1.0254,
+      "step": 4041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3032571667610144,
+      "learning_rate": 0.00018896993386769602,
+      "loss": 0.9515,
+      "step": 4042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2567507749894945,
+      "learning_rate": 0.0001889627104147762,
+      "loss": 1.054,
+      "step": 4043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3048904231170557,
+      "learning_rate": 0.00018895548473550866,
+      "loss": 1.0738,
+      "step": 4044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27093062547894164,
+      "learning_rate": 0.00018894825683007417,
+      "loss": 1.1361,
+      "step": 4045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2907476290882117,
+      "learning_rate": 0.00018894102669865368,
+      "loss": 1.2182,
+      "step": 4046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27932158114529015,
+      "learning_rate": 0.0001889337943414281,
+      "loss": 1.0292,
+      "step": 4047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.303163185427342,
+      "learning_rate": 0.00018892655975857842,
+      "loss": 1.0809,
+      "step": 4048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28389248942158896,
+      "learning_rate": 0.0001889193229502857,
+      "loss": 0.9802,
+      "step": 4049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.23743556853791362,
+      "learning_rate": 0.000188912083916731,
+      "loss": 1.0415,
+      "step": 4050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2607618534899767,
+      "learning_rate": 0.00018890484265809558,
+      "loss": 1.1187,
+      "step": 4051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.282981189123625,
+      "learning_rate": 0.00018889759917456057,
+      "loss": 1.1592,
+      "step": 4052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2875209522933082,
+      "learning_rate": 0.00018889035346630726,
+      "loss": 1.1722,
+      "step": 4053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28469142917045703,
+      "learning_rate": 0.000188883105533517,
+      "loss": 1.0714,
+      "step": 4054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27563667585065,
+      "learning_rate": 0.00018887585537637116,
+      "loss": 0.9172,
+      "step": 4055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2735767551268766,
+      "learning_rate": 0.00018886860299505118,
+      "loss": 1.1189,
+      "step": 4056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30194229419916335,
+      "learning_rate": 0.00018886134838973857,
+      "loss": 1.0886,
+      "step": 4057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2932088348927251,
+      "learning_rate": 0.00018885409156061488,
+      "loss": 1.115,
+      "step": 4058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27110810019888465,
+      "learning_rate": 0.00018884683250786167,
+      "loss": 1.0398,
+      "step": 4059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.33237664113893134,
+      "learning_rate": 0.00018883957123166066,
+      "loss": 1.1189,
+      "step": 4060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2507604291402785,
+      "learning_rate": 0.00018883230773219354,
+      "loss": 1.0053,
+      "step": 4061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2972413921137987,
+      "learning_rate": 0.00018882504200964207,
+      "loss": 1.1487,
+      "step": 4062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26170579658272336,
+      "learning_rate": 0.00018881777406418816,
+      "loss": 1.1655,
+      "step": 4063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30453124626017863,
+      "learning_rate": 0.00018881050389601357,
+      "loss": 0.9793,
+      "step": 4064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27274778300480246,
+      "learning_rate": 0.00018880323150530034,
+      "loss": 1.019,
+      "step": 4065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2597880316659366,
+      "learning_rate": 0.0001887959568922304,
+      "loss": 1.13,
+      "step": 4066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30696703463261027,
+      "learning_rate": 0.00018878868005698586,
+      "loss": 1.068,
+      "step": 4067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.254482856569276,
+      "learning_rate": 0.0001887814009997488,
+      "loss": 1.156,
+      "step": 4068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29100192996613405,
+      "learning_rate": 0.00018877411972070135,
+      "loss": 1.1195,
+      "step": 4069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2592354961216204,
+      "learning_rate": 0.0001887668362200258,
+      "loss": 1.1087,
+      "step": 4070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28232033919806615,
+      "learning_rate": 0.00018875955049790438,
+      "loss": 1.0374,
+      "step": 4071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2710379764058582,
+      "learning_rate": 0.00018875226255451942,
+      "loss": 1.0692,
+      "step": 4072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3211139224345676,
+      "learning_rate": 0.00018874497239005332,
+      "loss": 1.0422,
+      "step": 4073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26011924754300747,
+      "learning_rate": 0.0001887376800046885,
+      "loss": 1.0197,
+      "step": 4074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2596113402217534,
+      "learning_rate": 0.00018873038539860747,
+      "loss": 1.0965,
+      "step": 4075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2666964093730143,
+      "learning_rate": 0.0001887230885719928,
+      "loss": 1.1003,
+      "step": 4076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24755453906410413,
+      "learning_rate": 0.00018871578952502703,
+      "loss": 1.0753,
+      "step": 4077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25330256715469024,
+      "learning_rate": 0.0001887084882578929,
+      "loss": 1.0962,
+      "step": 4078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2664417858327983,
+      "learning_rate": 0.00018870118477077309,
+      "loss": 1.1026,
+      "step": 4079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2887769686304296,
+      "learning_rate": 0.00018869387906385044,
+      "loss": 1.085,
+      "step": 4080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2945622145717073,
+      "learning_rate": 0.00018868657113730764,
+      "loss": 1.0453,
+      "step": 4081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2914962906932826,
+      "learning_rate": 0.0001886792609913277,
+      "loss": 1.1669,
+      "step": 4082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2982094102526812,
+      "learning_rate": 0.00018867194862609354,
+      "loss": 1.1398,
+      "step": 4083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2943744623045212,
+      "learning_rate": 0.0001886646340417881,
+      "loss": 1.1049,
+      "step": 4084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2808334985745818,
+      "learning_rate": 0.0001886573172385945,
+      "loss": 1.0434,
+      "step": 4085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28765243366041743,
+      "learning_rate": 0.0001886499982166958,
+      "loss": 1.1537,
+      "step": 4086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2838538473305344,
+      "learning_rate": 0.0001886426769762752,
+      "loss": 1.1621,
+      "step": 4087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2779123375968445,
+      "learning_rate": 0.00018863535351751586,
+      "loss": 1.0426,
+      "step": 4088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26234781180349137,
+      "learning_rate": 0.00018862802784060115,
+      "loss": 1.1284,
+      "step": 4089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27792379078088625,
+      "learning_rate": 0.00018862069994571428,
+      "loss": 0.9789,
+      "step": 4090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3213818567797798,
+      "learning_rate": 0.00018861336983303875,
+      "loss": 1.0857,
+      "step": 4091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2942679485146761,
+      "learning_rate": 0.0001886060375027579,
+      "loss": 1.0244,
+      "step": 4092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2934982143646054,
+      "learning_rate": 0.0001885987029550553,
+      "loss": 1.0908,
+      "step": 4093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2705085832069158,
+      "learning_rate": 0.00018859136619011447,
+      "loss": 1.0971,
+      "step": 4094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2645736238106595,
+      "learning_rate": 0.00018858402720811905,
+      "loss": 1.0806,
+      "step": 4095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25805002703777125,
+      "learning_rate": 0.00018857668600925264,
+      "loss": 1.0815,
+      "step": 4096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2768106728894951,
+      "learning_rate": 0.00018856934259369902,
+      "loss": 1.0963,
+      "step": 4097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2952517992902994,
+      "learning_rate": 0.00018856199696164194,
+      "loss": 1.0913,
+      "step": 4098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.25015646961672283,
+      "learning_rate": 0.0001885546491132652,
+      "loss": 1.0521,
+      "step": 4099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2933146156325368,
+      "learning_rate": 0.00018854729904875273,
+      "loss": 1.1188,
+      "step": 4100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2906550664804429,
+      "learning_rate": 0.00018853994676828846,
+      "loss": 1.1433,
+      "step": 4101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.26884740955424163,
+      "learning_rate": 0.00018853259227205634,
+      "loss": 1.1072,
+      "step": 4102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3113311790033681,
+      "learning_rate": 0.0001885252355602405,
+      "loss": 1.1282,
+      "step": 4103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.28155446924580935,
+      "learning_rate": 0.00018851787663302498,
+      "loss": 1.083,
+      "step": 4104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29809592194674794,
+      "learning_rate": 0.00018851051549059397,
+      "loss": 1.0569,
+      "step": 4105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.24384181868224059,
+      "learning_rate": 0.0001885031521331317,
+      "loss": 1.0561,
+      "step": 4106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27108319846787104,
+      "learning_rate": 0.0001884957865608224,
+      "loss": 1.058,
+      "step": 4107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2943767607152993,
+      "learning_rate": 0.00018848841877385045,
+      "loss": 1.1393,
+      "step": 4108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29474696198802675,
+      "learning_rate": 0.00018848104877240015,
+      "loss": 1.1804,
+      "step": 4109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2951250932743141,
+      "learning_rate": 0.00018847367655665606,
+      "loss": 1.2261,
+      "step": 4110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2608593829479291,
+      "learning_rate": 0.0001884663021268026,
+      "loss": 1.0769,
+      "step": 4111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29600065532000475,
+      "learning_rate": 0.0001884589254830243,
+      "loss": 1.0772,
+      "step": 4112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2990313407932539,
+      "learning_rate": 0.0001884515466255058,
+      "loss": 1.1775,
+      "step": 4113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2843669340987528,
+      "learning_rate": 0.00018844416555443178,
+      "loss": 1.1374,
+      "step": 4114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2508179755628321,
+      "learning_rate": 0.00018843678226998693,
+      "loss": 1.1355,
+      "step": 4115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2754853279028242,
+      "learning_rate": 0.000188429396772356,
+      "loss": 1.0528,
+      "step": 4116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2692652041119292,
+      "learning_rate": 0.00018842200906172386,
+      "loss": 1.081,
+      "step": 4117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2581020541470848,
+      "learning_rate": 0.00018841461913827537,
+      "loss": 1.1674,
+      "step": 4118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.30950720631732437,
+      "learning_rate": 0.0001884072270021955,
+      "loss": 1.1665,
+      "step": 4119
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2485313737562046,
+      "learning_rate": 0.00018839983265366917,
+      "loss": 1.1825,
+      "step": 4120
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2644284997612591,
+      "learning_rate": 0.0001883924360928815,
+      "loss": 1.1264,
+      "step": 4121
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.29112561286000443,
+      "learning_rate": 0.0001883850373200175,
+      "loss": 1.1339,
+      "step": 4122
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27090825993266066,
+      "learning_rate": 0.00018837763633526247,
+      "loss": 1.1998,
+      "step": 4123
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3146596625378274,
+      "learning_rate": 0.0001883702331388015,
+      "loss": 1.1325,
+      "step": 4124
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.31943559497066387,
+      "learning_rate": 0.00018836282773081992,
+      "loss": 1.0685,
+      "step": 4125
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3067968110428308,
+      "learning_rate": 0.00018835542011150303,
+      "loss": 1.1561,
+      "step": 4126
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.27469103129370837,
+      "learning_rate": 0.00018834801028103627,
+      "loss": 1.0606,
+      "step": 4127
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.2565388108324173,
+      "learning_rate": 0.00018834059823960497,
+      "loss": 1.1264,
+      "step": 4128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26620235976121565,
+      "learning_rate": 0.0001883331839873947,
+      "loss": 0.9498,
+      "step": 4129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2736769679639783,
+      "learning_rate": 0.00018832576752459099,
+      "loss": 1.1824,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28039373078224655,
+      "learning_rate": 0.00018831834885137943,
+      "loss": 1.0567,
+      "step": 4131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28360004650576354,
+      "learning_rate": 0.00018831092796794572,
+      "loss": 1.1355,
+      "step": 4132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3007266751162003,
+      "learning_rate": 0.0001883035048744755,
+      "loss": 1.1805,
+      "step": 4133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27426609206190533,
+      "learning_rate": 0.00018829607957115458,
+      "loss": 1.1591,
+      "step": 4134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2975062542878613,
+      "learning_rate": 0.00018828865205816877,
+      "loss": 1.1326,
+      "step": 4135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24138546192014423,
+      "learning_rate": 0.00018828122233570396,
+      "loss": 1.0062,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3259428242836633,
+      "learning_rate": 0.00018827379040394607,
+      "loss": 1.0638,
+      "step": 4137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2893519126284774,
+      "learning_rate": 0.00018826635626308113,
+      "loss": 1.0938,
+      "step": 4138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2649524956937564,
+      "learning_rate": 0.00018825891991329513,
+      "loss": 1.0869,
+      "step": 4139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28942290188391856,
+      "learning_rate": 0.00018825148135477417,
+      "loss": 0.9653,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.303875931112067,
+      "learning_rate": 0.00018824404058770443,
+      "loss": 1.1875,
+      "step": 4141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27790521276702573,
+      "learning_rate": 0.00018823659761227216,
+      "loss": 1.1442,
+      "step": 4142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2499769206341798,
+      "learning_rate": 0.00018822915242866354,
+      "loss": 0.9754,
+      "step": 4143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30156640795095413,
+      "learning_rate": 0.00018822170503706494,
+      "loss": 1.1384,
+      "step": 4144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2551627106250161,
+      "learning_rate": 0.00018821425543766275,
+      "loss": 0.9835,
+      "step": 4145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2699256367377609,
+      "learning_rate": 0.00018820680363064335,
+      "loss": 1.0811,
+      "step": 4146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2685887519613943,
+      "learning_rate": 0.00018819934961619323,
+      "loss": 1.1884,
+      "step": 4147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27522442156799526,
+      "learning_rate": 0.000188191893394499,
+      "loss": 1.1393,
+      "step": 4148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25866933190885216,
+      "learning_rate": 0.0001881844349657472,
+      "loss": 1.0642,
+      "step": 4149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2767798366498473,
+      "learning_rate": 0.00018817697433012447,
+      "loss": 1.1294,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25741272329454395,
+      "learning_rate": 0.00018816951148781756,
+      "loss": 1.2252,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2703985849781623,
+      "learning_rate": 0.00018816204643901322,
+      "loss": 1.0734,
+      "step": 4152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3049033588778724,
+      "learning_rate": 0.00018815457918389822,
+      "loss": 1.053,
+      "step": 4153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.22456734765420605,
+      "learning_rate": 0.00018814710972265953,
+      "loss": 0.9969,
+      "step": 4154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2773921790872893,
+      "learning_rate": 0.00018813963805548397,
+      "loss": 1.0376,
+      "step": 4155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34700507029136146,
+      "learning_rate": 0.0001881321641825586,
+      "loss": 1.1532,
+      "step": 4156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3113422889257847,
+      "learning_rate": 0.00018812468810407043,
+      "loss": 1.0595,
+      "step": 4157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29393485580712975,
+      "learning_rate": 0.00018811720982020655,
+      "loss": 1.1448,
+      "step": 4158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2958948874029899,
+      "learning_rate": 0.00018810972933115412,
+      "loss": 1.1131,
+      "step": 4159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2465662069981653,
+      "learning_rate": 0.00018810224663710033,
+      "loss": 0.8469,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2876082268093369,
+      "learning_rate": 0.00018809476173823247,
+      "loss": 1.0811,
+      "step": 4161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3097135244404772,
+      "learning_rate": 0.0001880872746347378,
+      "loss": 1.1378,
+      "step": 4162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2793139934409337,
+      "learning_rate": 0.00018807978532680374,
+      "loss": 1.0819,
+      "step": 4163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2657493887824555,
+      "learning_rate": 0.0001880722938146177,
+      "loss": 1.1998,
+      "step": 4164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2504323188479509,
+      "learning_rate": 0.00018806480009836716,
+      "loss": 1.136,
+      "step": 4165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2770886792479756,
+      "learning_rate": 0.00018805730417823964,
+      "loss": 1.0626,
+      "step": 4166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28327442888376786,
+      "learning_rate": 0.00018804980605442273,
+      "loss": 1.2524,
+      "step": 4167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2738803155422817,
+      "learning_rate": 0.00018804230572710411,
+      "loss": 1.2178,
+      "step": 4168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2621464862724536,
+      "learning_rate": 0.00018803480319647145,
+      "loss": 1.1003,
+      "step": 4169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2829875367760781,
+      "learning_rate": 0.0001880272984627125,
+      "loss": 1.1652,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25110962790800634,
+      "learning_rate": 0.00018801979152601508,
+      "loss": 1.0741,
+      "step": 4171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3021846850869496,
+      "learning_rate": 0.0001880122823865671,
+      "loss": 1.1427,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25970989561092755,
+      "learning_rate": 0.00018800477104455638,
+      "loss": 1.1105,
+      "step": 4173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3037823815880924,
+      "learning_rate": 0.00018799725750017098,
+      "loss": 1.04,
+      "step": 4174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28340725697761254,
+      "learning_rate": 0.00018798974175359892,
+      "loss": 1.0287,
+      "step": 4175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2907181694596812,
+      "learning_rate": 0.00018798222380502825,
+      "loss": 1.1459,
+      "step": 4176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2798317614131016,
+      "learning_rate": 0.00018797470365464718,
+      "loss": 1.0269,
+      "step": 4177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2451699568307834,
+      "learning_rate": 0.0001879671813026438,
+      "loss": 1.0953,
+      "step": 4178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26115132881368164,
+      "learning_rate": 0.00018795965674920647,
+      "loss": 0.95,
+      "step": 4179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28240121314882904,
+      "learning_rate": 0.00018795212999452344,
+      "loss": 1.0437,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.30408179576379274,
+      "learning_rate": 0.00018794460103878306,
+      "loss": 1.1509,
+      "step": 4181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2613648521070335,
+      "learning_rate": 0.00018793706988217378,
+      "loss": 1.0649,
+      "step": 4182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2807710295201798,
+      "learning_rate": 0.00018792953652488405,
+      "loss": 1.2016,
+      "step": 4183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27595401782163975,
+      "learning_rate": 0.0001879220009671024,
+      "loss": 1.12,
+      "step": 4184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28286269456870644,
+      "learning_rate": 0.00018791446320901747,
+      "loss": 1.0609,
+      "step": 4185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28563194802618863,
+      "learning_rate": 0.0001879069232508178,
+      "loss": 1.0778,
+      "step": 4186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.270002889739277,
+      "learning_rate": 0.00018789938109269215,
+      "loss": 1.0788,
+      "step": 4187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25196325847643697,
+      "learning_rate": 0.00018789183673482924,
+      "loss": 1.1108,
+      "step": 4188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25601666120815847,
+      "learning_rate": 0.00018788429017741785,
+      "loss": 1.043,
+      "step": 4189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2743217941272287,
+      "learning_rate": 0.0001878767414206469,
+      "loss": 1.0379,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2971315421501434,
+      "learning_rate": 0.00018786919046470527,
+      "loss": 1.1175,
+      "step": 4191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24701289964452056,
+      "learning_rate": 0.0001878616373097819,
+      "loss": 1.047,
+      "step": 4192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28093581228292824,
+      "learning_rate": 0.00018785408195606587,
+      "loss": 1.1656,
+      "step": 4193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2738956970254292,
+      "learning_rate": 0.0001878465244037462,
+      "loss": 1.0897,
+      "step": 4194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3300931758890875,
+      "learning_rate": 0.00018783896465301205,
+      "loss": 1.1008,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2339082478589511,
+      "learning_rate": 0.0001878314027040526,
+      "loss": 1.1647,
+      "step": 4196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2575164996133438,
+      "learning_rate": 0.0001878238385570571,
+      "loss": 1.1316,
+      "step": 4197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28065885782328104,
+      "learning_rate": 0.00018781627221221484,
+      "loss": 1.1204,
+      "step": 4198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27240538013841803,
+      "learning_rate": 0.0001878087036697152,
+      "loss": 1.1995,
+      "step": 4199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28177346756683214,
+      "learning_rate": 0.00018780113292974756,
+      "loss": 1.0969,
+      "step": 4200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2869344753327611,
+      "learning_rate": 0.00018779355999250135,
+      "loss": 1.02,
+      "step": 4201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.317258761790818,
+      "learning_rate": 0.00018778598485816618,
+      "loss": 1.0979,
+      "step": 4202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28088629903560547,
+      "learning_rate": 0.00018777840752693152,
+      "loss": 1.0249,
+      "step": 4203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2796415407726609,
+      "learning_rate": 0.00018777082799898705,
+      "loss": 1.2466,
+      "step": 4204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28573228367442755,
+      "learning_rate": 0.00018776324627452247,
+      "loss": 1.1008,
+      "step": 4205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.36653558077222376,
+      "learning_rate": 0.0001877556623537275,
+      "loss": 1.1889,
+      "step": 4206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.257060176220037,
+      "learning_rate": 0.00018774807623679192,
+      "loss": 1.096,
+      "step": 4207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26689142065335364,
+      "learning_rate": 0.00018774048792390559,
+      "loss": 1.1396,
+      "step": 4208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27797803343333893,
+      "learning_rate": 0.0001877328974152584,
+      "loss": 1.0059,
+      "step": 4209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26890403496189735,
+      "learning_rate": 0.00018772530471104028,
+      "loss": 1.1074,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29380901154741923,
+      "learning_rate": 0.00018771770981144132,
+      "loss": 1.1014,
+      "step": 4211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2777299581369351,
+      "learning_rate": 0.00018771011271665153,
+      "loss": 1.0938,
+      "step": 4212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26403372932978386,
+      "learning_rate": 0.00018770251342686104,
+      "loss": 1.1595,
+      "step": 4213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28636002615157274,
+      "learning_rate": 0.00018769491194226006,
+      "loss": 1.0455,
+      "step": 4214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2844974569194305,
+      "learning_rate": 0.00018768730826303876,
+      "loss": 1.0772,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25117753184917296,
+      "learning_rate": 0.0001876797023893875,
+      "loss": 1.1215,
+      "step": 4216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.25290828237696705,
+      "learning_rate": 0.00018767209432149652,
+      "loss": 1.0751,
+      "step": 4217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2641125806261801,
+      "learning_rate": 0.0001876644840595563,
+      "loss": 1.1248,
+      "step": 4218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.26789458127024735,
+      "learning_rate": 0.00018765687160375732,
+      "loss": 1.058,
+      "step": 4219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2656234103720025,
+      "learning_rate": 0.00018764925695428998,
+      "loss": 1.0255,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.24975067381842872,
+      "learning_rate": 0.00018764164011134495,
+      "loss": 1.1212,
+      "step": 4221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2395528550635482,
+      "learning_rate": 0.00018763402107511276,
+      "loss": 0.9989,
+      "step": 4222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2532908262725297,
+      "learning_rate": 0.00018762639984578412,
+      "loss": 1.1901,
+      "step": 4223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31719585665339967,
+      "learning_rate": 0.00018761877642354977,
+      "loss": 1.2269,
+      "step": 4224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.31874436060610234,
+      "learning_rate": 0.00018761115080860046,
+      "loss": 0.9891,
+      "step": 4225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.29050294955105876,
+      "learning_rate": 0.00018760352300112705,
+      "loss": 1.1358,
+      "step": 4226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27399249257978747,
+      "learning_rate": 0.00018759589300132041,
+      "loss": 1.0868,
+      "step": 4227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2696276778720872,
+      "learning_rate": 0.00018758826080937148,
+      "loss": 1.1893,
+      "step": 4228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.28473533529326145,
+      "learning_rate": 0.00018758062642547133,
+      "loss": 1.1131,
+      "step": 4229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.27137640986272094,
+      "learning_rate": 0.00018757298984981092,
+      "loss": 0.976,
+      "step": 4230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3035287993677886,
+      "learning_rate": 0.0001875653510825814,
+      "loss": 1.2036,
+      "step": 4231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2730310702863054,
+      "learning_rate": 0.00018755771012397393,
+      "loss": 1.1172,
+      "step": 4232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.2515152904223668,
+      "learning_rate": 0.00018755006697417976,
+      "loss": 1.0117,
+      "step": 4233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2628436719293594,
+      "learning_rate": 0.00018754242163339014,
+      "loss": 1.0782,
+      "step": 4234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26021326272006884,
+      "learning_rate": 0.0001875347741017964,
+      "loss": 1.1058,
+      "step": 4235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2739556248198109,
+      "learning_rate": 0.0001875271243795899,
+      "loss": 0.9982,
+      "step": 4236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30134796736592306,
+      "learning_rate": 0.00018751947246696212,
+      "loss": 1.0949,
+      "step": 4237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.271764675869869,
+      "learning_rate": 0.00018751181836410455,
+      "loss": 1.0182,
+      "step": 4238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24250081027639273,
+      "learning_rate": 0.0001875041620712087,
+      "loss": 1.0884,
+      "step": 4239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2869641568770455,
+      "learning_rate": 0.0001874965035884662,
+      "loss": 1.1567,
+      "step": 4240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2969381271520231,
+      "learning_rate": 0.00018748884291606874,
+      "loss": 1.0765,
+      "step": 4241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26178640355726157,
+      "learning_rate": 0.00018748118005420798,
+      "loss": 1.1142,
+      "step": 4242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2800257202847653,
+      "learning_rate": 0.0001874735150030757,
+      "loss": 1.1046,
+      "step": 4243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2944863333696688,
+      "learning_rate": 0.00018746584776286376,
+      "loss": 1.038,
+      "step": 4244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.30216624143418275,
+      "learning_rate": 0.00018745817833376398,
+      "loss": 1.2037,
+      "step": 4245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2663185997548643,
+      "learning_rate": 0.00018745050671596834,
+      "loss": 1.1279,
+      "step": 4246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2447130737177253,
+      "learning_rate": 0.00018744283290966882,
+      "loss": 1.0917,
+      "step": 4247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.290268659831609,
+      "learning_rate": 0.00018743515691505743,
+      "loss": 1.031,
+      "step": 4248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28648517218701464,
+      "learning_rate": 0.0001874274787323263,
+      "loss": 1.0971,
+      "step": 4249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2682660047938872,
+      "learning_rate": 0.00018741979836166755,
+      "loss": 1.1424,
+      "step": 4250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3030211943382245,
+      "learning_rate": 0.00018741211580327344,
+      "loss": 1.254,
+      "step": 4251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3098473572044177,
+      "learning_rate": 0.00018740443105733613,
+      "loss": 0.9803,
+      "step": 4252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2797263417102469,
+      "learning_rate": 0.00018739674412404807,
+      "loss": 1.0723,
+      "step": 4253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27611130556479674,
+      "learning_rate": 0.00018738905500360154,
+      "loss": 1.0926,
+      "step": 4254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2787572653472917,
+      "learning_rate": 0.00018738136369618897,
+      "loss": 1.0504,
+      "step": 4255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28676439879084453,
+      "learning_rate": 0.00018737367020200285,
+      "loss": 1.0677,
+      "step": 4256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2898408979989211,
+      "learning_rate": 0.00018736597452123575,
+      "loss": 1.0629,
+      "step": 4257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.265340677178945,
+      "learning_rate": 0.00018735827665408022,
+      "loss": 1.0293,
+      "step": 4258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2910287914202519,
+      "learning_rate": 0.0001873505766007289,
+      "loss": 0.9698,
+      "step": 4259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2899174476618089,
+      "learning_rate": 0.00018734287436137452,
+      "loss": 1.1953,
+      "step": 4260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2566780706363218,
+      "learning_rate": 0.0001873351699362098,
+      "loss": 1.0803,
+      "step": 4261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25026487846526846,
+      "learning_rate": 0.00018732746332542758,
+      "loss": 1.0512,
+      "step": 4262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2634865581228525,
+      "learning_rate": 0.00018731975452922073,
+      "loss": 1.0902,
+      "step": 4263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2739235925675129,
+      "learning_rate": 0.0001873120435477821,
+      "loss": 1.2044,
+      "step": 4264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3392678199356469,
+      "learning_rate": 0.00018730433038130473,
+      "loss": 1.1055,
+      "step": 4265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2394900592357321,
+      "learning_rate": 0.0001872966150299816,
+      "loss": 1.0701,
+      "step": 4266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2700640038637131,
+      "learning_rate": 0.00018728889749400584,
+      "loss": 1.0162,
+      "step": 4267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2912889000502306,
+      "learning_rate": 0.00018728117777357055,
+      "loss": 1.1419,
+      "step": 4268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2760619669010758,
+      "learning_rate": 0.00018727345586886892,
+      "loss": 1.0408,
+      "step": 4269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2927376911193287,
+      "learning_rate": 0.00018726573178009422,
+      "loss": 1.0906,
+      "step": 4270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25519561721394524,
+      "learning_rate": 0.00018725800550743976,
+      "loss": 1.0385,
+      "step": 4271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2830277309204389,
+      "learning_rate": 0.00018725027705109886,
+      "loss": 1.0272,
+      "step": 4272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2664819202376281,
+      "learning_rate": 0.00018724254641126493,
+      "loss": 1.1779,
+      "step": 4273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2574469843738646,
+      "learning_rate": 0.00018723481358813145,
+      "loss": 1.0864,
+      "step": 4274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2757193054824571,
+      "learning_rate": 0.00018722707858189193,
+      "loss": 1.0145,
+      "step": 4275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29257610913040466,
+      "learning_rate": 0.00018721934139273998,
+      "loss": 0.9679,
+      "step": 4276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.289091197361481,
+      "learning_rate": 0.00018721160202086914,
+      "loss": 1.092,
+      "step": 4277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24344559699327611,
+      "learning_rate": 0.0001872038604664732,
+      "loss": 1.1082,
+      "step": 4278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2810770812752845,
+      "learning_rate": 0.0001871961167297458,
+      "loss": 1.0673,
+      "step": 4279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2861371909971846,
+      "learning_rate": 0.0001871883708108808,
+      "loss": 1.0974,
+      "step": 4280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.31326751338971903,
+      "learning_rate": 0.00018718062271007204,
+      "loss": 1.1654,
+      "step": 4281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2695673863745955,
+      "learning_rate": 0.00018717287242751341,
+      "loss": 1.045,
+      "step": 4282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3078415478161515,
+      "learning_rate": 0.00018716511996339885,
+      "loss": 1.0805,
+      "step": 4283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26900716775364764,
+      "learning_rate": 0.00018715736531792237,
+      "loss": 1.1007,
+      "step": 4284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28023295676688076,
+      "learning_rate": 0.00018714960849127804,
+      "loss": 1.0815,
+      "step": 4285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2898989628594484,
+      "learning_rate": 0.00018714184948366,
+      "loss": 1.0418,
+      "step": 4286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26061571588826044,
+      "learning_rate": 0.00018713408829526242,
+      "loss": 1.1429,
+      "step": 4287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2745641023863704,
+      "learning_rate": 0.0001871263249262795,
+      "loss": 1.1773,
+      "step": 4288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26470950342417765,
+      "learning_rate": 0.00018711855937690556,
+      "loss": 1.1311,
+      "step": 4289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29468498976148555,
+      "learning_rate": 0.00018711079164733491,
+      "loss": 1.0414,
+      "step": 4290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26226927485345436,
+      "learning_rate": 0.00018710302173776194,
+      "loss": 1.0876,
+      "step": 4291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24393119618376538,
+      "learning_rate": 0.00018709524964838115,
+      "loss": 1.0936,
+      "step": 4292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2871335693752671,
+      "learning_rate": 0.00018708747537938696,
+      "loss": 1.0862,
+      "step": 4293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25454898572621976,
+      "learning_rate": 0.00018707969893097399,
+      "loss": 0.9553,
+      "step": 4294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2612510649191512,
+      "learning_rate": 0.0001870719203033368,
+      "loss": 1.015,
+      "step": 4295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2735824234461878,
+      "learning_rate": 0.0001870641394966701,
+      "loss": 1.1366,
+      "step": 4296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2561249258130775,
+      "learning_rate": 0.00018705635651116857,
+      "loss": 0.9771,
+      "step": 4297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2724097621208985,
+      "learning_rate": 0.00018704857134702705,
+      "loss": 1.0373,
+      "step": 4298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2624090546858682,
+      "learning_rate": 0.00018704078400444028,
+      "loss": 1.0986,
+      "step": 4299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2933893818570987,
+      "learning_rate": 0.0001870329944836032,
+      "loss": 1.0659,
+      "step": 4300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25403733960977537,
+      "learning_rate": 0.00018702520278471074,
+      "loss": 1.1726,
+      "step": 4301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27542609220308545,
+      "learning_rate": 0.00018701740890795788,
+      "loss": 1.1237,
+      "step": 4302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2536691041486995,
+      "learning_rate": 0.0001870096128535397,
+      "loss": 0.9981,
+      "step": 4303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2732550236694586,
+      "learning_rate": 0.00018700181462165126,
+      "loss": 1.0504,
+      "step": 4304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26940609611280475,
+      "learning_rate": 0.0001869940142124877,
+      "loss": 1.0977,
+      "step": 4305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2912691081559738,
+      "learning_rate": 0.0001869862116262443,
+      "loss": 1.0703,
+      "step": 4306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27335596729191,
+      "learning_rate": 0.00018697840686311628,
+      "loss": 1.1206,
+      "step": 4307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2882836785046496,
+      "learning_rate": 0.00018697059992329895,
+      "loss": 1.0942,
+      "step": 4308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2736335086651974,
+      "learning_rate": 0.0001869627908069877,
+      "loss": 1.09,
+      "step": 4309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.25546265700852766,
+      "learning_rate": 0.00018695497951437795,
+      "loss": 1.1653,
+      "step": 4310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27099441480074316,
+      "learning_rate": 0.0001869471660456652,
+      "loss": 1.0797,
+      "step": 4311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2734354890739324,
+      "learning_rate": 0.00018693935040104497,
+      "loss": 1.0753,
+      "step": 4312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3009809400535224,
+      "learning_rate": 0.00018693153258071286,
+      "loss": 1.1689,
+      "step": 4313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29754474015094556,
+      "learning_rate": 0.00018692371258486451,
+      "loss": 1.0379,
+      "step": 4314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2996282403269194,
+      "learning_rate": 0.00018691589041369564,
+      "loss": 1.1081,
+      "step": 4315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24314970149389548,
+      "learning_rate": 0.000186908066067402,
+      "loss": 1.0464,
+      "step": 4316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24701089925472527,
+      "learning_rate": 0.00018690023954617932,
+      "loss": 1.1741,
+      "step": 4317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27607286639486517,
+      "learning_rate": 0.0001868924108502236,
+      "loss": 1.0594,
+      "step": 4318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3017023954760616,
+      "learning_rate": 0.00018688457997973065,
+      "loss": 1.0294,
+      "step": 4319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.24937940795309205,
+      "learning_rate": 0.00018687674693489647,
+      "loss": 1.0421,
+      "step": 4320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.23892217344372102,
+      "learning_rate": 0.00018686891171591712,
+      "loss": 1.0908,
+      "step": 4321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28068063323508174,
+      "learning_rate": 0.00018686107432298868,
+      "loss": 1.1319,
+      "step": 4322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.26786810608469297,
+      "learning_rate": 0.00018685323475630723,
+      "loss": 1.0355,
+      "step": 4323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2435421171517636,
+      "learning_rate": 0.000186845393016069,
+      "loss": 1.0141,
+      "step": 4324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.27884414310128647,
+      "learning_rate": 0.00018683754910247025,
+      "loss": 1.1473,
+      "step": 4325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2890938734276742,
+      "learning_rate": 0.00018682970301570726,
+      "loss": 1.1203,
+      "step": 4326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2769615698559789,
+      "learning_rate": 0.00018682185475597636,
+      "loss": 1.0541,
+      "step": 4327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2785590604016616,
+      "learning_rate": 0.00018681400432347397,
+      "loss": 1.0544,
+      "step": 4328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29418313155446657,
+      "learning_rate": 0.00018680615171839658,
+      "loss": 1.1175,
+      "step": 4329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3025538424684769,
+      "learning_rate": 0.00018679829694094068,
+      "loss": 1.2113,
+      "step": 4330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2910730421005133,
+      "learning_rate": 0.00018679043999130288,
+      "loss": 1.132,
+      "step": 4331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.29253393692155966,
+      "learning_rate": 0.00018678258086967975,
+      "loss": 1.095,
+      "step": 4332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.28305224842813903,
+      "learning_rate": 0.00018677471957626797,
+      "loss": 1.1026,
+      "step": 4333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2924260897779223,
+      "learning_rate": 0.0001867668561112643,
+      "loss": 1.0386,
+      "step": 4334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.293768434817011,
+      "learning_rate": 0.00018675899047486557,
+      "loss": 1.0448,
+      "step": 4335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2941288478482527,
+      "learning_rate": 0.00018675112266726854,
+      "loss": 1.0662,
+      "step": 4336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.2957044806998864,
+      "learning_rate": 0.00018674325268867016,
+      "loss": 0.9872,
+      "step": 4337
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3170426741111006,
+      "learning_rate": 0.00018673538053926735,
+      "loss": 1.0532,
+      "step": 4338
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3000921754928047,
+      "learning_rate": 0.00018672750621925714,
+      "loss": 1.1413,
+      "step": 4339
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2998372677155844,
+      "learning_rate": 0.00018671962972883658,
+      "loss": 1.1425,
+      "step": 4340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2645728037896027,
+      "learning_rate": 0.00018671175106820277,
+      "loss": 1.0747,
+      "step": 4341
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28183953656999067,
+      "learning_rate": 0.00018670387023755295,
+      "loss": 0.9737,
+      "step": 4342
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.256299564446901,
+      "learning_rate": 0.00018669598723708422,
+      "loss": 1.1139,
+      "step": 4343
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26935203909823185,
+      "learning_rate": 0.00018668810206699395,
+      "loss": 1.0717,
+      "step": 4344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23352656548172643,
+      "learning_rate": 0.00018668021472747944,
+      "loss": 1.0427,
+      "step": 4345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2778191273219274,
+      "learning_rate": 0.00018667232521873807,
+      "loss": 1.0442,
+      "step": 4346
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2708092397068009,
+      "learning_rate": 0.00018666443354096733,
+      "loss": 1.1642,
+      "step": 4347
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2628086430153175,
+      "learning_rate": 0.00018665653969436466,
+      "loss": 1.0567,
+      "step": 4348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2689137058024899,
+      "learning_rate": 0.00018664864367912758,
+      "loss": 1.055,
+      "step": 4349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27522982747891467,
+      "learning_rate": 0.00018664074549545377,
+      "loss": 0.9796,
+      "step": 4350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2585842007123571,
+      "learning_rate": 0.00018663284514354084,
+      "loss": 1.0592,
+      "step": 4351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2698306914125685,
+      "learning_rate": 0.0001866249426235865,
+      "loss": 1.084,
+      "step": 4352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.30565305866824105,
+      "learning_rate": 0.00018661703793578855,
+      "loss": 1.0316,
+      "step": 4353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26114362713963013,
+      "learning_rate": 0.00018660913108034478,
+      "loss": 1.1677,
+      "step": 4354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3016235809178839,
+      "learning_rate": 0.00018660122205745313,
+      "loss": 1.1933,
+      "step": 4355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2863853851725708,
+      "learning_rate": 0.0001865933108673114,
+      "loss": 0.9809,
+      "step": 4356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2912386308622707,
+      "learning_rate": 0.00018658539751011767,
+      "loss": 1.1289,
+      "step": 4357
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2776875205740512,
+      "learning_rate": 0.00018657748198606995,
+      "loss": 1.0565,
+      "step": 4358
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2748712764200445,
+      "learning_rate": 0.00018656956429536633,
+      "loss": 1.0782,
+      "step": 4359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.255166953661312,
+      "learning_rate": 0.00018656164443820494,
+      "loss": 1.0291,
+      "step": 4360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2665620217061101,
+      "learning_rate": 0.00018655372241478403,
+      "loss": 1.114,
+      "step": 4361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2544779490108305,
+      "learning_rate": 0.00018654579822530179,
+      "loss": 1.2265,
+      "step": 4362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23854453003136591,
+      "learning_rate": 0.00018653787186995654,
+      "loss": 1.033,
+      "step": 4363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2710547109595493,
+      "learning_rate": 0.00018652994334894668,
+      "loss": 1.0567,
+      "step": 4364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27823107428416105,
+      "learning_rate": 0.00018652201266247063,
+      "loss": 1.0357,
+      "step": 4365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3065920295229727,
+      "learning_rate": 0.0001865140798107268,
+      "loss": 0.9394,
+      "step": 4366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29060026851055215,
+      "learning_rate": 0.00018650614479391378,
+      "loss": 1.0582,
+      "step": 4367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2798363947842491,
+      "learning_rate": 0.00018649820761223012,
+      "loss": 1.1008,
+      "step": 4368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2524281791152147,
+      "learning_rate": 0.00018649026826587442,
+      "loss": 1.0788,
+      "step": 4369
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3175783594422539,
+      "learning_rate": 0.00018648232675504543,
+      "loss": 1.0091,
+      "step": 4370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2605770620877395,
+      "learning_rate": 0.00018647438307994185,
+      "loss": 1.0315,
+      "step": 4371
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3074184313253669,
+      "learning_rate": 0.0001864664372407625,
+      "loss": 1.11,
+      "step": 4372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2804678028999845,
+      "learning_rate": 0.0001864584892377062,
+      "loss": 1.0521,
+      "step": 4373
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25285663187552027,
+      "learning_rate": 0.00018645053907097187,
+      "loss": 1.0407,
+      "step": 4374
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32876163956484417,
+      "learning_rate": 0.00018644258674075848,
+      "loss": 0.9872,
+      "step": 4375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25777775379310247,
+      "learning_rate": 0.000186434632247265,
+      "loss": 1.0567,
+      "step": 4376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29826590288671295,
+      "learning_rate": 0.00018642667559069055,
+      "loss": 1.0757,
+      "step": 4377
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27317079924980964,
+      "learning_rate": 0.0001864187167712342,
+      "loss": 1.1448,
+      "step": 4378
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3070229873978444,
+      "learning_rate": 0.00018641075578909518,
+      "loss": 1.0379,
+      "step": 4379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2718896024874906,
+      "learning_rate": 0.0001864027926444727,
+      "loss": 1.1072,
+      "step": 4380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2620818390598612,
+      "learning_rate": 0.00018639482733756601,
+      "loss": 1.0537,
+      "step": 4381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2289284295029631,
+      "learning_rate": 0.00018638685986857448,
+      "loss": 1.0194,
+      "step": 4382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29452954070001014,
+      "learning_rate": 0.00018637889023769748,
+      "loss": 1.1051,
+      "step": 4383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24409495634240713,
+      "learning_rate": 0.00018637091844513445,
+      "loss": 1.089,
+      "step": 4384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2444141149494506,
+      "learning_rate": 0.00018636294449108493,
+      "loss": 0.9955,
+      "step": 4385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25886299860296796,
+      "learning_rate": 0.00018635496837574844,
+      "loss": 1.0442,
+      "step": 4386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2788912752950406,
+      "learning_rate": 0.00018634699009932462,
+      "loss": 1.1965,
+      "step": 4387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2671936525699648,
+      "learning_rate": 0.00018633900966201304,
+      "loss": 1.0077,
+      "step": 4388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28071558942931124,
+      "learning_rate": 0.00018633102706401355,
+      "loss": 1.1462,
+      "step": 4389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.34829423214490346,
+      "learning_rate": 0.00018632304230552582,
+      "loss": 1.0623,
+      "step": 4390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24191132707171756,
+      "learning_rate": 0.0001863150553867497,
+      "loss": 1.0765,
+      "step": 4391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27536728708801206,
+      "learning_rate": 0.00018630706630788505,
+      "loss": 1.0193,
+      "step": 4392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2697499088562797,
+      "learning_rate": 0.00018629907506913186,
+      "loss": 1.1339,
+      "step": 4393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26546348358382443,
+      "learning_rate": 0.00018629108167069006,
+      "loss": 1.1509,
+      "step": 4394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2627657556993223,
+      "learning_rate": 0.00018628308611275972,
+      "loss": 1.1698,
+      "step": 4395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2681938025996225,
+      "learning_rate": 0.00018627508839554093,
+      "loss": 1.0299,
+      "step": 4396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26016966609609554,
+      "learning_rate": 0.00018626708851923382,
+      "loss": 1.132,
+      "step": 4397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28096390529787246,
+      "learning_rate": 0.0001862590864840386,
+      "loss": 1.1004,
+      "step": 4398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2914134240329839,
+      "learning_rate": 0.00018625108229015555,
+      "loss": 1.0259,
+      "step": 4399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27633727064382735,
+      "learning_rate": 0.00018624307593778495,
+      "loss": 1.0006,
+      "step": 4400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26565720029010786,
+      "learning_rate": 0.00018623506742712715,
+      "loss": 1.0687,
+      "step": 4401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24736946692607406,
+      "learning_rate": 0.00018622705675838263,
+      "loss": 1.1022,
+      "step": 4402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.23869319091908384,
+      "learning_rate": 0.0001862190439317518,
+      "loss": 1.1136,
+      "step": 4403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26085938439573125,
+      "learning_rate": 0.0001862110289474352,
+      "loss": 1.0647,
+      "step": 4404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2615746057557681,
+      "learning_rate": 0.00018620301180563342,
+      "loss": 1.026,
+      "step": 4405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.265385119223784,
+      "learning_rate": 0.0001861949925065471,
+      "loss": 1.0762,
+      "step": 4406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2621600644598475,
+      "learning_rate": 0.00018618697105037693,
+      "loss": 1.0342,
+      "step": 4407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28497896413460505,
+      "learning_rate": 0.00018617894743732361,
+      "loss": 1.0353,
+      "step": 4408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2537528026008887,
+      "learning_rate": 0.00018617092166758802,
+      "loss": 0.9979,
+      "step": 4409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.273548636864498,
+      "learning_rate": 0.00018616289374137092,
+      "loss": 1.1967,
+      "step": 4410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32894219686067755,
+      "learning_rate": 0.0001861548636588733,
+      "loss": 1.082,
+      "step": 4411
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26569438454996513,
+      "learning_rate": 0.00018614683142029602,
+      "loss": 0.8886,
+      "step": 4412
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25788494163824816,
+      "learning_rate": 0.00018613879702584013,
+      "loss": 1.0712,
+      "step": 4413
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.272910401506832,
+      "learning_rate": 0.00018613076047570678,
+      "loss": 1.0169,
+      "step": 4414
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27161074974964666,
+      "learning_rate": 0.00018612272177009694,
+      "loss": 1.1233,
+      "step": 4415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28453465725372484,
+      "learning_rate": 0.0001861146809092119,
+      "loss": 1.0644,
+      "step": 4416
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26695328078689184,
+      "learning_rate": 0.00018610663789325288,
+      "loss": 1.0859,
+      "step": 4417
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28644342117192206,
+      "learning_rate": 0.00018609859272242108,
+      "loss": 1.1014,
+      "step": 4418
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28232507692715614,
+      "learning_rate": 0.0001860905453969179,
+      "loss": 1.0183,
+      "step": 4419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28914880496465856,
+      "learning_rate": 0.0001860824959169447,
+      "loss": 1.0779,
+      "step": 4420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27341424599975833,
+      "learning_rate": 0.000186074444282703,
+      "loss": 0.9407,
+      "step": 4421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.31013888814670343,
+      "learning_rate": 0.00018606639049439415,
+      "loss": 1.1336,
+      "step": 4422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2804627470142664,
+      "learning_rate": 0.00018605833455221984,
+      "loss": 1.1237,
+      "step": 4423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27060408840367484,
+      "learning_rate": 0.00018605027645638163,
+      "loss": 1.0914,
+      "step": 4424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2949572609779446,
+      "learning_rate": 0.00018604221620708113,
+      "loss": 1.0142,
+      "step": 4425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2978775593613743,
+      "learning_rate": 0.00018603415380452013,
+      "loss": 1.0299,
+      "step": 4426
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.27422717885209724,
+      "learning_rate": 0.00018602608924890034,
+      "loss": 0.9475,
+      "step": 4427
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.24881221794442596,
+      "learning_rate": 0.0001860180225404236,
+      "loss": 1.0083,
+      "step": 4428
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28296329045097385,
+      "learning_rate": 0.00018600995367929182,
+      "loss": 1.1519,
+      "step": 4429
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3132917134168698,
+      "learning_rate": 0.00018600188266570687,
+      "loss": 1.1355,
+      "step": 4430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2656930447471748,
+      "learning_rate": 0.00018599380949987072,
+      "loss": 1.1187,
+      "step": 4431
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.29159128798152506,
+      "learning_rate": 0.0001859857341819855,
+      "loss": 1.0272,
+      "step": 4432
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26189338837080134,
+      "learning_rate": 0.00018597765671225322,
+      "loss": 1.0971,
+      "step": 4433
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.28052538093117707,
+      "learning_rate": 0.00018596957709087603,
+      "loss": 1.0194,
+      "step": 4434
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2652637453080487,
+      "learning_rate": 0.0001859614953180562,
+      "loss": 1.0746,
+      "step": 4435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2829916437751392,
+      "learning_rate": 0.00018595341139399584,
+      "loss": 1.1021,
+      "step": 4436
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.25812189396799773,
+      "learning_rate": 0.0001859453253188974,
+      "loss": 1.0858,
+      "step": 4437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.26939995257598304,
+      "learning_rate": 0.00018593723709296316,
+      "loss": 1.0994,
+      "step": 4438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2860192381686427,
+      "learning_rate": 0.00018592914671639553,
+      "loss": 1.0819,
+      "step": 4439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2791379239990837,
+      "learning_rate": 0.00018592105418939705,
+      "loss": 1.0618,
+      "step": 4440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2740072559598784,
+      "learning_rate": 0.00018591295951217015,
+      "loss": 1.1346,
+      "step": 4441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.2889247502983577,
+      "learning_rate": 0.00018590486268491748,
+      "loss": 1.0459,
+      "step": 4442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2978227410800553,
+      "learning_rate": 0.0001858967637078416,
+      "loss": 1.0501,
+      "step": 4443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3018285492782609,
+      "learning_rate": 0.00018588866258114524,
+      "loss": 1.0894,
+      "step": 4444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28346415215862286,
+      "learning_rate": 0.0001858805593050311,
+      "loss": 0.9718,
+      "step": 4445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26039184338131427,
+      "learning_rate": 0.000185872453879702,
+      "loss": 1.1089,
+      "step": 4446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3206288442226837,
+      "learning_rate": 0.0001858643463053608,
+      "loss": 1.0651,
+      "step": 4447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.265449658618547,
+      "learning_rate": 0.00018585623658221034,
+      "loss": 1.0637,
+      "step": 4448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29272693696344226,
+      "learning_rate": 0.0001858481247104536,
+      "loss": 1.0931,
+      "step": 4449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29578928026152973,
+      "learning_rate": 0.0001858400106902936,
+      "loss": 1.1793,
+      "step": 4450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27791665518502323,
+      "learning_rate": 0.00018583189452193338,
+      "loss": 1.0318,
+      "step": 4451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28708172386603614,
+      "learning_rate": 0.00018582377620557602,
+      "loss": 1.0001,
+      "step": 4452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2945398920234109,
+      "learning_rate": 0.0001858156557414248,
+      "loss": 1.0574,
+      "step": 4453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23921340949830877,
+      "learning_rate": 0.0001858075331296828,
+      "loss": 1.0104,
+      "step": 4454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3568337762102039,
+      "learning_rate": 0.00018579940837055338,
+      "loss": 0.9811,
+      "step": 4455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3247650997857825,
+      "learning_rate": 0.00018579128146423984,
+      "loss": 1.1677,
+      "step": 4456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3235988446494336,
+      "learning_rate": 0.00018578315241094554,
+      "loss": 1.1063,
+      "step": 4457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2910859031590426,
+      "learning_rate": 0.00018577502121087396,
+      "loss": 1.1122,
+      "step": 4458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2855175850221935,
+      "learning_rate": 0.00018576688786422856,
+      "loss": 1.0616,
+      "step": 4459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24829230055660786,
+      "learning_rate": 0.0001857587523712129,
+      "loss": 1.0424,
+      "step": 4460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2840641696870963,
+      "learning_rate": 0.00018575061473203054,
+      "loss": 1.2132,
+      "step": 4461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27078441085947547,
+      "learning_rate": 0.0001857424749468852,
+      "loss": 1.0858,
+      "step": 4462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.285656535390425,
+      "learning_rate": 0.0001857343330159805,
+      "loss": 1.0003,
+      "step": 4463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2687099265187277,
+      "learning_rate": 0.00018572618893952024,
+      "loss": 1.0664,
+      "step": 4464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3181574270847497,
+      "learning_rate": 0.00018571804271770822,
+      "loss": 0.9845,
+      "step": 4465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25226801914556535,
+      "learning_rate": 0.0001857098943507483,
+      "loss": 1.1452,
+      "step": 4466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3055598706754259,
+      "learning_rate": 0.00018570174383884442,
+      "loss": 1.0501,
+      "step": 4467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2382755864156548,
+      "learning_rate": 0.00018569359118220056,
+      "loss": 1.0189,
+      "step": 4468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26719151556229226,
+      "learning_rate": 0.00018568543638102072,
+      "loss": 1.1856,
+      "step": 4469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2671695573995491,
+      "learning_rate": 0.00018567727943550897,
+      "loss": 1.0382,
+      "step": 4470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27765193324490084,
+      "learning_rate": 0.00018566912034586946,
+      "loss": 1.2756,
+      "step": 4471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27186513834244824,
+      "learning_rate": 0.00018566095911230638,
+      "loss": 1.0309,
+      "step": 4472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29391579590132144,
+      "learning_rate": 0.00018565279573502392,
+      "loss": 1.1029,
+      "step": 4473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30436200911544314,
+      "learning_rate": 0.00018564463021422645,
+      "loss": 1.0607,
+      "step": 4474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29885017590562324,
+      "learning_rate": 0.00018563646255011828,
+      "loss": 1.1022,
+      "step": 4475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30524559756359343,
+      "learning_rate": 0.0001856282927429038,
+      "loss": 1.1297,
+      "step": 4476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2862736844725348,
+      "learning_rate": 0.0001856201207927875,
+      "loss": 1.0444,
+      "step": 4477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27357248947278107,
+      "learning_rate": 0.00018561194669997386,
+      "loss": 1.1338,
+      "step": 4478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2642596700408733,
+      "learning_rate": 0.00018560377046466747,
+      "loss": 1.0775,
+      "step": 4479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28175880695705074,
+      "learning_rate": 0.00018559559208707288,
+      "loss": 1.1622,
+      "step": 4480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29569453857541267,
+      "learning_rate": 0.00018558741156739483,
+      "loss": 1.0761,
+      "step": 4481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2751922546444621,
+      "learning_rate": 0.000185579228905838,
+      "loss": 1.0507,
+      "step": 4482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27122561641213205,
+      "learning_rate": 0.00018557104410260722,
+      "loss": 1.174,
+      "step": 4483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2912196997412804,
+      "learning_rate": 0.00018556285715790724,
+      "loss": 1.0137,
+      "step": 4484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2696109196990019,
+      "learning_rate": 0.00018555466807194303,
+      "loss": 0.8727,
+      "step": 4485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27947099836689787,
+      "learning_rate": 0.00018554647684491943,
+      "loss": 1.1609,
+      "step": 4486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26675781929194203,
+      "learning_rate": 0.00018553828347704152,
+      "loss": 1.0248,
+      "step": 4487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2588683196044053,
+      "learning_rate": 0.00018553008796851428,
+      "loss": 1.0563,
+      "step": 4488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27999097742810836,
+      "learning_rate": 0.00018552189031954285,
+      "loss": 1.0313,
+      "step": 4489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2895515425733008,
+      "learning_rate": 0.00018551369053033237,
+      "loss": 1.0168,
+      "step": 4490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2545597330861161,
+      "learning_rate": 0.00018550548860108804,
+      "loss": 1.0956,
+      "step": 4491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26512014338659784,
+      "learning_rate": 0.00018549728453201513,
+      "loss": 1.1138,
+      "step": 4492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27434370289340704,
+      "learning_rate": 0.0001854890783233189,
+      "loss": 1.0605,
+      "step": 4493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2936485774839461,
+      "learning_rate": 0.0001854808699752048,
+      "loss": 1.0747,
+      "step": 4494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27403335709343807,
+      "learning_rate": 0.00018547265948787818,
+      "loss": 1.0193,
+      "step": 4495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.28508114808522955,
+      "learning_rate": 0.00018546444686154455,
+      "loss": 1.0224,
+      "step": 4496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3342643589174626,
+      "learning_rate": 0.00018545623209640941,
+      "loss": 1.0092,
+      "step": 4497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26332940264787036,
+      "learning_rate": 0.0001854480151926784,
+      "loss": 1.1283,
+      "step": 4498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30979393657373167,
+      "learning_rate": 0.00018543979615055705,
+      "loss": 1.1229,
+      "step": 4499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29782251228163314,
+      "learning_rate": 0.00018543157497025113,
+      "loss": 1.0053,
+      "step": 4500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26274351011419134,
+      "learning_rate": 0.00018542335165196635,
+      "loss": 1.1258,
+      "step": 4501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3085288258220842,
+      "learning_rate": 0.00018541512619590854,
+      "loss": 1.0965,
+      "step": 4502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.30190633074116374,
+      "learning_rate": 0.00018540689860228348,
+      "loss": 1.2104,
+      "step": 4503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2754205944826528,
+      "learning_rate": 0.0001853986688712971,
+      "loss": 1.1199,
+      "step": 4504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2590596310721408,
+      "learning_rate": 0.00018539043700315538,
+      "loss": 1.1025,
+      "step": 4505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24485764570082005,
+      "learning_rate": 0.0001853822029980643,
+      "loss": 0.9729,
+      "step": 4506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27774076617481674,
+      "learning_rate": 0.00018537396685622994,
+      "loss": 1.103,
+      "step": 4507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2682689587887445,
+      "learning_rate": 0.00018536572857785842,
+      "loss": 1.105,
+      "step": 4508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29122611914479773,
+      "learning_rate": 0.00018535748816315585,
+      "loss": 1.1096,
+      "step": 4509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.23291917772580553,
+      "learning_rate": 0.0001853492456123285,
+      "loss": 1.1044,
+      "step": 4510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33929513760098673,
+      "learning_rate": 0.00018534100092558266,
+      "loss": 1.1069,
+      "step": 4511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2541464302266371,
+      "learning_rate": 0.00018533275410312464,
+      "loss": 1.043,
+      "step": 4512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2729757201162812,
+      "learning_rate": 0.0001853245051451608,
+      "loss": 1.0917,
+      "step": 4513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2881511710111317,
+      "learning_rate": 0.00018531625405189761,
+      "loss": 1.0334,
+      "step": 4514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3004052066232476,
+      "learning_rate": 0.00018530800082354153,
+      "loss": 1.2131,
+      "step": 4515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.296301949311051,
+      "learning_rate": 0.0001852997454602991,
+      "loss": 1.153,
+      "step": 4516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26666605205862975,
+      "learning_rate": 0.00018529148796237696,
+      "loss": 1.0267,
+      "step": 4517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27466066458804184,
+      "learning_rate": 0.00018528322832998172,
+      "loss": 1.0631,
+      "step": 4518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2913602893847047,
+      "learning_rate": 0.0001852749665633201,
+      "loss": 1.1576,
+      "step": 4519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2631891203792342,
+      "learning_rate": 0.00018526670266259885,
+      "loss": 1.0626,
+      "step": 4520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.33347156766562963,
+      "learning_rate": 0.00018525843662802477,
+      "loss": 1.0737,
+      "step": 4521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2790606063839219,
+      "learning_rate": 0.00018525016845980473,
+      "loss": 1.0466,
+      "step": 4522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2836733588666177,
+      "learning_rate": 0.00018524189815814565,
+      "loss": 1.0932,
+      "step": 4523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3028905196190124,
+      "learning_rate": 0.0001852336257232545,
+      "loss": 1.0616,
+      "step": 4524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2845873599717972,
+      "learning_rate": 0.00018522535115533828,
+      "loss": 1.0551,
+      "step": 4525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24471061186570162,
+      "learning_rate": 0.0001852170744546041,
+      "loss": 1.0166,
+      "step": 4526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2877270171258153,
+      "learning_rate": 0.00018520879562125905,
+      "loss": 1.0145,
+      "step": 4527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2770560293900152,
+      "learning_rate": 0.00018520051465551038,
+      "loss": 1.1089,
+      "step": 4528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27958349827422824,
+      "learning_rate": 0.00018519223155756526,
+      "loss": 1.1539,
+      "step": 4529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.24834333728465557,
+      "learning_rate": 0.000185183946327631,
+      "loss": 1.1034,
+      "step": 4530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2749131019387281,
+      "learning_rate": 0.00018517565896591494,
+      "loss": 0.9705,
+      "step": 4531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2717292532230968,
+      "learning_rate": 0.00018516736947262453,
+      "loss": 1.0913,
+      "step": 4532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.22308007763116086,
+      "learning_rate": 0.00018515907784796712,
+      "loss": 1.0885,
+      "step": 4533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2723580969909008,
+      "learning_rate": 0.00018515078409215029,
+      "loss": 1.0822,
+      "step": 4534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2723449059110442,
+      "learning_rate": 0.00018514248820538157,
+      "loss": 1.1214,
+      "step": 4535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2985439043563554,
+      "learning_rate": 0.0001851341901878686,
+      "loss": 1.0749,
+      "step": 4536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.26995390366164945,
+      "learning_rate": 0.000185125890039819,
+      "loss": 1.1233,
+      "step": 4537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2852650589577388,
+      "learning_rate": 0.00018511758776144048,
+      "loss": 1.135,
+      "step": 4538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.25495506751354075,
+      "learning_rate": 0.0001851092833529408,
+      "loss": 1.0329,
+      "step": 4539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2788212963607499,
+      "learning_rate": 0.0001851009768145279,
+      "loss": 1.0452,
+      "step": 4540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.29286510945056726,
+      "learning_rate": 0.00018509266814640952,
+      "loss": 1.0701,
+      "step": 4541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.298753776585502,
+      "learning_rate": 0.00018508435734879367,
+      "loss": 1.0797,
+      "step": 4542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.2893400312544143,
+      "learning_rate": 0.00018507604442188826,
+      "loss": 1.1236,
+      "step": 4543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.297108222760269,
+      "learning_rate": 0.0001850677293659014,
+      "loss": 1.1958,
+      "step": 4544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3026573746982335,
+      "learning_rate": 0.00018505941218104112,
+      "loss": 1.0817,
+      "step": 4545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.27275443083535883,
+      "learning_rate": 0.00018505109286751564,
+      "loss": 1.0752,
+      "step": 4546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27885346389765003,
+      "learning_rate": 0.00018504277142553308,
+      "loss": 1.0219,
+      "step": 4547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3180339146414116,
+      "learning_rate": 0.00018503444785530172,
+      "loss": 1.0987,
+      "step": 4548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2808155483149112,
+      "learning_rate": 0.00018502612215702988,
+      "loss": 1.073,
+      "step": 4549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25391611727242125,
+      "learning_rate": 0.00018501779433092587,
+      "loss": 1.0876,
+      "step": 4550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2261609383943027,
+      "learning_rate": 0.00018500946437719813,
+      "loss": 1.0297,
+      "step": 4551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2754443425644491,
+      "learning_rate": 0.00018500113229605512,
+      "loss": 1.082,
+      "step": 4552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27586769729030647,
+      "learning_rate": 0.00018499279808770536,
+      "loss": 1.137,
+      "step": 4553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31905598519632444,
+      "learning_rate": 0.0001849844617523574,
+      "loss": 1.1845,
+      "step": 4554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.23729193980262855,
+      "learning_rate": 0.00018497612329021988,
+      "loss": 0.9061,
+      "step": 4555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2782155585428631,
+      "learning_rate": 0.00018496778270150145,
+      "loss": 1.1527,
+      "step": 4556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2710366956832406,
+      "learning_rate": 0.0001849594399864109,
+      "loss": 1.0705,
+      "step": 4557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22796189873292866,
+      "learning_rate": 0.00018495109514515693,
+      "loss": 1.0695,
+      "step": 4558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2831241401573468,
+      "learning_rate": 0.00018494274817794842,
+      "loss": 0.9582,
+      "step": 4559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2824780217557077,
+      "learning_rate": 0.0001849343990849943,
+      "loss": 1.0315,
+      "step": 4560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.270556872472652,
+      "learning_rate": 0.0001849260478665034,
+      "loss": 1.0019,
+      "step": 4561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25499089437909833,
+      "learning_rate": 0.00018491769452268482,
+      "loss": 1.0728,
+      "step": 4562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2908571742460526,
+      "learning_rate": 0.00018490933905374754,
+      "loss": 1.0077,
+      "step": 4563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28487382701620256,
+      "learning_rate": 0.0001849009814599007,
+      "loss": 1.0629,
+      "step": 4564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26175206299618,
+      "learning_rate": 0.00018489262174135345,
+      "loss": 1.1802,
+      "step": 4565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2881582041583295,
+      "learning_rate": 0.00018488425989831496,
+      "loss": 0.9577,
+      "step": 4566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2833504131778664,
+      "learning_rate": 0.00018487589593099455,
+      "loss": 1.2034,
+      "step": 4567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2623734320038995,
+      "learning_rate": 0.00018486752983960146,
+      "loss": 1.1153,
+      "step": 4568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2522517398867482,
+      "learning_rate": 0.00018485916162434515,
+      "loss": 1.0618,
+      "step": 4569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24153711019667667,
+      "learning_rate": 0.00018485079128543496,
+      "loss": 1.0822,
+      "step": 4570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2882476151134348,
+      "learning_rate": 0.0001848424188230804,
+      "loss": 1.1657,
+      "step": 4571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2964951820277572,
+      "learning_rate": 0.00018483404423749096,
+      "loss": 1.1222,
+      "step": 4572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27784406217795626,
+      "learning_rate": 0.00018482566752887628,
+      "loss": 1.0545,
+      "step": 4573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26707431862694997,
+      "learning_rate": 0.00018481728869744596,
+      "loss": 1.0567,
+      "step": 4574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3017539840667631,
+      "learning_rate": 0.00018480890774340964,
+      "loss": 1.1777,
+      "step": 4575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3113524234554564,
+      "learning_rate": 0.00018480052466697715,
+      "loss": 1.1013,
+      "step": 4576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.308306701323019,
+      "learning_rate": 0.00018479213946835822,
+      "loss": 1.1211,
+      "step": 4577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2710413287763368,
+      "learning_rate": 0.00018478375214776272,
+      "loss": 1.1882,
+      "step": 4578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31364976820788937,
+      "learning_rate": 0.00018477536270540052,
+      "loss": 1.0807,
+      "step": 4579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25048911528630063,
+      "learning_rate": 0.00018476697114148158,
+      "loss": 1.0959,
+      "step": 4580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2663984503966,
+      "learning_rate": 0.00018475857745621594,
+      "loss": 1.0236,
+      "step": 4581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26614213923311075,
+      "learning_rate": 0.00018475018164981362,
+      "loss": 1.0751,
+      "step": 4582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2814084782252928,
+      "learning_rate": 0.00018474178372248474,
+      "loss": 0.9935,
+      "step": 4583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25798031430362306,
+      "learning_rate": 0.00018473338367443946,
+      "loss": 0.9938,
+      "step": 4584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2865428384395744,
+      "learning_rate": 0.00018472498150588803,
+      "loss": 1.0878,
+      "step": 4585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3077652321307214,
+      "learning_rate": 0.00018471657721704066,
+      "loss": 1.1567,
+      "step": 4586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31452880319912974,
+      "learning_rate": 0.0001847081708081077,
+      "loss": 1.1304,
+      "step": 4587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2763324776356194,
+      "learning_rate": 0.00018469976227929955,
+      "loss": 1.0634,
+      "step": 4588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29691042342244073,
+      "learning_rate": 0.0001846913516308266,
+      "loss": 1.1236,
+      "step": 4589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2846334629692399,
+      "learning_rate": 0.00018468293886289935,
+      "loss": 0.9717,
+      "step": 4590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2668935421303119,
+      "learning_rate": 0.00018467452397572833,
+      "loss": 1.082,
+      "step": 4591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28376985073992644,
+      "learning_rate": 0.00018466610696952416,
+      "loss": 1.088,
+      "step": 4592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3014664421717131,
+      "learning_rate": 0.00018465768784449742,
+      "loss": 0.9671,
+      "step": 4593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27756706076476795,
+      "learning_rate": 0.00018464926660085885,
+      "loss": 1.0486,
+      "step": 4594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31276907084202593,
+      "learning_rate": 0.00018464084323881918,
+      "loss": 1.1944,
+      "step": 4595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2710547392306981,
+      "learning_rate": 0.00018463241775858923,
+      "loss": 1.0873,
+      "step": 4596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2726412664274147,
+      "learning_rate": 0.00018462399016037982,
+      "loss": 0.975,
+      "step": 4597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28160638881053474,
+      "learning_rate": 0.00018461556044440186,
+      "loss": 1.092,
+      "step": 4598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2725964852327896,
+      "learning_rate": 0.00018460712861086633,
+      "loss": 1.1128,
+      "step": 4599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27984134666173177,
+      "learning_rate": 0.00018459869465998425,
+      "loss": 1.1321,
+      "step": 4600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2863490711350883,
+      "learning_rate": 0.00018459025859196663,
+      "loss": 1.1366,
+      "step": 4601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2879506583130578,
+      "learning_rate": 0.00018458182040702466,
+      "loss": 1.1197,
+      "step": 4602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26719713518673577,
+      "learning_rate": 0.00018457338010536946,
+      "loss": 0.9944,
+      "step": 4603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3121012835925408,
+      "learning_rate": 0.0001845649376872123,
+      "loss": 1.198,
+      "step": 4604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.30448939179247264,
+      "learning_rate": 0.0001845564931527644,
+      "loss": 0.9822,
+      "step": 4605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2735559381750156,
+      "learning_rate": 0.00018454804650223713,
+      "loss": 1.0978,
+      "step": 4606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2555127503868371,
+      "learning_rate": 0.0001845395977358418,
+      "loss": 1.0871,
+      "step": 4607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2784512056488346,
+      "learning_rate": 0.00018453114685379,
+      "loss": 1.1117,
+      "step": 4608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27412619492911333,
+      "learning_rate": 0.0001845226938562931,
+      "loss": 0.9858,
+      "step": 4609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3198996291332153,
+      "learning_rate": 0.00018451423874356261,
+      "loss": 1.0908,
+      "step": 4610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2893727277713405,
+      "learning_rate": 0.00018450578151581022,
+      "loss": 1.1565,
+      "step": 4611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3012087634762126,
+      "learning_rate": 0.00018449732217324754,
+      "loss": 1.1766,
+      "step": 4612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29522825483614856,
+      "learning_rate": 0.00018448886071608625,
+      "loss": 1.2191,
+      "step": 4613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27761023884520314,
+      "learning_rate": 0.00018448039714453814,
+      "loss": 1.2029,
+      "step": 4614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.22548736584501833,
+      "learning_rate": 0.000184471931458815,
+      "loss": 1.0772,
+      "step": 4615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28551218282274554,
+      "learning_rate": 0.00018446346365912867,
+      "loss": 1.1001,
+      "step": 4616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29640261684492736,
+      "learning_rate": 0.0001844549937456911,
+      "loss": 1.0567,
+      "step": 4617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26167728449746586,
+      "learning_rate": 0.0001844465217187142,
+      "loss": 1.0648,
+      "step": 4618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.27822656705075727,
+      "learning_rate": 0.00018443804757841003,
+      "loss": 1.1127,
+      "step": 4619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2904755993031546,
+      "learning_rate": 0.00018442957132499069,
+      "loss": 1.0909,
+      "step": 4620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28052003011517346,
+      "learning_rate": 0.00018442109295866823,
+      "loss": 1.0709,
+      "step": 4621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25770690828327825,
+      "learning_rate": 0.00018441261247965487,
+      "loss": 1.1385,
+      "step": 4622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.256212600739427,
+      "learning_rate": 0.00018440412988816283,
+      "loss": 1.0721,
+      "step": 4623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2749581272974963,
+      "learning_rate": 0.0001843956451844044,
+      "loss": 1.1453,
+      "step": 4624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3046942593317576,
+      "learning_rate": 0.0001843871583685919,
+      "loss": 1.189,
+      "step": 4625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28752913128850266,
+      "learning_rate": 0.00018437866944093773,
+      "loss": 1.2043,
+      "step": 4626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2900180388718663,
+      "learning_rate": 0.00018437017840165434,
+      "loss": 1.1533,
+      "step": 4627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26073951801493656,
+      "learning_rate": 0.0001843616852509542,
+      "loss": 1.0558,
+      "step": 4628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3077527251913004,
+      "learning_rate": 0.00018435318998904986,
+      "loss": 1.0387,
+      "step": 4629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29483737359579054,
+      "learning_rate": 0.00018434469261615393,
+      "loss": 1.0913,
+      "step": 4630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2933924970818525,
+      "learning_rate": 0.00018433619313247906,
+      "loss": 1.0347,
+      "step": 4631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.26212171413860036,
+      "learning_rate": 0.00018432769153823797,
+      "loss": 1.0702,
+      "step": 4632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2546684009326573,
+      "learning_rate": 0.00018431918783364337,
+      "loss": 1.028,
+      "step": 4633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.24838393365899913,
+      "learning_rate": 0.00018431068201890812,
+      "loss": 1.0165,
+      "step": 4634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3099036311014584,
+      "learning_rate": 0.00018430217409424505,
+      "loss": 0.9548,
+      "step": 4635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2759430315688091,
+      "learning_rate": 0.00018429366405986713,
+      "loss": 1.149,
+      "step": 4636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2953749611748448,
+      "learning_rate": 0.00018428515191598726,
+      "loss": 1.1464,
+      "step": 4637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2654249301676195,
+      "learning_rate": 0.0001842766376628185,
+      "loss": 1.0327,
+      "step": 4638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2802886640388649,
+      "learning_rate": 0.0001842681213005739,
+      "loss": 1.1143,
+      "step": 4639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29564078608147715,
+      "learning_rate": 0.00018425960282946661,
+      "loss": 0.8881,
+      "step": 4640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2925251530239038,
+      "learning_rate": 0.00018425108224970983,
+      "loss": 1.1731,
+      "step": 4641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2900983438301286,
+      "learning_rate": 0.00018424255956151674,
+      "loss": 1.177,
+      "step": 4642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.31019044957642683,
+      "learning_rate": 0.00018423403476510065,
+      "loss": 1.1355,
+      "step": 4643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3213591456521516,
+      "learning_rate": 0.00018422550786067492,
+      "loss": 1.077,
+      "step": 4644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2677075004163226,
+      "learning_rate": 0.0001842169788484529,
+      "loss": 1.0312,
+      "step": 4645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.25726221862210036,
+      "learning_rate": 0.0001842084477286481,
+      "loss": 0.9965,
+      "step": 4646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.29706048227507303,
+      "learning_rate": 0.00018419991450147394,
+      "loss": 1.0823,
+      "step": 4647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2567670140513427,
+      "learning_rate": 0.000184191379167144,
+      "loss": 1.0802,
+      "step": 4648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2730948766094166,
+      "learning_rate": 0.00018418284172587188,
+      "loss": 1.0743,
+      "step": 4649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.2576764085593571,
+      "learning_rate": 0.00018417430217787124,
+      "loss": 1.0818,
+      "step": 4650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.28479973480576726,
+      "learning_rate": 0.00018416576052335582,
+      "loss": 1.0513,
+      "step": 4651
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28752546745083385,
+      "learning_rate": 0.0001841572167625393,
+      "loss": 1.1685,
+      "step": 4652
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2846736795661952,
+      "learning_rate": 0.00018414867089563557,
+      "loss": 1.2087,
+      "step": 4653
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27455162051832954,
+      "learning_rate": 0.00018414012292285845,
+      "loss": 1.0672,
+      "step": 4654
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26145375283055017,
+      "learning_rate": 0.00018413157284442186,
+      "loss": 1.0861,
+      "step": 4655
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2871748977574908,
+      "learning_rate": 0.0001841230206605398,
+      "loss": 1.168,
+      "step": 4656
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26086270240508314,
+      "learning_rate": 0.00018411446637142632,
+      "loss": 1.1131,
+      "step": 4657
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26968127971681205,
+      "learning_rate": 0.0001841059099772954,
+      "loss": 1.0536,
+      "step": 4658
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2665866276637478,
+      "learning_rate": 0.00018409735147836124,
+      "loss": 1.2219,
+      "step": 4659
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2788104688731091,
+      "learning_rate": 0.000184088790874838,
+      "loss": 1.0498,
+      "step": 4660
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3042257990864148,
+      "learning_rate": 0.00018408022816693994,
+      "loss": 1.1032,
+      "step": 4661
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23891458103453578,
+      "learning_rate": 0.0001840716633548813,
+      "loss": 1.1667,
+      "step": 4662
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2668685305184029,
+      "learning_rate": 0.00018406309643887649,
+      "loss": 1.0313,
+      "step": 4663
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26923096074860575,
+      "learning_rate": 0.0001840545274191398,
+      "loss": 1.096,
+      "step": 4664
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26120691585456396,
+      "learning_rate": 0.0001840459562958858,
+      "loss": 1.0147,
+      "step": 4665
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.279287076136616,
+      "learning_rate": 0.0001840373830693289,
+      "loss": 1.12,
+      "step": 4666
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2517878873784173,
+      "learning_rate": 0.00018402880773968363,
+      "loss": 1.0207,
+      "step": 4667
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2554752277219163,
+      "learning_rate": 0.00018402023030716469,
+      "loss": 1.1272,
+      "step": 4668
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24985971217131164,
+      "learning_rate": 0.00018401165077198666,
+      "loss": 1.0905,
+      "step": 4669
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2515290610443281,
+      "learning_rate": 0.0001840030691343643,
+      "loss": 1.1553,
+      "step": 4670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2632716739172169,
+      "learning_rate": 0.00018399448539451228,
+      "loss": 1.0571,
+      "step": 4671
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29388108923128464,
+      "learning_rate": 0.00018398589955264552,
+      "loss": 1.0905,
+      "step": 4672
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.34065302184228613,
+      "learning_rate": 0.00018397731160897882,
+      "loss": 1.118,
+      "step": 4673
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26474232885623084,
+      "learning_rate": 0.00018396872156372713,
+      "loss": 1.1804,
+      "step": 4674
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23096264177639642,
+      "learning_rate": 0.00018396012941710542,
+      "loss": 1.0124,
+      "step": 4675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26143495936449157,
+      "learning_rate": 0.00018395153516932868,
+      "loss": 1.0354,
+      "step": 4676
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2909688243868478,
+      "learning_rate": 0.00018394293882061203,
+      "loss": 1.0682,
+      "step": 4677
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2786275126719699,
+      "learning_rate": 0.00018393434037117056,
+      "loss": 1.1167,
+      "step": 4678
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31697489334219714,
+      "learning_rate": 0.0001839257398212195,
+      "loss": 1.0991,
+      "step": 4679
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27077228765384886,
+      "learning_rate": 0.00018391713717097404,
+      "loss": 0.9453,
+      "step": 4680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2762723786687761,
+      "learning_rate": 0.0001839085324206495,
+      "loss": 1.1394,
+      "step": 4681
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2904037438324207,
+      "learning_rate": 0.00018389992557046116,
+      "loss": 0.9806,
+      "step": 4682
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.33416394497118107,
+      "learning_rate": 0.00018389131662062449,
+      "loss": 1.0519,
+      "step": 4683
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.280572824730302,
+      "learning_rate": 0.00018388270557135488,
+      "loss": 1.0679,
+      "step": 4684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.318148944370186,
+      "learning_rate": 0.00018387409242286786,
+      "loss": 1.0326,
+      "step": 4685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2592469384169005,
+      "learning_rate": 0.00018386547717537895,
+      "loss": 1.154,
+      "step": 4686
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.24639073939502068,
+      "learning_rate": 0.00018385685982910376,
+      "loss": 1.0404,
+      "step": 4687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.305988694893456,
+      "learning_rate": 0.00018384824038425796,
+      "loss": 1.0658,
+      "step": 4688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29161415294102544,
+      "learning_rate": 0.00018383961884105724,
+      "loss": 1.0198,
+      "step": 4689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2679552986606862,
+      "learning_rate": 0.00018383099519971737,
+      "loss": 1.081,
+      "step": 4690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27720828972286177,
+      "learning_rate": 0.00018382236946045416,
+      "loss": 1.1043,
+      "step": 4691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26516410427337717,
+      "learning_rate": 0.0001838137416234835,
+      "loss": 0.9742,
+      "step": 4692
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27803791295610497,
+      "learning_rate": 0.00018380511168902128,
+      "loss": 1.0465,
+      "step": 4693
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3744328239378393,
+      "learning_rate": 0.00018379647965728344,
+      "loss": 1.1089,
+      "step": 4694
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2973942314834355,
+      "learning_rate": 0.00018378784552848605,
+      "loss": 1.0153,
+      "step": 4695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26659388968085823,
+      "learning_rate": 0.00018377920930284515,
+      "loss": 1.1312,
+      "step": 4696
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25981005331538953,
+      "learning_rate": 0.0001837705709805769,
+      "loss": 1.0054,
+      "step": 4697
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.22213075334205962,
+      "learning_rate": 0.00018376193056189745,
+      "loss": 1.015,
+      "step": 4698
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.275135826725204,
+      "learning_rate": 0.00018375328804702304,
+      "loss": 0.965,
+      "step": 4699
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29035632968889363,
+      "learning_rate": 0.00018374464343617,
+      "loss": 1.0289,
+      "step": 4700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27764015636676737,
+      "learning_rate": 0.0001837359967295546,
+      "loss": 1.0597,
+      "step": 4701
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25677252798632294,
+      "learning_rate": 0.00018372734792739323,
+      "loss": 1.0364,
+      "step": 4702
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3040774439028437,
+      "learning_rate": 0.0001837186970299024,
+      "loss": 1.157,
+      "step": 4703
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2693075199539607,
+      "learning_rate": 0.00018371004403729853,
+      "loss": 1.0576,
+      "step": 4704
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29196540371409324,
+      "learning_rate": 0.0001837013889497982,
+      "loss": 1.1876,
+      "step": 4705
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2926321409398997,
+      "learning_rate": 0.00018369273176761802,
+      "loss": 1.1449,
+      "step": 4706
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26895329592482214,
+      "learning_rate": 0.00018368407249097466,
+      "loss": 1.0901,
+      "step": 4707
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2606964589821753,
+      "learning_rate": 0.00018367541112008476,
+      "loss": 1.1665,
+      "step": 4708
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3222647861912937,
+      "learning_rate": 0.0001836667476551651,
+      "loss": 1.0423,
+      "step": 4709
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28474742820692894,
+      "learning_rate": 0.00018365808209643253,
+      "loss": 1.0235,
+      "step": 4710
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.29857635596649795,
+      "learning_rate": 0.00018364941444410385,
+      "loss": 1.1574,
+      "step": 4711
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3222102111839974,
+      "learning_rate": 0.00018364074469839602,
+      "loss": 1.1667,
+      "step": 4712
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.31112582358071006,
+      "learning_rate": 0.00018363207285952595,
+      "loss": 1.152,
+      "step": 4713
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.26706361932201095,
+      "learning_rate": 0.00018362339892771072,
+      "loss": 1.1072,
+      "step": 4714
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3008566236074501,
+      "learning_rate": 0.00018361472290316736,
+      "loss": 1.0795,
+      "step": 4715
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2802364960499926,
+      "learning_rate": 0.00018360604478611303,
+      "loss": 1.062,
+      "step": 4716
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25442275739177317,
+      "learning_rate": 0.00018359736457676488,
+      "loss": 1.0775,
+      "step": 4717
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28931573899470037,
+      "learning_rate": 0.00018358868227534014,
+      "loss": 1.1024,
+      "step": 4718
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.256028110974589,
+      "learning_rate": 0.0001835799978820561,
+      "loss": 1.1924,
+      "step": 4719
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2645246955746137,
+      "learning_rate": 0.00018357131139713008,
+      "loss": 1.1301,
+      "step": 4720
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3130330679624715,
+      "learning_rate": 0.0001835626228207795,
+      "loss": 1.0055,
+      "step": 4721
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28521400562617844,
+      "learning_rate": 0.00018355393215322173,
+      "loss": 1.1705,
+      "step": 4722
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23137236678760192,
+      "learning_rate": 0.0001835452393946743,
+      "loss": 1.0553,
+      "step": 4723
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2716401502501313,
+      "learning_rate": 0.00018353654454535473,
+      "loss": 0.9819,
+      "step": 4724
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25530206196101923,
+      "learning_rate": 0.00018352784760548066,
+      "loss": 1.0459,
+      "step": 4725
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3028518361428745,
+      "learning_rate": 0.0001835191485752697,
+      "loss": 1.1051,
+      "step": 4726
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.266666477668348,
+      "learning_rate": 0.00018351044745493957,
+      "loss": 1.0546,
+      "step": 4727
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.270858102859276,
+      "learning_rate": 0.000183501744244708,
+      "loss": 1.1381,
+      "step": 4728
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2626211922077772,
+      "learning_rate": 0.0001834930389447928,
+      "loss": 1.0571,
+      "step": 4729
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25526998893008135,
+      "learning_rate": 0.00018348433155541182,
+      "loss": 1.0052,
+      "step": 4730
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2578903383304173,
+      "learning_rate": 0.000183475622076783,
+      "loss": 0.9924,
+      "step": 4731
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2413218945569206,
+      "learning_rate": 0.00018346691050912423,
+      "loss": 1.1513,
+      "step": 4732
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2446885085320821,
+      "learning_rate": 0.0001834581968526536,
+      "loss": 1.0376,
+      "step": 4733
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3680136627218759,
+      "learning_rate": 0.00018344948110758912,
+      "loss": 0.9561,
+      "step": 4734
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2723843485796278,
+      "learning_rate": 0.00018344076327414896,
+      "loss": 1.0291,
+      "step": 4735
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3027706624868208,
+      "learning_rate": 0.00018343204335255123,
+      "loss": 1.1075,
+      "step": 4736
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.27483257758567703,
+      "learning_rate": 0.00018342332134301418,
+      "loss": 1.0461,
+      "step": 4737
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30639774434247813,
+      "learning_rate": 0.00018341459724575612,
+      "loss": 1.1396,
+      "step": 4738
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30130628173745716,
+      "learning_rate": 0.00018340587106099532,
+      "loss": 1.0496,
+      "step": 4739
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2951096190186332,
+      "learning_rate": 0.00018339714278895017,
+      "loss": 1.0468,
+      "step": 4740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.255641750420738,
+      "learning_rate": 0.0001833884124298391,
+      "loss": 0.9368,
+      "step": 4741
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.28182596587401804,
+      "learning_rate": 0.00018337967998388062,
+      "loss": 1.1615,
+      "step": 4742
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2654967829673649,
+      "learning_rate": 0.00018337094545129327,
+      "loss": 1.1393,
+      "step": 4743
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2809627165424522,
+      "learning_rate": 0.00018336220883229557,
+      "loss": 1.0288,
+      "step": 4744
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3063917931895973,
+      "learning_rate": 0.0001833534701271062,
+      "loss": 1.1096,
+      "step": 4745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2978525520912866,
+      "learning_rate": 0.00018334472933594388,
+      "loss": 1.1158,
+      "step": 4746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2957907023562577,
+      "learning_rate": 0.00018333598645902733,
+      "loss": 1.0369,
+      "step": 4747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.23971767785327205,
+      "learning_rate": 0.00018332724149657534,
+      "loss": 1.0324,
+      "step": 4748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.25958963718826317,
+      "learning_rate": 0.00018331849444880676,
+      "loss": 1.0323,
+      "step": 4749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2702035936636161,
+      "learning_rate": 0.00018330974531594046,
+      "loss": 1.146,
+      "step": 4750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30569382745175333,
+      "learning_rate": 0.00018330099409819548,
+      "loss": 1.1357,
+      "step": 4751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2820062684104576,
+      "learning_rate": 0.00018329224079579072,
+      "loss": 1.1124,
+      "step": 4752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2829441998287804,
+      "learning_rate": 0.0001832834854089453,
+      "loss": 1.0883,
+      "step": 4753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.30058225164600477,
+      "learning_rate": 0.00018327472793787833,
+      "loss": 1.1226,
+      "step": 4754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.2806946975067004,
+      "learning_rate": 0.00018326596838280897,
+      "loss": 1.0858,
+      "step": 4755
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26835184504189624,
+      "learning_rate": 0.0001832572067439564,
+      "loss": 1.0372,
+      "step": 4756
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2284738744280186,
+      "learning_rate": 0.00018324844302153992,
+      "loss": 0.9814,
+      "step": 4757
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2678772918055495,
+      "learning_rate": 0.00018323967721577881,
+      "loss": 1.1667,
+      "step": 4758
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2842258914883526,
+      "learning_rate": 0.00018323090932689248,
+      "loss": 1.0497,
+      "step": 4759
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2909027276625095,
+      "learning_rate": 0.00018322213935510035,
+      "loss": 1.083,
+      "step": 4760
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2861834538414527,
+      "learning_rate": 0.00018321336730062185,
+      "loss": 1.2472,
+      "step": 4761
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3061550280271623,
+      "learning_rate": 0.00018320459316367656,
+      "loss": 1.0457,
+      "step": 4762
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29646221655536087,
+      "learning_rate": 0.00018319581694448402,
+      "loss": 1.2093,
+      "step": 4763
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24948311138239332,
+      "learning_rate": 0.00018318703864326387,
+      "loss": 1.0602,
+      "step": 4764
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2972189282962595,
+      "learning_rate": 0.0001831782582602358,
+      "loss": 1.0545,
+      "step": 4765
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2766121044623654,
+      "learning_rate": 0.00018316947579561955,
+      "loss": 1.1347,
+      "step": 4766
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24848006867021413,
+      "learning_rate": 0.0001831606912496349,
+      "loss": 0.9744,
+      "step": 4767
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31237542451843275,
+      "learning_rate": 0.00018315190462250166,
+      "loss": 1.1843,
+      "step": 4768
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29004056616059715,
+      "learning_rate": 0.00018314311591443978,
+      "loss": 1.1249,
+      "step": 4769
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26116055277908656,
+      "learning_rate": 0.00018313432512566914,
+      "loss": 1.1072,
+      "step": 4770
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28883119567906423,
+      "learning_rate": 0.0001831255322564098,
+      "loss": 1.0457,
+      "step": 4771
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.303739333682482,
+      "learning_rate": 0.00018311673730688174,
+      "loss": 1.0541,
+      "step": 4772
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29152935043740263,
+      "learning_rate": 0.00018310794027730513,
+      "loss": 0.9989,
+      "step": 4773
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2642591264939651,
+      "learning_rate": 0.00018309914116790006,
+      "loss": 1.0887,
+      "step": 4774
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2737816497235381,
+      "learning_rate": 0.00018309033997888677,
+      "loss": 0.9973,
+      "step": 4775
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27257528194967345,
+      "learning_rate": 0.0001830815367104855,
+      "loss": 1.0495,
+      "step": 4776
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28425805359861567,
+      "learning_rate": 0.00018307273136291654,
+      "loss": 1.0373,
+      "step": 4777
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29026318888377095,
+      "learning_rate": 0.00018306392393640025,
+      "loss": 1.0727,
+      "step": 4778
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27166227285923494,
+      "learning_rate": 0.0001830551144311571,
+      "loss": 1.0168,
+      "step": 4779
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2755929576909545,
+      "learning_rate": 0.00018304630284740752,
+      "loss": 1.1526,
+      "step": 4780
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2899920317630215,
+      "learning_rate": 0.00018303748918537197,
+      "loss": 1.1746,
+      "step": 4781
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2986410508568978,
+      "learning_rate": 0.00018302867344527113,
+      "loss": 0.9996,
+      "step": 4782
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.265948409051619,
+      "learning_rate": 0.00018301985562732548,
+      "loss": 1.0769,
+      "step": 4783
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28694971143700526,
+      "learning_rate": 0.0001830110357317558,
+      "loss": 1.1081,
+      "step": 4784
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27742537509438936,
+      "learning_rate": 0.00018300221375878282,
+      "loss": 1.0131,
+      "step": 4785
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2625193602157892,
+      "learning_rate": 0.00018299338970862724,
+      "loss": 1.1648,
+      "step": 4786
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2738231281645282,
+      "learning_rate": 0.00018298456358150996,
+      "loss": 1.0869,
+      "step": 4787
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28533535787671566,
+      "learning_rate": 0.00018297573537765175,
+      "loss": 1.1657,
+      "step": 4788
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24149533195194994,
+      "learning_rate": 0.00018296690509727367,
+      "loss": 1.1252,
+      "step": 4789
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2513397282973885,
+      "learning_rate": 0.00018295807274059663,
+      "loss": 1.1539,
+      "step": 4790
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26483605282680023,
+      "learning_rate": 0.00018294923830784168,
+      "loss": 1.1032,
+      "step": 4791
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24791217448830763,
+      "learning_rate": 0.0001829404017992299,
+      "loss": 1.019,
+      "step": 4792
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2666231822052287,
+      "learning_rate": 0.00018293156321498247,
+      "loss": 1.0347,
+      "step": 4793
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29874993409578854,
+      "learning_rate": 0.0001829227225553205,
+      "loss": 1.178,
+      "step": 4794
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2738977464194366,
+      "learning_rate": 0.00018291387982046536,
+      "loss": 1.1841,
+      "step": 4795
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26913204839505855,
+      "learning_rate": 0.00018290503501063819,
+      "loss": 1.1249,
+      "step": 4796
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2642140627592377,
+      "learning_rate": 0.00018289618812606046,
+      "loss": 1.1228,
+      "step": 4797
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2923403068856712,
+      "learning_rate": 0.00018288733916695351,
+      "loss": 1.0207,
+      "step": 4798
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2574187448061279,
+      "learning_rate": 0.0001828784881335388,
+      "loss": 1.022,
+      "step": 4799
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2705462102994065,
+      "learning_rate": 0.00018286963502603786,
+      "loss": 1.1524,
+      "step": 4800
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2570338839265344,
+      "learning_rate": 0.0001828607798446722,
+      "loss": 1.0328,
+      "step": 4801
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3210236176915547,
+      "learning_rate": 0.00018285192258966343,
+      "loss": 1.1115,
+      "step": 4802
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28608741562007683,
+      "learning_rate": 0.00018284306326123327,
+      "loss": 1.1011,
+      "step": 4803
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2945292323022467,
+      "learning_rate": 0.00018283420185960338,
+      "loss": 1.0293,
+      "step": 4804
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2921369853346895,
+      "learning_rate": 0.00018282533838499552,
+      "loss": 1.0956,
+      "step": 4805
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2508835050525997,
+      "learning_rate": 0.0001828164728376315,
+      "loss": 0.9944,
+      "step": 4806
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31162618222530697,
+      "learning_rate": 0.00018280760521773322,
+      "loss": 1.1071,
+      "step": 4807
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26840278702905573,
+      "learning_rate": 0.00018279873552552256,
+      "loss": 1.0196,
+      "step": 4808
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28949640843261154,
+      "learning_rate": 0.0001827898637612215,
+      "loss": 1.0671,
+      "step": 4809
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26451664824875853,
+      "learning_rate": 0.00018278098992505207,
+      "loss": 1.037,
+      "step": 4810
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3163920468320238,
+      "learning_rate": 0.00018277211401723634,
+      "loss": 1.0914,
+      "step": 4811
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.293156846011273,
+      "learning_rate": 0.00018276323603799645,
+      "loss": 1.0247,
+      "step": 4812
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3167908615898505,
+      "learning_rate": 0.00018275435598755457,
+      "loss": 1.0896,
+      "step": 4813
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29195900263136115,
+      "learning_rate": 0.0001827454738661329,
+      "loss": 1.106,
+      "step": 4814
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24880439045988673,
+      "learning_rate": 0.00018273658967395378,
+      "loss": 1.0249,
+      "step": 4815
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27945847139446783,
+      "learning_rate": 0.00018272770341123948,
+      "loss": 1.1502,
+      "step": 4816
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2826003957306108,
+      "learning_rate": 0.0001827188150782124,
+      "loss": 1.0295,
+      "step": 4817
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2733756686159319,
+      "learning_rate": 0.000182709924675095,
+      "loss": 1.1557,
+      "step": 4818
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2719146164107582,
+      "learning_rate": 0.00018270103220210975,
+      "loss": 1.1701,
+      "step": 4819
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27143925540828834,
+      "learning_rate": 0.0001826921376594792,
+      "loss": 1.094,
+      "step": 4820
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3064338525945154,
+      "learning_rate": 0.00018268324104742592,
+      "loss": 1.1517,
+      "step": 4821
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31652916261666003,
+      "learning_rate": 0.00018267434236617257,
+      "loss": 1.1124,
+      "step": 4822
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28456338644193074,
+      "learning_rate": 0.00018266544161594185,
+      "loss": 1.1934,
+      "step": 4823
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24199134880492237,
+      "learning_rate": 0.0001826565387969565,
+      "loss": 0.9981,
+      "step": 4824
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26907228489033597,
+      "learning_rate": 0.00018264763390943932,
+      "loss": 0.9906,
+      "step": 4825
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2885973158602934,
+      "learning_rate": 0.00018263872695361316,
+      "loss": 1.0922,
+      "step": 4826
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2618403285446365,
+      "learning_rate": 0.00018262981792970093,
+      "loss": 1.0619,
+      "step": 4827
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2508955108209627,
+      "learning_rate": 0.00018262090683792556,
+      "loss": 1.0296,
+      "step": 4828
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2805927803935353,
+      "learning_rate": 0.00018261199367851008,
+      "loss": 1.0554,
+      "step": 4829
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2703573079658423,
+      "learning_rate": 0.00018260307845167754,
+      "loss": 1.0561,
+      "step": 4830
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2674244714124627,
+      "learning_rate": 0.00018259416115765103,
+      "loss": 1.1403,
+      "step": 4831
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28142267132464976,
+      "learning_rate": 0.00018258524179665377,
+      "loss": 1.142,
+      "step": 4832
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28299880159536434,
+      "learning_rate": 0.00018257632036890891,
+      "loss": 1.0733,
+      "step": 4833
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26287599628896613,
+      "learning_rate": 0.0001825673968746397,
+      "loss": 1.036,
+      "step": 4834
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27196439255395516,
+      "learning_rate": 0.00018255847131406954,
+      "loss": 1.0408,
+      "step": 4835
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27846018633896047,
+      "learning_rate": 0.00018254954368742172,
+      "loss": 1.0871,
+      "step": 4836
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.29707264756040697,
+      "learning_rate": 0.00018254061399491968,
+      "loss": 1.0648,
+      "step": 4837
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26505419324604984,
+      "learning_rate": 0.00018253168223678694,
+      "loss": 1.1028,
+      "step": 4838
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.255410178872581,
+      "learning_rate": 0.00018252274841324697,
+      "loss": 1.0632,
+      "step": 4839
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.31740617878546407,
+      "learning_rate": 0.00018251381252452334,
+      "loss": 1.0536,
+      "step": 4840
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.283151238470317,
+      "learning_rate": 0.0001825048745708397,
+      "loss": 1.0617,
+      "step": 4841
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2754257741606846,
+      "learning_rate": 0.0001824959345524197,
+      "loss": 1.1151,
+      "step": 4842
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.27843031061325174,
+      "learning_rate": 0.00018248699246948714,
+      "loss": 1.0461,
+      "step": 4843
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.28725916243319277,
+      "learning_rate": 0.00018247804832226573,
+      "loss": 1.1349,
+      "step": 4844
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2656780830493282,
+      "learning_rate": 0.00018246910211097933,
+      "loss": 1.0609,
+      "step": 4845
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2735044242141262,
+      "learning_rate": 0.0001824601538358518,
+      "loss": 1.1276,
+      "step": 4846
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2758621380929105,
+      "learning_rate": 0.00018245120349710708,
+      "loss": 0.9056,
+      "step": 4847
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2399901693338529,
+      "learning_rate": 0.00018244225109496922,
+      "loss": 1.0778,
+      "step": 4848
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24569048440162722,
+      "learning_rate": 0.0001824332966296622,
+      "loss": 1.1481,
+      "step": 4849
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.24697964457020974,
+      "learning_rate": 0.00018242434010141013,
+      "loss": 1.0456,
+      "step": 4850
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.26864292763639014,
+      "learning_rate": 0.0001824153815104371,
+      "loss": 1.0611,
+      "step": 4851
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32764299735013946,
+      "learning_rate": 0.0001824064208569674,
+      "loss": 1.1193,
+      "step": 4852
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2841433629068385,
+      "learning_rate": 0.00018239745814122523,
+      "loss": 1.1592,
+      "step": 4853
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2803323057557425,
+      "learning_rate": 0.00018238849336343487,
+      "loss": 1.103,
+      "step": 4854
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2403783130731807,
+      "learning_rate": 0.00018237952652382067,
+      "loss": 1.1279,
+      "step": 4855
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2644107034178308,
+      "learning_rate": 0.00018237055762260708,
+      "loss": 1.0965,
+      "step": 4856
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.30697640163453677,
+      "learning_rate": 0.0001823615866600185,
+      "loss": 1.2236,
+      "step": 4857
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2735457493491036,
+      "learning_rate": 0.00018235261363627945,
+      "loss": 1.0618,
+      "step": 4858
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2406737920100741,
+      "learning_rate": 0.00018234363855161448,
+      "loss": 1.004,
+      "step": 4859
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.2484692248494346,
+      "learning_rate": 0.00018233466140624822,
+      "loss": 0.9887,
+      "step": 4860
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2896830439213678,
+      "learning_rate": 0.00018232568220040532,
+      "loss": 1.1294,
+      "step": 4861
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26324509988169137,
+      "learning_rate": 0.00018231670093431042,
+      "loss": 1.1409,
+      "step": 4862
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2545994960965595,
+      "learning_rate": 0.00018230771760818844,
+      "loss": 1.0028,
+      "step": 4863
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2708183217546789,
+      "learning_rate": 0.000182298732222264,
+      "loss": 1.0793,
+      "step": 4864
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2917108244983868,
+      "learning_rate": 0.00018228974477676216,
+      "loss": 1.0369,
+      "step": 4865
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.261745568565937,
+      "learning_rate": 0.0001822807552719077,
+      "loss": 0.962,
+      "step": 4866
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2743799469305386,
+      "learning_rate": 0.0001822717637079256,
+      "loss": 1.2131,
+      "step": 4867
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2979549118848966,
+      "learning_rate": 0.0001822627700850409,
+      "loss": 1.0951,
+      "step": 4868
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2589301792916495,
+      "learning_rate": 0.00018225377440347874,
+      "loss": 1.1224,
+      "step": 4869
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27151217602723077,
+      "learning_rate": 0.00018224477666346414,
+      "loss": 1.2002,
+      "step": 4870
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2678272434705896,
+      "learning_rate": 0.00018223577686522232,
+      "loss": 1.0903,
+      "step": 4871
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28664523382221585,
+      "learning_rate": 0.0001822267750089785,
+      "loss": 1.1004,
+      "step": 4872
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2947186333319964,
+      "learning_rate": 0.00018221777109495797,
+      "loss": 1.0248,
+      "step": 4873
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2911818485318527,
+      "learning_rate": 0.00018220876512338604,
+      "loss": 1.1243,
+      "step": 4874
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2389316378471969,
+      "learning_rate": 0.0001821997570944881,
+      "loss": 1.0098,
+      "step": 4875
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29112691824378567,
+      "learning_rate": 0.00018219074700848956,
+      "loss": 1.1284,
+      "step": 4876
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.32234639877925986,
+      "learning_rate": 0.00018218173486561593,
+      "loss": 1.1974,
+      "step": 4877
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28693422182466444,
+      "learning_rate": 0.00018217272066609275,
+      "loss": 1.07,
+      "step": 4878
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24948119387159542,
+      "learning_rate": 0.00018216370441014558,
+      "loss": 1.0697,
+      "step": 4879
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2766198420685631,
+      "learning_rate": 0.00018215468609800007,
+      "loss": 1.1055,
+      "step": 4880
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2913825743782603,
+      "learning_rate": 0.0001821456657298819,
+      "loss": 1.0749,
+      "step": 4881
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2688271558729764,
+      "learning_rate": 0.00018213664330601683,
+      "loss": 0.9326,
+      "step": 4882
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30551493928355505,
+      "learning_rate": 0.00018212761882663062,
+      "loss": 1.1667,
+      "step": 4883
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3272843823903593,
+      "learning_rate": 0.00018211859229194918,
+      "loss": 1.0908,
+      "step": 4884
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2683326547425987,
+      "learning_rate": 0.00018210956370219832,
+      "loss": 1.1501,
+      "step": 4885
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2508457951874583,
+      "learning_rate": 0.00018210053305760403,
+      "loss": 1.0294,
+      "step": 4886
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3066517541478313,
+      "learning_rate": 0.0001820915003583923,
+      "loss": 1.1289,
+      "step": 4887
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2979558635215368,
+      "learning_rate": 0.0001820824656047892,
+      "loss": 1.0369,
+      "step": 4888
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24297953137434605,
+      "learning_rate": 0.0001820734287970208,
+      "loss": 1.0952,
+      "step": 4889
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2642818716032307,
+      "learning_rate": 0.00018206438993531324,
+      "loss": 1.1495,
+      "step": 4890
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28250099631091213,
+      "learning_rate": 0.0001820553490198928,
+      "loss": 1.2023,
+      "step": 4891
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24244241524853985,
+      "learning_rate": 0.00018204630605098563,
+      "loss": 1.1123,
+      "step": 4892
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28117721265077,
+      "learning_rate": 0.00018203726102881807,
+      "loss": 1.1124,
+      "step": 4893
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24950954002584746,
+      "learning_rate": 0.00018202821395361656,
+      "loss": 1.0812,
+      "step": 4894
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29484743456459533,
+      "learning_rate": 0.0001820191648256074,
+      "loss": 1.019,
+      "step": 4895
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26478910448375825,
+      "learning_rate": 0.00018201011364501712,
+      "loss": 1.061,
+      "step": 4896
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2802200200065187,
+      "learning_rate": 0.00018200106041207218,
+      "loss": 1.2153,
+      "step": 4897
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2931006305889275,
+      "learning_rate": 0.00018199200512699918,
+      "loss": 1.1586,
+      "step": 4898
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28693848326326754,
+      "learning_rate": 0.00018198294779002473,
+      "loss": 0.8363,
+      "step": 4899
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2488053532878251,
+      "learning_rate": 0.00018197388840137548,
+      "loss": 1.0084,
+      "step": 4900
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30279834945613493,
+      "learning_rate": 0.00018196482696127814,
+      "loss": 1.0889,
+      "step": 4901
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27315676586595977,
+      "learning_rate": 0.0001819557634699595,
+      "loss": 1.194,
+      "step": 4902
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27870113697280036,
+      "learning_rate": 0.0001819466979276464,
+      "loss": 1.0237,
+      "step": 4903
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26068511407455874,
+      "learning_rate": 0.00018193763033456565,
+      "loss": 1.141,
+      "step": 4904
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25403699988609485,
+      "learning_rate": 0.0001819285606909442,
+      "loss": 0.985,
+      "step": 4905
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25520183475888464,
+      "learning_rate": 0.00018191948899700904,
+      "loss": 0.9452,
+      "step": 4906
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2890128875837949,
+      "learning_rate": 0.00018191041525298719,
+      "loss": 0.9948,
+      "step": 4907
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2826684783407896,
+      "learning_rate": 0.00018190133945910573,
+      "loss": 1.1965,
+      "step": 4908
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27927424785621435,
+      "learning_rate": 0.00018189226161559175,
+      "loss": 1.0749,
+      "step": 4909
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28389430568060914,
+      "learning_rate": 0.00018188318172267245,
+      "loss": 1.1575,
+      "step": 4910
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.21938938517365983,
+      "learning_rate": 0.0001818740997805751,
+      "loss": 1.0206,
+      "step": 4911
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29366370308740425,
+      "learning_rate": 0.00018186501578952693,
+      "loss": 1.0559,
+      "step": 4912
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29917686177952224,
+      "learning_rate": 0.0001818559297497553,
+      "loss": 1.164,
+      "step": 4913
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24934981554280172,
+      "learning_rate": 0.00018184684166148754,
+      "loss": 0.978,
+      "step": 4914
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25613501770670105,
+      "learning_rate": 0.00018183775152495117,
+      "loss": 1.1179,
+      "step": 4915
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28150065228278026,
+      "learning_rate": 0.00018182865934037362,
+      "loss": 1.0,
+      "step": 4916
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2736305655565639,
+      "learning_rate": 0.00018181956510798246,
+      "loss": 0.9436,
+      "step": 4917
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23963815077384734,
+      "learning_rate": 0.00018181046882800525,
+      "loss": 1.1872,
+      "step": 4918
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26272710687943224,
+      "learning_rate": 0.00018180137050066963,
+      "loss": 1.1077,
+      "step": 4919
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2491631914332162,
+      "learning_rate": 0.00018179227012620332,
+      "loss": 1.0311,
+      "step": 4920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2850474055751791,
+      "learning_rate": 0.00018178316770483405,
+      "loss": 1.072,
+      "step": 4921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2563447984596202,
+      "learning_rate": 0.0001817740632367896,
+      "loss": 1.0942,
+      "step": 4922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28963605095483796,
+      "learning_rate": 0.00018176495672229782,
+      "loss": 1.1151,
+      "step": 4923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29799210873069965,
+      "learning_rate": 0.0001817558481615866,
+      "loss": 1.0651,
+      "step": 4924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2900355369937844,
+      "learning_rate": 0.0001817467375548839,
+      "loss": 1.0395,
+      "step": 4925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.26790765666005045,
+      "learning_rate": 0.00018173762490241777,
+      "loss": 1.0195,
+      "step": 4926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27029423974434824,
+      "learning_rate": 0.00018172851020441616,
+      "loss": 1.0376,
+      "step": 4927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29781631086031485,
+      "learning_rate": 0.00018171939346110723,
+      "loss": 1.1626,
+      "step": 4928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24153000189250098,
+      "learning_rate": 0.0001817102746727191,
+      "loss": 1.1416,
+      "step": 4929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2832270079995993,
+      "learning_rate": 0.00018170115383948001,
+      "loss": 1.0366,
+      "step": 4930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28980962720099673,
+      "learning_rate": 0.0001816920309616182,
+      "loss": 1.1199,
+      "step": 4931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2885127084279082,
+      "learning_rate": 0.00018168290603936198,
+      "loss": 0.9624,
+      "step": 4932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.264979114595309,
+      "learning_rate": 0.00018167377907293966,
+      "loss": 1.1314,
+      "step": 4933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29190107567170176,
+      "learning_rate": 0.00018166465006257972,
+      "loss": 1.1603,
+      "step": 4934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.29218193365264744,
+      "learning_rate": 0.0001816555190085106,
+      "loss": 1.1024,
+      "step": 4935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.290286097531124,
+      "learning_rate": 0.00018164638591096078,
+      "loss": 1.0179,
+      "step": 4936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3045929894370889,
+      "learning_rate": 0.00018163725077015883,
+      "loss": 1.1458,
+      "step": 4937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3096182938272493,
+      "learning_rate": 0.0001816281135863334,
+      "loss": 1.0445,
+      "step": 4938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2989704049787465,
+      "learning_rate": 0.00018161897435971312,
+      "loss": 1.0597,
+      "step": 4939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.33521100047691355,
+      "learning_rate": 0.00018160983309052671,
+      "loss": 1.0351,
+      "step": 4940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27937510049751796,
+      "learning_rate": 0.00018160068977900293,
+      "loss": 1.0309,
+      "step": 4941
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.30607280379755825,
+      "learning_rate": 0.00018159154442537058,
+      "loss": 1.0614,
+      "step": 4942
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2942443003492716,
+      "learning_rate": 0.0001815823970298586,
+      "loss": 1.0296,
+      "step": 4943
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27168124667388854,
+      "learning_rate": 0.00018157324759269583,
+      "loss": 1.2025,
+      "step": 4944
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3040911262976451,
+      "learning_rate": 0.00018156409611411127,
+      "loss": 1.1002,
+      "step": 4945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2932905809708139,
+      "learning_rate": 0.00018155494259433397,
+      "loss": 1.1174,
+      "step": 4946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27057969270876014,
+      "learning_rate": 0.00018154578703359294,
+      "loss": 1.1873,
+      "step": 4947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.23291125191630732,
+      "learning_rate": 0.00018153662943211737,
+      "loss": 1.0286,
+      "step": 4948
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2935414983637233,
+      "learning_rate": 0.00018152746979013638,
+      "loss": 1.0808,
+      "step": 4949
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.28416378913000195,
+      "learning_rate": 0.00018151830810787925,
+      "loss": 1.1802,
+      "step": 4950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2804898105323499,
+      "learning_rate": 0.00018150914438557522,
+      "loss": 1.0825,
+      "step": 4951
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2908103000886573,
+      "learning_rate": 0.0001814999786234536,
+      "loss": 1.1133,
+      "step": 4952
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.25695695427126014,
+      "learning_rate": 0.0001814908108217438,
+      "loss": 1.1093,
+      "step": 4953
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27219006337780743,
+      "learning_rate": 0.0001814816409806753,
+      "loss": 1.1098,
+      "step": 4954
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3013175495467479,
+      "learning_rate": 0.00018147246910047747,
+      "loss": 1.0026,
+      "step": 4955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.24847902288191978,
+      "learning_rate": 0.0001814632951813799,
+      "loss": 0.999,
+      "step": 4956
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2691757208602004,
+      "learning_rate": 0.00018145411922361219,
+      "loss": 1.0259,
+      "step": 4957
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2601243300778845,
+      "learning_rate": 0.00018144494122740394,
+      "loss": 0.9758,
+      "step": 4958
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.27967276767781657,
+      "learning_rate": 0.00018143576119298484,
+      "loss": 1.0271,
+      "step": 4959
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3093700281906271,
+      "learning_rate": 0.00018142657912058465,
+      "loss": 0.9627,
+      "step": 4960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2951333334350976,
+      "learning_rate": 0.00018141739501043315,
+      "loss": 1.0788,
+      "step": 4961
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2754995868240653,
+      "learning_rate": 0.00018140820886276018,
+      "loss": 0.987,
+      "step": 4962
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2650597795296595,
+      "learning_rate": 0.0001813990206777956,
+      "loss": 1.108,
+      "step": 4963
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.2680629620506122,
+      "learning_rate": 0.00018138983045576937,
+      "loss": 1.0824,
+      "step": 4964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2745197902744774,
+      "learning_rate": 0.00018138063819691147,
+      "loss": 1.087,
+      "step": 4965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2951908263527352,
+      "learning_rate": 0.00018137144390145194,
+      "loss": 1.0986,
+      "step": 4966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3006255204787449,
+      "learning_rate": 0.00018136224756962093,
+      "loss": 1.0538,
+      "step": 4967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28348543906022167,
+      "learning_rate": 0.00018135304920164854,
+      "loss": 1.0905,
+      "step": 4968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2572498672154097,
+      "learning_rate": 0.00018134384879776497,
+      "loss": 1.1133,
+      "step": 4969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30267258785481194,
+      "learning_rate": 0.00018133464635820042,
+      "loss": 1.1097,
+      "step": 4970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2645184466065096,
+      "learning_rate": 0.00018132544188318526,
+      "loss": 1.111,
+      "step": 4971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2981463331584167,
+      "learning_rate": 0.0001813162353729498,
+      "loss": 1.0594,
+      "step": 4972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3028092831683072,
+      "learning_rate": 0.0001813070268277244,
+      "loss": 1.227,
+      "step": 4973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.276916640360599,
+      "learning_rate": 0.00018129781624773961,
+      "loss": 1.1563,
+      "step": 4974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26497515296266566,
+      "learning_rate": 0.00018128860363322586,
+      "loss": 1.0489,
+      "step": 4975
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2870089379191648,
+      "learning_rate": 0.00018127938898441373,
+      "loss": 1.1085,
+      "step": 4976
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2804072923871364,
+      "learning_rate": 0.00018127017230153378,
+      "loss": 1.1697,
+      "step": 4977
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26600268228720364,
+      "learning_rate": 0.0001812609535848167,
+      "loss": 1.0651,
+      "step": 4978
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3017636575581846,
+      "learning_rate": 0.0001812517328344932,
+      "loss": 1.0479,
+      "step": 4979
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2848522851547357,
+      "learning_rate": 0.000181242510050794,
+      "loss": 1.0961,
+      "step": 4980
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29706092141646745,
+      "learning_rate": 0.00018123328523394992,
+      "loss": 1.0572,
+      "step": 4981
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24623214065631238,
+      "learning_rate": 0.00018122405838419186,
+      "loss": 1.0209,
+      "step": 4982
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27148089077495985,
+      "learning_rate": 0.00018121482950175067,
+      "loss": 1.0425,
+      "step": 4983
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27374692095758846,
+      "learning_rate": 0.00018120559858685734,
+      "loss": 1.0984,
+      "step": 4984
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28314322190570707,
+      "learning_rate": 0.00018119636563974285,
+      "loss": 1.0245,
+      "step": 4985
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27988607306147967,
+      "learning_rate": 0.0001811871306606383,
+      "loss": 1.0786,
+      "step": 4986
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28485376383594274,
+      "learning_rate": 0.0001811778936497748,
+      "loss": 1.0639,
+      "step": 4987
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23607692697159477,
+      "learning_rate": 0.00018116865460738343,
+      "loss": 1.0506,
+      "step": 4988
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29135188062241957,
+      "learning_rate": 0.0001811594135336955,
+      "loss": 1.0689,
+      "step": 4989
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26704486335325794,
+      "learning_rate": 0.00018115017042894227,
+      "loss": 1.1443,
+      "step": 4990
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26331319852301815,
+      "learning_rate": 0.00018114092529335497,
+      "loss": 1.0934,
+      "step": 4991
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26377752564879786,
+      "learning_rate": 0.00018113167812716506,
+      "loss": 0.9711,
+      "step": 4992
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26534776082769385,
+      "learning_rate": 0.0001811224289306039,
+      "loss": 1.0002,
+      "step": 4993
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2711195707498488,
+      "learning_rate": 0.00018111317770390297,
+      "loss": 1.2213,
+      "step": 4994
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24845192655050088,
+      "learning_rate": 0.0001811039244472938,
+      "loss": 1.1603,
+      "step": 4995
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2774236359347863,
+      "learning_rate": 0.00018109466916100793,
+      "loss": 1.0153,
+      "step": 4996
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.291748829551597,
+      "learning_rate": 0.000181085411845277,
+      "loss": 1.139,
+      "step": 4997
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2707487926064819,
+      "learning_rate": 0.0001810761525003327,
+      "loss": 0.9485,
+      "step": 4998
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31763719053401285,
+      "learning_rate": 0.0001810668911264067,
+      "loss": 1.0691,
+      "step": 4999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.271398772008147,
+      "learning_rate": 0.00018105762772373086,
+      "loss": 1.0501,
+      "step": 5000
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2723475184642883,
+      "learning_rate": 0.00018104836229253688,
+      "loss": 1.0046,
+      "step": 5001
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23738205025152834,
+      "learning_rate": 0.00018103909483305672,
+      "loss": 1.1247,
+      "step": 5002
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28212585900181886,
+      "learning_rate": 0.00018102982534552226,
+      "loss": 0.966,
+      "step": 5003
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2923303986208393,
+      "learning_rate": 0.00018102055383016554,
+      "loss": 1.177,
+      "step": 5004
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.261063011725201,
+      "learning_rate": 0.0001810112802872185,
+      "loss": 0.9956,
+      "step": 5005
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28452157947524576,
+      "learning_rate": 0.0001810020047169133,
+      "loss": 1.1531,
+      "step": 5006
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30435514362331084,
+      "learning_rate": 0.00018099272711948197,
+      "loss": 1.101,
+      "step": 5007
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3171881336122615,
+      "learning_rate": 0.0001809834474951568,
+      "loss": 1.1477,
+      "step": 5008
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3021432074002858,
+      "learning_rate": 0.00018097416584416992,
+      "loss": 1.1089,
+      "step": 5009
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2584355712285805,
+      "learning_rate": 0.00018096488216675364,
+      "loss": 1.0171,
+      "step": 5010
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2784662384579421,
+      "learning_rate": 0.00018095559646314033,
+      "loss": 1.1563,
+      "step": 5011
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2590883277277557,
+      "learning_rate": 0.00018094630873356234,
+      "loss": 0.9934,
+      "step": 5012
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.31013602501461507,
+      "learning_rate": 0.0001809370189782521,
+      "loss": 1.0347,
+      "step": 5013
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2908179871344749,
+      "learning_rate": 0.00018092772719744207,
+      "loss": 1.1276,
+      "step": 5014
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27894866086973424,
+      "learning_rate": 0.0001809184333913648,
+      "loss": 1.1603,
+      "step": 5015
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.30799725337128564,
+      "learning_rate": 0.0001809091375602529,
+      "loss": 0.9902,
+      "step": 5016
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26185924608898803,
+      "learning_rate": 0.00018089983970433896,
+      "loss": 1.0665,
+      "step": 5017
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23371144120926612,
+      "learning_rate": 0.0001808905398238557,
+      "loss": 1.0455,
+      "step": 5018
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27647347653440757,
+      "learning_rate": 0.00018088123791903588,
+      "loss": 1.1008,
+      "step": 5019
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2619795224095397,
+      "learning_rate": 0.0001808719339901122,
+      "loss": 1.111,
+      "step": 5020
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28812383140009523,
+      "learning_rate": 0.00018086262803731758,
+      "loss": 1.1625,
+      "step": 5021
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29790943849759494,
+      "learning_rate": 0.00018085332006088486,
+      "loss": 1.0231,
+      "step": 5022
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2584586663721738,
+      "learning_rate": 0.00018084401006104699,
+      "loss": 0.9985,
+      "step": 5023
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3204107870274487,
+      "learning_rate": 0.00018083469803803696,
+      "loss": 1.0371,
+      "step": 5024
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.32851014719630056,
+      "learning_rate": 0.0001808253839920878,
+      "loss": 1.0865,
+      "step": 5025
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28813789079186414,
+      "learning_rate": 0.00018081606792343262,
+      "loss": 1.0732,
+      "step": 5026
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2778577229940317,
+      "learning_rate": 0.00018080674983230455,
+      "loss": 0.9632,
+      "step": 5027
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25371536354442387,
+      "learning_rate": 0.00018079742971893677,
+      "loss": 1.0637,
+      "step": 5028
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2608878940566144,
+      "learning_rate": 0.00018078810758356256,
+      "loss": 1.1195,
+      "step": 5029
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2880007088671419,
+      "learning_rate": 0.00018077878342641514,
+      "loss": 1.0541,
+      "step": 5030
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27719099372774664,
+      "learning_rate": 0.0001807694572477279,
+      "loss": 1.14,
+      "step": 5031
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26236080400560896,
+      "learning_rate": 0.00018076012904773427,
+      "loss": 1.0992,
+      "step": 5032
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28662749205806315,
+      "learning_rate": 0.00018075079882666763,
+      "loss": 1.1681,
+      "step": 5033
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24357430650710507,
+      "learning_rate": 0.0001807414665847615,
+      "loss": 1.0774,
+      "step": 5034
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.22511154512758563,
+      "learning_rate": 0.00018073213232224945,
+      "loss": 0.9704,
+      "step": 5035
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25551533972771157,
+      "learning_rate": 0.000180722796039365,
+      "loss": 1.1366,
+      "step": 5036
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27262724569941255,
+      "learning_rate": 0.0001807134577363419,
+      "loss": 0.9719,
+      "step": 5037
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2640695671336433,
+      "learning_rate": 0.00018070411741341377,
+      "loss": 1.0389,
+      "step": 5038
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25865882131840123,
+      "learning_rate": 0.00018069477507081438,
+      "loss": 1.0127,
+      "step": 5039
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3041466915560575,
+      "learning_rate": 0.00018068543070877752,
+      "loss": 1.1345,
+      "step": 5040
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.280532311208517,
+      "learning_rate": 0.00018067608432753706,
+      "loss": 0.98,
+      "step": 5041
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3067608848042604,
+      "learning_rate": 0.0001806667359273269,
+      "loss": 1.1467,
+      "step": 5042
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28794591958027166,
+      "learning_rate": 0.00018065738550838094,
+      "loss": 1.1362,
+      "step": 5043
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25975883602188704,
+      "learning_rate": 0.00018064803307093325,
+      "loss": 1.0846,
+      "step": 5044
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28326599642955586,
+      "learning_rate": 0.00018063867861521784,
+      "loss": 1.0675,
+      "step": 5045
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2558529563820642,
+      "learning_rate": 0.00018062932214146882,
+      "loss": 1.1281,
+      "step": 5046
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28221647075585127,
+      "learning_rate": 0.0001806199636499203,
+      "loss": 1.0499,
+      "step": 5047
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25047524197955967,
+      "learning_rate": 0.00018061060314080658,
+      "loss": 1.1112,
+      "step": 5048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25460102742828555,
+      "learning_rate": 0.00018060124061436184,
+      "loss": 1.139,
+      "step": 5049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2684967059066573,
+      "learning_rate": 0.00018059187607082037,
+      "loss": 1.0121,
+      "step": 5050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2740879178050141,
+      "learning_rate": 0.00018058250951041656,
+      "loss": 1.2213,
+      "step": 5051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.23814111073074376,
+      "learning_rate": 0.0001805731409333848,
+      "loss": 0.946,
+      "step": 5052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2846550890405717,
+      "learning_rate": 0.00018056377033995959,
+      "loss": 1.0843,
+      "step": 5053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.26517227533600335,
+      "learning_rate": 0.00018055439773037536,
+      "loss": 1.0066,
+      "step": 5054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27053031387998794,
+      "learning_rate": 0.0001805450231048667,
+      "loss": 0.9459,
+      "step": 5055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3019675939625031,
+      "learning_rate": 0.00018053564646366822,
+      "loss": 1.2236,
+      "step": 5056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25798882391808686,
+      "learning_rate": 0.00018052626780701457,
+      "loss": 1.1254,
+      "step": 5057
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.27311813604505014,
+      "learning_rate": 0.00018051688713514047,
+      "loss": 1.0087,
+      "step": 5058
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24271338025668723,
+      "learning_rate": 0.00018050750444828067,
+      "loss": 1.0985,
+      "step": 5059
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28513204364807454,
+      "learning_rate": 0.00018049811974666996,
+      "loss": 0.968,
+      "step": 5060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2684639300431927,
+      "learning_rate": 0.00018048873303054324,
+      "loss": 1.0187,
+      "step": 5061
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28668337168219155,
+      "learning_rate": 0.00018047934430013535,
+      "loss": 1.1447,
+      "step": 5062
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2526134197715635,
+      "learning_rate": 0.0001804699535556813,
+      "loss": 1.1222,
+      "step": 5063
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2734155691633195,
+      "learning_rate": 0.0001804605607974161,
+      "loss": 1.0668,
+      "step": 5064
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.29009495050980677,
+      "learning_rate": 0.0001804511660255748,
+      "loss": 1.143,
+      "step": 5065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.24381730481982744,
+      "learning_rate": 0.0001804417692403925,
+      "loss": 1.2,
+      "step": 5066
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.28667698702842986,
+      "learning_rate": 0.00018043237044210438,
+      "loss": 0.9782,
+      "step": 5067
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.2290214469722657,
+      "learning_rate": 0.00018042296963094562,
+      "loss": 1.0481,
+      "step": 5068
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.25712013816908574,
+      "learning_rate": 0.00018041356680715152,
+      "loss": 1.0808,
+      "step": 5069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27191198243103304,
+      "learning_rate": 0.00018040416197095737,
+      "loss": 1.0839,
+      "step": 5070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2896020255383838,
+      "learning_rate": 0.00018039475512259855,
+      "loss": 1.0719,
+      "step": 5071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2722755649971534,
+      "learning_rate": 0.0001803853462623104,
+      "loss": 1.111,
+      "step": 5072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.296096353606665,
+      "learning_rate": 0.0001803759353903285,
+      "loss": 1.1522,
+      "step": 5073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2516486412186821,
+      "learning_rate": 0.0001803665225068883,
+      "loss": 1.0666,
+      "step": 5074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28162660247716065,
+      "learning_rate": 0.00018035710761222533,
+      "loss": 1.0581,
+      "step": 5075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2573369100267432,
+      "learning_rate": 0.00018034769070657524,
+      "loss": 1.0313,
+      "step": 5076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25964129622911775,
+      "learning_rate": 0.00018033827179017372,
+      "loss": 1.0417,
+      "step": 5077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2584869469852085,
+      "learning_rate": 0.00018032885086325645,
+      "loss": 1.0705,
+      "step": 5078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2803573003244443,
+      "learning_rate": 0.0001803194279260592,
+      "loss": 1.1134,
+      "step": 5079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25948244844278057,
+      "learning_rate": 0.00018031000297881778,
+      "loss": 1.1251,
+      "step": 5080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27477615836677105,
+      "learning_rate": 0.00018030057602176806,
+      "loss": 1.0987,
+      "step": 5081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2949596444499225,
+      "learning_rate": 0.00018029114705514596,
+      "loss": 1.1267,
+      "step": 5082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3126152488004548,
+      "learning_rate": 0.00018028171607918747,
+      "loss": 1.1043,
+      "step": 5083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3158944201979104,
+      "learning_rate": 0.00018027228309412853,
+      "loss": 1.1333,
+      "step": 5084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23721139976987007,
+      "learning_rate": 0.00018026284810020532,
+      "loss": 1.0958,
+      "step": 5085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2620591461537738,
+      "learning_rate": 0.00018025341109765384,
+      "loss": 1.0396,
+      "step": 5086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29187944723111464,
+      "learning_rate": 0.00018024397208671035,
+      "loss": 1.1689,
+      "step": 5087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2740567842867544,
+      "learning_rate": 0.000180234531067611,
+      "loss": 1.0742,
+      "step": 5088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28997111975839956,
+      "learning_rate": 0.00018022508804059207,
+      "loss": 1.1359,
+      "step": 5089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.264169160568814,
+      "learning_rate": 0.00018021564300588994,
+      "loss": 0.9856,
+      "step": 5090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.25799277848982,
+      "learning_rate": 0.0001802061959637409,
+      "loss": 1.0802,
+      "step": 5091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27789264617161874,
+      "learning_rate": 0.0001801967469143814,
+      "loss": 1.1495,
+      "step": 5092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.284905774011098,
+      "learning_rate": 0.0001801872958580479,
+      "loss": 1.1043,
+      "step": 5093
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.23840852278323477,
+      "learning_rate": 0.00018017784279497693,
+      "loss": 1.0545,
+      "step": 5094
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29365200369529876,
+      "learning_rate": 0.00018016838772540506,
+      "loss": 1.0454,
+      "step": 5095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2905391396194257,
+      "learning_rate": 0.0001801589306495689,
+      "loss": 1.0989,
+      "step": 5096
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30717926924064015,
+      "learning_rate": 0.00018014947156770513,
+      "loss": 1.0989,
+      "step": 5097
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24633770830565768,
+      "learning_rate": 0.00018014001048005044,
+      "loss": 0.9554,
+      "step": 5098
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.269877103345373,
+      "learning_rate": 0.00018013054738684166,
+      "loss": 1.1133,
+      "step": 5099
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2995678521690841,
+      "learning_rate": 0.00018012108228831556,
+      "loss": 1.1077,
+      "step": 5100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24440039782838777,
+      "learning_rate": 0.000180111615184709,
+      "loss": 1.0012,
+      "step": 5101
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27251554211773565,
+      "learning_rate": 0.00018010214607625894,
+      "loss": 1.1004,
+      "step": 5102
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2810534155604437,
+      "learning_rate": 0.0001800926749632023,
+      "loss": 1.0193,
+      "step": 5103
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27100726353528887,
+      "learning_rate": 0.0001800832018457762,
+      "loss": 1.0602,
+      "step": 5104
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26564758624041535,
+      "learning_rate": 0.00018007372672421756,
+      "loss": 1.079,
+      "step": 5105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2974717566929695,
+      "learning_rate": 0.00018006424959876363,
+      "loss": 1.1326,
+      "step": 5106
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30488571850565943,
+      "learning_rate": 0.00018005477046965153,
+      "loss": 1.0309,
+      "step": 5107
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2145246313949886,
+      "learning_rate": 0.0001800452893371185,
+      "loss": 1.0762,
+      "step": 5108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2636249075639135,
+      "learning_rate": 0.00018003580620140177,
+      "loss": 1.153,
+      "step": 5109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27782368506119703,
+      "learning_rate": 0.0001800263210627387,
+      "loss": 1.117,
+      "step": 5110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3007327930211106,
+      "learning_rate": 0.00018001683392136666,
+      "loss": 1.1193,
+      "step": 5111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28929437068723235,
+      "learning_rate": 0.00018000734477752306,
+      "loss": 1.0561,
+      "step": 5112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28549761815962377,
+      "learning_rate": 0.00017999785363144536,
+      "loss": 1.1016,
+      "step": 5113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2863390229819487,
+      "learning_rate": 0.0001799883604833711,
+      "loss": 1.1143,
+      "step": 5114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2758612911370492,
+      "learning_rate": 0.00017997886533353786,
+      "loss": 1.1496,
+      "step": 5115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26382558236784287,
+      "learning_rate": 0.00017996936818218324,
+      "loss": 1.0343,
+      "step": 5116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2695740841257572,
+      "learning_rate": 0.00017995986902954493,
+      "loss": 1.1589,
+      "step": 5117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.315794471443836,
+      "learning_rate": 0.00017995036787586064,
+      "loss": 0.9799,
+      "step": 5118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2707540246275962,
+      "learning_rate": 0.00017994086472136815,
+      "loss": 1.089,
+      "step": 5119
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3337912375773768,
+      "learning_rate": 0.0001799313595663053,
+      "loss": 1.0462,
+      "step": 5120
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30849482259896643,
+      "learning_rate": 0.0001799218524109099,
+      "loss": 0.9913,
+      "step": 5121
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2794905681957685,
+      "learning_rate": 0.00017991234325541995,
+      "loss": 1.0798,
+      "step": 5122
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26906629376110763,
+      "learning_rate": 0.00017990283210007335,
+      "loss": 0.9918,
+      "step": 5123
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2640810736886688,
+      "learning_rate": 0.00017989331894510818,
+      "loss": 1.0668,
+      "step": 5124
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3403000076240741,
+      "learning_rate": 0.0001798838037907625,
+      "loss": 1.1871,
+      "step": 5125
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29390271181136557,
+      "learning_rate": 0.00017987428663727441,
+      "loss": 1.0834,
+      "step": 5126
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29242522687517936,
+      "learning_rate": 0.00017986476748488214,
+      "loss": 0.9395,
+      "step": 5127
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30030389910750893,
+      "learning_rate": 0.00017985524633382381,
+      "loss": 1.0064,
+      "step": 5128
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26767260775313606,
+      "learning_rate": 0.00017984572318433778,
+      "loss": 1.2037,
+      "step": 5129
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2883783786958614,
+      "learning_rate": 0.00017983619803666235,
+      "loss": 1.0901,
+      "step": 5130
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2615336186142248,
+      "learning_rate": 0.00017982667089103588,
+      "loss": 1.1084,
+      "step": 5131
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2806392247590465,
+      "learning_rate": 0.0001798171417476968,
+      "loss": 1.0532,
+      "step": 5132
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2563049389592253,
+      "learning_rate": 0.0001798076106068836,
+      "loss": 1.0789,
+      "step": 5133
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.32787128373934044,
+      "learning_rate": 0.0001797980774688348,
+      "loss": 1.0862,
+      "step": 5134
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26980316995262976,
+      "learning_rate": 0.00017978854233378891,
+      "loss": 1.2006,
+      "step": 5135
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2359678414179539,
+      "learning_rate": 0.00017977900520198465,
+      "loss": 0.9747,
+      "step": 5136
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28300881314834403,
+      "learning_rate": 0.00017976946607366063,
+      "loss": 1.0696,
+      "step": 5137
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27366802597990486,
+      "learning_rate": 0.0001797599249490556,
+      "loss": 0.9662,
+      "step": 5138
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2714308944537401,
+      "learning_rate": 0.00017975038182840828,
+      "loss": 1.083,
+      "step": 5139
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2644809976461488,
+      "learning_rate": 0.00017974083671195757,
+      "loss": 1.081,
+      "step": 5140
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2813538147596539,
+      "learning_rate": 0.0001797312895999423,
+      "loss": 1.1335,
+      "step": 5141
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2910923380431433,
+      "learning_rate": 0.0001797217404926014,
+      "loss": 1.0232,
+      "step": 5142
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26419502096815084,
+      "learning_rate": 0.00017971218939017382,
+      "loss": 1.106,
+      "step": 5143
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30598831197367454,
+      "learning_rate": 0.00017970263629289864,
+      "loss": 1.1303,
+      "step": 5144
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24615417333770134,
+      "learning_rate": 0.00017969308120101488,
+      "loss": 0.969,
+      "step": 5145
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2592721166455555,
+      "learning_rate": 0.00017968352411476166,
+      "loss": 1.0971,
+      "step": 5146
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2868843022304741,
+      "learning_rate": 0.00017967396503437816,
+      "loss": 0.986,
+      "step": 5147
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2936385306501026,
+      "learning_rate": 0.00017966440396010366,
+      "loss": 1.0833,
+      "step": 5148
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2947617623317851,
+      "learning_rate": 0.00017965484089217735,
+      "loss": 1.0711,
+      "step": 5149
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2628043083883782,
+      "learning_rate": 0.0001796452758308386,
+      "loss": 1.1146,
+      "step": 5150
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.31452969252198026,
+      "learning_rate": 0.00017963570877632676,
+      "loss": 1.1144,
+      "step": 5151
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2596821467826261,
+      "learning_rate": 0.00017962613972888125,
+      "loss": 1.0938,
+      "step": 5152
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27739044259675993,
+      "learning_rate": 0.00017961656868874156,
+      "loss": 1.0211,
+      "step": 5153
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2937922180821596,
+      "learning_rate": 0.0001796069956561472,
+      "loss": 1.0647,
+      "step": 5154
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27965187717460716,
+      "learning_rate": 0.00017959742063133774,
+      "loss": 1.0117,
+      "step": 5155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2840025373723606,
+      "learning_rate": 0.00017958784361455282,
+      "loss": 0.9706,
+      "step": 5156
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3178568601522771,
+      "learning_rate": 0.00017957826460603205,
+      "loss": 1.018,
+      "step": 5157
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2860032058314914,
+      "learning_rate": 0.00017956868360601526,
+      "loss": 1.1566,
+      "step": 5158
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28443581551498176,
+      "learning_rate": 0.00017955910061474213,
+      "loss": 1.0321,
+      "step": 5159
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27187438960276544,
+      "learning_rate": 0.0001795495156324525,
+      "loss": 1.0505,
+      "step": 5160
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29288869429900377,
+      "learning_rate": 0.00017953992865938622,
+      "loss": 1.0511,
+      "step": 5161
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.30056027600724244,
+      "learning_rate": 0.00017953033969578326,
+      "loss": 1.1062,
+      "step": 5162
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26821431519549427,
+      "learning_rate": 0.00017952074874188356,
+      "loss": 1.092,
+      "step": 5163
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24906638557874644,
+      "learning_rate": 0.00017951115579792717,
+      "loss": 1.1022,
+      "step": 5164
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26309218013010355,
+      "learning_rate": 0.0001795015608641541,
+      "loss": 1.0265,
+      "step": 5165
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.26268394260057937,
+      "learning_rate": 0.00017949196394080453,
+      "loss": 1.0653,
+      "step": 5166
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2896832205130746,
+      "learning_rate": 0.00017948236502811859,
+      "loss": 1.0854,
+      "step": 5167
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2625237394571958,
+      "learning_rate": 0.00017947276412633652,
+      "loss": 0.9619,
+      "step": 5168
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.27986610289816,
+      "learning_rate": 0.0001794631612356986,
+      "loss": 1.1716,
+      "step": 5169
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.28211340979754457,
+      "learning_rate": 0.0001794535563564451,
+      "loss": 1.1021,
+      "step": 5170
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.24926141918189934,
+      "learning_rate": 0.00017944394948881642,
+      "loss": 1.056,
+      "step": 5171
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2869730653075155,
+      "learning_rate": 0.00017943434063305298,
+      "loss": 0.9492,
+      "step": 5172
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.29187379432960536,
+      "learning_rate": 0.00017942472978939525,
+      "loss": 1.1833,
+      "step": 5173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28973741108012707,
+      "learning_rate": 0.00017941511695808372,
+      "loss": 1.0489,
+      "step": 5174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27689121010931406,
+      "learning_rate": 0.000179405502139359,
+      "loss": 1.1769,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3067616467203378,
+      "learning_rate": 0.00017939588533346168,
+      "loss": 1.1473,
+      "step": 5176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2757139090006968,
+      "learning_rate": 0.0001793862665406324,
+      "loss": 1.1183,
+      "step": 5177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27386836919563157,
+      "learning_rate": 0.00017937664576111198,
+      "loss": 1.0154,
+      "step": 5178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2918051909413474,
+      "learning_rate": 0.00017936702299514105,
+      "loss": 1.1863,
+      "step": 5179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2964159774568591,
+      "learning_rate": 0.00017935739824296052,
+      "loss": 1.0043,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31639186034892197,
+      "learning_rate": 0.0001793477715048112,
+      "loss": 1.1946,
+      "step": 5181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3060832768680089,
+      "learning_rate": 0.00017933814278093407,
+      "loss": 0.9377,
+      "step": 5182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3453844813102695,
+      "learning_rate": 0.00017932851207157002,
+      "loss": 1.0465,
+      "step": 5183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25905466276473943,
+      "learning_rate": 0.0001793188793769601,
+      "loss": 1.1672,
+      "step": 5184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29227551232364024,
+      "learning_rate": 0.00017930924469734537,
+      "loss": 1.0898,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2513616979563531,
+      "learning_rate": 0.00017929960803296697,
+      "loss": 1.0656,
+      "step": 5186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.275133905403024,
+      "learning_rate": 0.00017928996938406603,
+      "loss": 1.002,
+      "step": 5187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28951906430321944,
+      "learning_rate": 0.00017928032875088375,
+      "loss": 1.1197,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26163177751335326,
+      "learning_rate": 0.00017927068613366145,
+      "loss": 1.1226,
+      "step": 5189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.308746805614135,
+      "learning_rate": 0.00017926104153264042,
+      "loss": 1.0741,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29075350653382204,
+      "learning_rate": 0.00017925139494806198,
+      "loss": 1.1404,
+      "step": 5191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2760956753624024,
+      "learning_rate": 0.0001792417463801676,
+      "loss": 1.132,
+      "step": 5192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.30467872599141255,
+      "learning_rate": 0.0001792320958291987,
+      "loss": 1.1227,
+      "step": 5193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.269599432070437,
+      "learning_rate": 0.0001792224432953968,
+      "loss": 1.1915,
+      "step": 5194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2540692313113988,
+      "learning_rate": 0.00017921278877900348,
+      "loss": 1.1215,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25608501554671015,
+      "learning_rate": 0.0001792031322802603,
+      "loss": 1.0529,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2646934249716702,
+      "learning_rate": 0.00017919347379940904,
+      "loss": 1.1269,
+      "step": 5197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2458734180042231,
+      "learning_rate": 0.00017918381333669126,
+      "loss": 1.0294,
+      "step": 5198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2629321130779039,
+      "learning_rate": 0.0001791741508923488,
+      "loss": 1.0545,
+      "step": 5199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29932796860263766,
+      "learning_rate": 0.00017916448646662346,
+      "loss": 1.1029,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2794636942689881,
+      "learning_rate": 0.00017915482005975708,
+      "loss": 0.9605,
+      "step": 5201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2805023432544276,
+      "learning_rate": 0.00017914515167199158,
+      "loss": 1.0897,
+      "step": 5202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25784189464219454,
+      "learning_rate": 0.00017913548130356894,
+      "loss": 1.013,
+      "step": 5203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31553870897854386,
+      "learning_rate": 0.00017912580895473114,
+      "loss": 1.1689,
+      "step": 5204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31827443288264134,
+      "learning_rate": 0.00017911613462572024,
+      "loss": 1.0521,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3025879844146086,
+      "learning_rate": 0.00017910645831677836,
+      "loss": 0.997,
+      "step": 5206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26057117438515826,
+      "learning_rate": 0.0001790967800281476,
+      "loss": 1.1074,
+      "step": 5207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2611725357612145,
+      "learning_rate": 0.00017908709976007024,
+      "loss": 1.0784,
+      "step": 5208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2344376304389919,
+      "learning_rate": 0.0001790774175127885,
+      "loss": 1.0707,
+      "step": 5209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29122733484693347,
+      "learning_rate": 0.00017906773328654472,
+      "loss": 1.1213,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2759187534926237,
+      "learning_rate": 0.00017905804708158118,
+      "loss": 1.0693,
+      "step": 5211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2777254558995927,
+      "learning_rate": 0.00017904835889814033,
+      "loss": 1.0366,
+      "step": 5212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2714206402525035,
+      "learning_rate": 0.00017903866873646463,
+      "loss": 1.1107,
+      "step": 5213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3012099609994761,
+      "learning_rate": 0.0001790289765967966,
+      "loss": 1.0848,
+      "step": 5214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3032845339156945,
+      "learning_rate": 0.00017901928247937872,
+      "loss": 1.1453,
+      "step": 5215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26396959858562596,
+      "learning_rate": 0.00017900958638445365,
+      "loss": 1.1375,
+      "step": 5216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2727644440686032,
+      "learning_rate": 0.00017899988831226402,
+      "loss": 1.0937,
+      "step": 5217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3053107974870675,
+      "learning_rate": 0.00017899018826305252,
+      "loss": 1.1792,
+      "step": 5218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26793488874015303,
+      "learning_rate": 0.00017898048623706195,
+      "loss": 1.1571,
+      "step": 5219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2597411208953679,
+      "learning_rate": 0.00017897078223453504,
+      "loss": 1.052,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.31195341891857165,
+      "learning_rate": 0.0001789610762557147,
+      "loss": 1.0902,
+      "step": 5221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2744676529459866,
+      "learning_rate": 0.0001789513683008438,
+      "loss": 1.0829,
+      "step": 5222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26093271474594537,
+      "learning_rate": 0.00017894165837016528,
+      "loss": 0.9847,
+      "step": 5223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.290962254361299,
+      "learning_rate": 0.00017893194646392214,
+      "loss": 1.0686,
+      "step": 5224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26375436627341603,
+      "learning_rate": 0.00017892223258235746,
+      "loss": 1.1454,
+      "step": 5225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2646502633063915,
+      "learning_rate": 0.00017891251672571428,
+      "loss": 1.0035,
+      "step": 5226
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.130812406539917,
+      "eval_runtime": 4229.0908,
+      "eval_samples_per_second": 19.772,
+      "eval_steps_per_second": 2.472,
+      "step": 5226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27086856649700053,
+      "learning_rate": 0.00017890279889423577,
+      "loss": 1.2004,
+      "step": 5227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26907032517761026,
+      "learning_rate": 0.00017889307908816514,
+      "loss": 1.1269,
+      "step": 5228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2559621504798168,
+      "learning_rate": 0.00017888335730774563,
+      "loss": 1.0099,
+      "step": 5229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2886629743550709,
+      "learning_rate": 0.00017887363355322054,
+      "loss": 1.0698,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3107523248743581,
+      "learning_rate": 0.00017886390782483318,
+      "loss": 1.086,
+      "step": 5231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2731347287540774,
+      "learning_rate": 0.00017885418012282696,
+      "loss": 1.1274,
+      "step": 5232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3032030305104551,
+      "learning_rate": 0.00017884445044744532,
+      "loss": 1.0157,
+      "step": 5233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29204175835395896,
+      "learning_rate": 0.00017883471879893176,
+      "loss": 1.0925,
+      "step": 5234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27875381267269567,
+      "learning_rate": 0.00017882498517752984,
+      "loss": 1.0,
+      "step": 5235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2744649218236222,
+      "learning_rate": 0.00017881524958348311,
+      "loss": 1.1309,
+      "step": 5236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2638597475413694,
+      "learning_rate": 0.00017880551201703522,
+      "loss": 1.1368,
+      "step": 5237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29740497162072044,
+      "learning_rate": 0.00017879577247842984,
+      "loss": 1.163,
+      "step": 5238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2792113343183714,
+      "learning_rate": 0.00017878603096791078,
+      "loss": 0.9453,
+      "step": 5239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3252881204864866,
+      "learning_rate": 0.00017877628748572176,
+      "loss": 0.9687,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2993197408644714,
+      "learning_rate": 0.00017876654203210666,
+      "loss": 1.1889,
+      "step": 5241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25588679975602296,
+      "learning_rate": 0.0001787567946073093,
+      "loss": 1.0073,
+      "step": 5242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25844292628782845,
+      "learning_rate": 0.00017874704521157368,
+      "loss": 1.1023,
+      "step": 5243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2701484124438951,
+      "learning_rate": 0.00017873729384514374,
+      "loss": 1.1446,
+      "step": 5244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28163616631177285,
+      "learning_rate": 0.00017872754050826358,
+      "loss": 1.1451,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24357842637206142,
+      "learning_rate": 0.00017871778520117722,
+      "loss": 1.0879,
+      "step": 5246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2854538423239079,
+      "learning_rate": 0.0001787080279241288,
+      "loss": 1.1346,
+      "step": 5247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28966646255088585,
+      "learning_rate": 0.00017869826867736253,
+      "loss": 1.1342,
+      "step": 5248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.269137927932761,
+      "learning_rate": 0.0001786885074611226,
+      "loss": 1.2118,
+      "step": 5249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24965315368924026,
+      "learning_rate": 0.00017867874427565336,
+      "loss": 1.0445,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3268586937562193,
+      "learning_rate": 0.00017866897912119907,
+      "loss": 1.0391,
+      "step": 5251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2628697700403388,
+      "learning_rate": 0.00017865921199800415,
+      "loss": 1.0788,
+      "step": 5252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.25466346365515374,
+      "learning_rate": 0.00017864944290631301,
+      "loss": 1.1108,
+      "step": 5253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2785430342878654,
+      "learning_rate": 0.00017863967184637014,
+      "loss": 0.9847,
+      "step": 5254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2821653592665947,
+      "learning_rate": 0.00017862989881842003,
+      "loss": 1.1659,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28381789349374786,
+      "learning_rate": 0.0001786201238227073,
+      "loss": 1.0368,
+      "step": 5256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28526194687906037,
+      "learning_rate": 0.00017861034685947658,
+      "loss": 1.0789,
+      "step": 5257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2646173880017078,
+      "learning_rate": 0.0001786005679289725,
+      "loss": 1.0564,
+      "step": 5258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.23710824870546326,
+      "learning_rate": 0.0001785907870314398,
+      "loss": 0.9625,
+      "step": 5259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.26259679119084095,
+      "learning_rate": 0.0001785810041671233,
+      "loss": 1.1021,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2709420502635169,
+      "learning_rate": 0.00017857121933626777,
+      "loss": 1.0062,
+      "step": 5261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2613609827266697,
+      "learning_rate": 0.0001785614325391181,
+      "loss": 1.0678,
+      "step": 5262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28813131561794453,
+      "learning_rate": 0.00017855164377591918,
+      "loss": 0.9172,
+      "step": 5263
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27234699286481917,
+      "learning_rate": 0.000178541853046916,
+      "loss": 1.1471,
+      "step": 5264
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2506386238760721,
+      "learning_rate": 0.0001785320603523536,
+      "loss": 1.0259,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2540092486293724,
+      "learning_rate": 0.00017852226569247708,
+      "loss": 1.0877,
+      "step": 5266
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.29602747120892153,
+      "learning_rate": 0.00017851246906753145,
+      "loss": 1.109,
+      "step": 5267
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2992977885795131,
+      "learning_rate": 0.00017850267047776197,
+      "loss": 1.0069,
+      "step": 5268
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28770619817428295,
+      "learning_rate": 0.0001784928699234138,
+      "loss": 1.0456,
+      "step": 5269
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2621648485300015,
+      "learning_rate": 0.00017848306740473227,
+      "loss": 1.0159,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.27475331658348695,
+      "learning_rate": 0.00017847326292196261,
+      "loss": 1.026,
+      "step": 5271
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2883171040107314,
+      "learning_rate": 0.00017846345647535026,
+      "loss": 1.0636,
+      "step": 5272
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2679836068828528,
+      "learning_rate": 0.0001784536480651406,
+      "loss": 1.1411,
+      "step": 5273
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2718315189359781,
+      "learning_rate": 0.00017844383769157905,
+      "loss": 1.0714,
+      "step": 5274
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2679104057163504,
+      "learning_rate": 0.0001784340253549112,
+      "loss": 1.0169,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.28568786629626025,
+      "learning_rate": 0.00017842421105538256,
+      "loss": 1.0771,
+      "step": 5276
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.2548173055279772,
+      "learning_rate": 0.00017841439479323877,
+      "loss": 1.1359,
+      "step": 5277
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.24999339748919044,
+      "learning_rate": 0.00017840457656872544,
+      "loss": 1.0132,
+      "step": 5278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3003717665696886,
+      "learning_rate": 0.00017839475638208832,
+      "loss": 1.0838,
+      "step": 5279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31887733553865744,
+      "learning_rate": 0.00017838493423357314,
+      "loss": 1.0885,
+      "step": 5280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2625967483202807,
+      "learning_rate": 0.00017837511012342572,
+      "loss": 1.1534,
+      "step": 5281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31614913625651914,
+      "learning_rate": 0.0001783652840518919,
+      "loss": 1.0403,
+      "step": 5282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26033042065792134,
+      "learning_rate": 0.00017835545601921764,
+      "loss": 1.0581,
+      "step": 5283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2856070132728536,
+      "learning_rate": 0.00017834562602564883,
+      "loss": 1.0107,
+      "step": 5284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2536339900536725,
+      "learning_rate": 0.00017833579407143147,
+      "loss": 0.9525,
+      "step": 5285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26624532509069715,
+      "learning_rate": 0.00017832596015681165,
+      "loss": 1.0309,
+      "step": 5286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26639398423968824,
+      "learning_rate": 0.00017831612428203543,
+      "loss": 1.103,
+      "step": 5287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2854143682160096,
+      "learning_rate": 0.00017830628644734898,
+      "loss": 1.1898,
+      "step": 5288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2647981085190368,
+      "learning_rate": 0.0001782964466529985,
+      "loss": 1.0559,
+      "step": 5289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2692642057535311,
+      "learning_rate": 0.00017828660489923025,
+      "loss": 1.1121,
+      "step": 5290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2790653912792434,
+      "learning_rate": 0.00017827676118629054,
+      "loss": 1.119,
+      "step": 5291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2731285711487075,
+      "learning_rate": 0.00017826691551442564,
+      "loss": 1.1326,
+      "step": 5292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27848520893646905,
+      "learning_rate": 0.000178257067883882,
+      "loss": 1.0876,
+      "step": 5293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27766691943147104,
+      "learning_rate": 0.00017824721829490608,
+      "loss": 1.0075,
+      "step": 5294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24812476086316548,
+      "learning_rate": 0.00017823736674774432,
+      "loss": 0.9859,
+      "step": 5295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24136690327253105,
+      "learning_rate": 0.00017822751324264328,
+      "loss": 0.9295,
+      "step": 5296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.22887560800126114,
+      "learning_rate": 0.00017821765777984957,
+      "loss": 1.1376,
+      "step": 5297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2746004567986452,
+      "learning_rate": 0.0001782078003596098,
+      "loss": 1.0057,
+      "step": 5298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2425791039008265,
+      "learning_rate": 0.0001781979409821707,
+      "loss": 1.1405,
+      "step": 5299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2718956254441623,
+      "learning_rate": 0.00017818807964777898,
+      "loss": 1.0884,
+      "step": 5300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2616875209648741,
+      "learning_rate": 0.0001781782163566814,
+      "loss": 1.0947,
+      "step": 5301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24842391055912055,
+      "learning_rate": 0.00017816835110912485,
+      "loss": 0.9207,
+      "step": 5302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2747180515251053,
+      "learning_rate": 0.00017815848390535617,
+      "loss": 0.9877,
+      "step": 5303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2875077951068339,
+      "learning_rate": 0.00017814861474562232,
+      "loss": 1.0289,
+      "step": 5304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2763116859597204,
+      "learning_rate": 0.00017813874363017027,
+      "loss": 1.0357,
+      "step": 5305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28069395829776694,
+      "learning_rate": 0.00017812887055924703,
+      "loss": 1.1048,
+      "step": 5306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30047513289045635,
+      "learning_rate": 0.00017811899553309975,
+      "loss": 1.0333,
+      "step": 5307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2650398736977853,
+      "learning_rate": 0.00017810911855197547,
+      "loss": 1.0565,
+      "step": 5308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24640154980602844,
+      "learning_rate": 0.0001780992396161214,
+      "loss": 1.1352,
+      "step": 5309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28475244026032503,
+      "learning_rate": 0.00017808935872578482,
+      "loss": 1.0479,
+      "step": 5310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2756255403166274,
+      "learning_rate": 0.00017807947588121295,
+      "loss": 1.0636,
+      "step": 5311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3137323153824364,
+      "learning_rate": 0.00017806959108265308,
+      "loss": 1.2068,
+      "step": 5312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2842470117681127,
+      "learning_rate": 0.00017805970433035266,
+      "loss": 1.229,
+      "step": 5313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2981813888336136,
+      "learning_rate": 0.00017804981562455908,
+      "loss": 1.1881,
+      "step": 5314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28279186295473696,
+      "learning_rate": 0.00017803992496551982,
+      "loss": 1.0685,
+      "step": 5315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2702456382826126,
+      "learning_rate": 0.0001780300323534824,
+      "loss": 0.9998,
+      "step": 5316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.276914892943855,
+      "learning_rate": 0.00017802013778869436,
+      "loss": 1.0531,
+      "step": 5317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27960841961549454,
+      "learning_rate": 0.0001780102412714033,
+      "loss": 1.1359,
+      "step": 5318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25797705830410184,
+      "learning_rate": 0.00017800034280185699,
+      "loss": 1.0185,
+      "step": 5319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2857057108150563,
+      "learning_rate": 0.00017799044238030307,
+      "loss": 1.0631,
+      "step": 5320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25584721058778875,
+      "learning_rate": 0.0001779805400069893,
+      "loss": 0.987,
+      "step": 5321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2325654159439427,
+      "learning_rate": 0.0001779706356821635,
+      "loss": 1.0651,
+      "step": 5322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2970569353859898,
+      "learning_rate": 0.00017796072940607353,
+      "loss": 1.13,
+      "step": 5323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2929022741608403,
+      "learning_rate": 0.00017795082117896734,
+      "loss": 1.0719,
+      "step": 5324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2609795745084068,
+      "learning_rate": 0.00017794091100109283,
+      "loss": 1.0067,
+      "step": 5325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30762587611696396,
+      "learning_rate": 0.0001779309988726981,
+      "loss": 1.147,
+      "step": 5326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2661981298944252,
+      "learning_rate": 0.00017792108479403106,
+      "loss": 1.0734,
+      "step": 5327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2764342632822514,
+      "learning_rate": 0.00017791116876533994,
+      "loss": 1.1101,
+      "step": 5328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24938502770088783,
+      "learning_rate": 0.00017790125078687288,
+      "loss": 1.0666,
+      "step": 5329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3336408049693267,
+      "learning_rate": 0.000177891330858878,
+      "loss": 1.0902,
+      "step": 5330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3218184473435925,
+      "learning_rate": 0.00017788140898160367,
+      "loss": 1.0904,
+      "step": 5331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2789740899659255,
+      "learning_rate": 0.0001778714851552981,
+      "loss": 1.0243,
+      "step": 5332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.266597050786924,
+      "learning_rate": 0.00017786155938020968,
+      "loss": 1.0069,
+      "step": 5333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28223064418935384,
+      "learning_rate": 0.00017785163165658685,
+      "loss": 1.1401,
+      "step": 5334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.272562171370601,
+      "learning_rate": 0.00017784170198467797,
+      "loss": 1.0984,
+      "step": 5335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.23364083669723149,
+      "learning_rate": 0.00017783177036473155,
+      "loss": 1.0585,
+      "step": 5336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2423799752683352,
+      "learning_rate": 0.0001778218367969962,
+      "loss": 0.9586,
+      "step": 5337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28056034525742624,
+      "learning_rate": 0.00017781190128172045,
+      "loss": 1.0374,
+      "step": 5338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31575138951832793,
+      "learning_rate": 0.000177801963819153,
+      "loss": 1.1025,
+      "step": 5339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2756564243357627,
+      "learning_rate": 0.00017779202440954247,
+      "loss": 1.073,
+      "step": 5340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2855725418706688,
+      "learning_rate": 0.00017778208305313766,
+      "loss": 0.9841,
+      "step": 5341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2622842600538124,
+      "learning_rate": 0.00017777213975018734,
+      "loss": 0.9923,
+      "step": 5342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27868325397356525,
+      "learning_rate": 0.00017776219450094032,
+      "loss": 1.0104,
+      "step": 5343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2910306279110282,
+      "learning_rate": 0.00017775224730564554,
+      "loss": 0.9543,
+      "step": 5344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30979629764973976,
+      "learning_rate": 0.0001777422981645519,
+      "loss": 1.1145,
+      "step": 5345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.30299227741795537,
+      "learning_rate": 0.00017773234707790838,
+      "loss": 1.0723,
+      "step": 5346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2925383103599133,
+      "learning_rate": 0.00017772239404596402,
+      "loss": 1.1494,
+      "step": 5347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25939612879354135,
+      "learning_rate": 0.00017771243906896793,
+      "loss": 1.0362,
+      "step": 5348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.292894121185352,
+      "learning_rate": 0.00017770248214716918,
+      "loss": 1.0974,
+      "step": 5349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2934306234314348,
+      "learning_rate": 0.000177692523280817,
+      "loss": 1.0471,
+      "step": 5350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2620038040723942,
+      "learning_rate": 0.0001776825624701606,
+      "loss": 1.05,
+      "step": 5351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.258906223183488,
+      "learning_rate": 0.00017767259971544923,
+      "loss": 1.0016,
+      "step": 5352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2693727580254897,
+      "learning_rate": 0.00017766263501693222,
+      "loss": 1.1004,
+      "step": 5353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2656780441277641,
+      "learning_rate": 0.000177652668374859,
+      "loss": 1.1314,
+      "step": 5354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26233985901713786,
+      "learning_rate": 0.00017764269978947893,
+      "loss": 1.0327,
+      "step": 5355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2816892908241542,
+      "learning_rate": 0.00017763272926104152,
+      "loss": 1.1384,
+      "step": 5356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.29263486374255826,
+      "learning_rate": 0.00017762275678979625,
+      "loss": 1.1649,
+      "step": 5357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25054193247048423,
+      "learning_rate": 0.00017761278237599272,
+      "loss": 0.9422,
+      "step": 5358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2728514545749119,
+      "learning_rate": 0.00017760280601988052,
+      "loss": 1.1264,
+      "step": 5359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2787897298192036,
+      "learning_rate": 0.00017759282772170933,
+      "loss": 1.069,
+      "step": 5360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28051149945852427,
+      "learning_rate": 0.00017758284748172889,
+      "loss": 1.0434,
+      "step": 5361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2434867215928191,
+      "learning_rate": 0.0001775728653001889,
+      "loss": 1.094,
+      "step": 5362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.298894712000934,
+      "learning_rate": 0.00017756288117733922,
+      "loss": 1.0858,
+      "step": 5363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2989152889611858,
+      "learning_rate": 0.00017755289511342968,
+      "loss": 1.0862,
+      "step": 5364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.26061563505956564,
+      "learning_rate": 0.0001775429071087102,
+      "loss": 1.1229,
+      "step": 5365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3078317769181313,
+      "learning_rate": 0.00017753291716343075,
+      "loss": 1.0443,
+      "step": 5366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.271922412205937,
+      "learning_rate": 0.00017752292527784132,
+      "loss": 1.0569,
+      "step": 5367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2981868268132513,
+      "learning_rate": 0.00017751293145219194,
+      "loss": 1.1245,
+      "step": 5368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28791587214094305,
+      "learning_rate": 0.00017750293568673275,
+      "loss": 1.1032,
+      "step": 5369
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2770400531641322,
+      "learning_rate": 0.00017749293798171388,
+      "loss": 1.1548,
+      "step": 5370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31035234976544707,
+      "learning_rate": 0.00017748293833738554,
+      "loss": 1.0248,
+      "step": 5371
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.32563997547553536,
+      "learning_rate": 0.000177472936753998,
+      "loss": 1.0142,
+      "step": 5372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2965449735119415,
+      "learning_rate": 0.0001774629332318015,
+      "loss": 1.074,
+      "step": 5373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.24827461718041824,
+      "learning_rate": 0.00017745292777104638,
+      "loss": 1.0601,
+      "step": 5374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28164100400780073,
+      "learning_rate": 0.00017744292037198312,
+      "loss": 1.1411,
+      "step": 5375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.255726016665569,
+      "learning_rate": 0.00017743291103486207,
+      "loss": 1.0318,
+      "step": 5376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.25998427104390703,
+      "learning_rate": 0.0001774228997599338,
+      "loss": 1.0986,
+      "step": 5377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28575851348545084,
+      "learning_rate": 0.00017741288654744874,
+      "loss": 1.0325,
+      "step": 5378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.31525644009615983,
+      "learning_rate": 0.0001774028713976576,
+      "loss": 1.0881,
+      "step": 5379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.27772393103651905,
+      "learning_rate": 0.00017739285431081093,
+      "loss": 1.0819,
+      "step": 5380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.28458960847371934,
+      "learning_rate": 0.00017738283528715944,
+      "loss": 1.053,
+      "step": 5381
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2579176716986825,
+      "learning_rate": 0.00017737281432695387,
+      "loss": 1.0221,
+      "step": 5382
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.2682071919326544,
+      "learning_rate": 0.000177362791430445,
+      "loss": 1.0632,
+      "step": 5383
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2817647505413598,
+      "learning_rate": 0.00017735276659788365,
+      "loss": 1.0892,
+      "step": 5384
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27370042339496314,
+      "learning_rate": 0.0001773427398295207,
+      "loss": 1.1051,
+      "step": 5385
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29120157753379705,
+      "learning_rate": 0.00017733271112560707,
+      "loss": 1.057,
+      "step": 5386
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.274235384389006,
+      "learning_rate": 0.00017732268048639376,
+      "loss": 1.1645,
+      "step": 5387
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2554060673031894,
+      "learning_rate": 0.00017731264791213177,
+      "loss": 1.1206,
+      "step": 5388
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29038444135208097,
+      "learning_rate": 0.00017730261340307216,
+      "loss": 1.0847,
+      "step": 5389
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2927716873164211,
+      "learning_rate": 0.00017729257695946608,
+      "loss": 1.1842,
+      "step": 5390
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28575627634335404,
+      "learning_rate": 0.00017728253858156467,
+      "loss": 0.9855,
+      "step": 5391
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2741064932408415,
+      "learning_rate": 0.0001772724982696192,
+      "loss": 1.0681,
+      "step": 5392
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3221555396964685,
+      "learning_rate": 0.00017726245602388087,
+      "loss": 1.1304,
+      "step": 5393
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29940916808744256,
+      "learning_rate": 0.00017725241184460101,
+      "loss": 1.1972,
+      "step": 5394
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2555325051925566,
+      "learning_rate": 0.000177242365732031,
+      "loss": 1.0663,
+      "step": 5395
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29478956996412253,
+      "learning_rate": 0.00017723231768642227,
+      "loss": 1.1792,
+      "step": 5396
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2691856309392673,
+      "learning_rate": 0.0001772222677080262,
+      "loss": 1.0935,
+      "step": 5397
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26554481426737436,
+      "learning_rate": 0.00017721221579709438,
+      "loss": 1.1013,
+      "step": 5398
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2974199816061973,
+      "learning_rate": 0.00017720216195387834,
+      "loss": 1.1026,
+      "step": 5399
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25988903587070844,
+      "learning_rate": 0.00017719210617862967,
+      "loss": 0.9989,
+      "step": 5400
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2644737230686996,
+      "learning_rate": 0.00017718204847160004,
+      "loss": 1.1928,
+      "step": 5401
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.321621116578329,
+      "learning_rate": 0.0001771719888330411,
+      "loss": 1.0699,
+      "step": 5402
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3117270874799254,
+      "learning_rate": 0.00017716192726320468,
+      "loss": 1.0265,
+      "step": 5403
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26397074260532366,
+      "learning_rate": 0.0001771518637623425,
+      "loss": 1.0595,
+      "step": 5404
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26070999892086566,
+      "learning_rate": 0.00017714179833070646,
+      "loss": 0.9732,
+      "step": 5405
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29516458483721847,
+      "learning_rate": 0.00017713173096854846,
+      "loss": 1.244,
+      "step": 5406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28006511868535894,
+      "learning_rate": 0.0001771216616761204,
+      "loss": 1.0229,
+      "step": 5407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2785160995661278,
+      "learning_rate": 0.0001771115904536743,
+      "loss": 1.0974,
+      "step": 5408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3039976285784444,
+      "learning_rate": 0.00017710151730146215,
+      "loss": 1.1096,
+      "step": 5409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2725502855772708,
+      "learning_rate": 0.0001770914422197361,
+      "loss": 1.0443,
+      "step": 5410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2968618423089058,
+      "learning_rate": 0.00017708136520874822,
+      "loss": 1.0383,
+      "step": 5411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2787402009547091,
+      "learning_rate": 0.00017707128626875078,
+      "loss": 1.1659,
+      "step": 5412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2873480043549193,
+      "learning_rate": 0.00017706120539999595,
+      "loss": 0.9287,
+      "step": 5413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27057828580034904,
+      "learning_rate": 0.00017705112260273602,
+      "loss": 0.9655,
+      "step": 5414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2705987981572107,
+      "learning_rate": 0.00017704103787722332,
+      "loss": 1.1033,
+      "step": 5415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2883167160040527,
+      "learning_rate": 0.00017703095122371024,
+      "loss": 1.1945,
+      "step": 5416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28040572260172153,
+      "learning_rate": 0.00017702086264244918,
+      "loss": 1.1136,
+      "step": 5417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23953916000277492,
+      "learning_rate": 0.0001770107721336926,
+      "loss": 1.127,
+      "step": 5418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2666697798476311,
+      "learning_rate": 0.0001770006796976931,
+      "loss": 1.1178,
+      "step": 5419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26595583947693385,
+      "learning_rate": 0.00017699058533470318,
+      "loss": 1.1537,
+      "step": 5420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3018520026520265,
+      "learning_rate": 0.00017698048904497547,
+      "loss": 1.0672,
+      "step": 5421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2585809987651865,
+      "learning_rate": 0.00017697039082876264,
+      "loss": 0.9036,
+      "step": 5422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30520095394033386,
+      "learning_rate": 0.0001769602906863174,
+      "loss": 1.1266,
+      "step": 5423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26703193999809494,
+      "learning_rate": 0.00017695018861789254,
+      "loss": 1.0113,
+      "step": 5424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2665069045940571,
+      "learning_rate": 0.00017694008462374082,
+      "loss": 1.1435,
+      "step": 5425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3066957091926963,
+      "learning_rate": 0.00017692997870411513,
+      "loss": 1.1477,
+      "step": 5426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3060426948221189,
+      "learning_rate": 0.0001769198708592684,
+      "loss": 1.0636,
+      "step": 5427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31035289985606923,
+      "learning_rate": 0.00017690976108945353,
+      "loss": 1.1531,
+      "step": 5428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27105532382356223,
+      "learning_rate": 0.00017689964939492358,
+      "loss": 1.1738,
+      "step": 5429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30441191653566924,
+      "learning_rate": 0.00017688953577593158,
+      "loss": 1.272,
+      "step": 5430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30963973002090833,
+      "learning_rate": 0.0001768794202327306,
+      "loss": 1.2062,
+      "step": 5431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27923640796457394,
+      "learning_rate": 0.0001768693027655738,
+      "loss": 1.1393,
+      "step": 5432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2714745780737401,
+      "learning_rate": 0.00017685918337471442,
+      "loss": 1.0229,
+      "step": 5433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2746763330279501,
+      "learning_rate": 0.00017684906206040567,
+      "loss": 1.025,
+      "step": 5434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25420475121865566,
+      "learning_rate": 0.0001768389388229008,
+      "loss": 1.0173,
+      "step": 5435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23858794911194617,
+      "learning_rate": 0.00017682881366245322,
+      "loss": 1.0893,
+      "step": 5436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30224482212537046,
+      "learning_rate": 0.0001768186865793163,
+      "loss": 1.1222,
+      "step": 5437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23278958691465643,
+      "learning_rate": 0.00017680855757374345,
+      "loss": 1.0458,
+      "step": 5438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26428873295787486,
+      "learning_rate": 0.0001767984266459882,
+      "loss": 1.0792,
+      "step": 5439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2594486178953975,
+      "learning_rate": 0.00017678829379630406,
+      "loss": 1.0737,
+      "step": 5440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2743596761285101,
+      "learning_rate": 0.0001767781590249446,
+      "loss": 1.0216,
+      "step": 5441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3014933928943263,
+      "learning_rate": 0.00017676802233216346,
+      "loss": 1.0686,
+      "step": 5442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30975481238566765,
+      "learning_rate": 0.00017675788371821432,
+      "loss": 1.1826,
+      "step": 5443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24533869531288546,
+      "learning_rate": 0.00017674774318335085,
+      "loss": 1.162,
+      "step": 5444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27791112618813957,
+      "learning_rate": 0.0001767376007278269,
+      "loss": 0.9609,
+      "step": 5445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.31278080093016536,
+      "learning_rate": 0.00017672745635189633,
+      "loss": 1.1661,
+      "step": 5446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29874080218471216,
+      "learning_rate": 0.00017671731005581287,
+      "loss": 1.1068,
+      "step": 5447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2569503493201101,
+      "learning_rate": 0.0001767071618398305,
+      "loss": 1.0538,
+      "step": 5448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29826323414906175,
+      "learning_rate": 0.00017669701170420322,
+      "loss": 1.1264,
+      "step": 5449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2812826097473197,
+      "learning_rate": 0.00017668685964918504,
+      "loss": 1.0982,
+      "step": 5450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2973251253473092,
+      "learning_rate": 0.00017667670567502998,
+      "loss": 1.0728,
+      "step": 5451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24818251204984365,
+      "learning_rate": 0.0001766665497819922,
+      "loss": 1.0567,
+      "step": 5452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2884278666490467,
+      "learning_rate": 0.00017665639197032582,
+      "loss": 1.0685,
+      "step": 5453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.262999153831761,
+      "learning_rate": 0.00017664623224028503,
+      "loss": 0.9473,
+      "step": 5454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3152269030127414,
+      "learning_rate": 0.0001766360705921241,
+      "loss": 1.169,
+      "step": 5455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25251227559062966,
+      "learning_rate": 0.00017662590702609737,
+      "loss": 1.0352,
+      "step": 5456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.32106633074490737,
+      "learning_rate": 0.00017661574154245914,
+      "loss": 1.0894,
+      "step": 5457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2693010524825926,
+      "learning_rate": 0.00017660557414146384,
+      "loss": 1.0703,
+      "step": 5458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33501583797096984,
+      "learning_rate": 0.0001765954048233659,
+      "loss": 1.2,
+      "step": 5459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2766667108631253,
+      "learning_rate": 0.0001765852335884198,
+      "loss": 1.1096,
+      "step": 5460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.28642621560137427,
+      "learning_rate": 0.0001765750604368801,
+      "loss": 1.0589,
+      "step": 5461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29446092888521896,
+      "learning_rate": 0.0001765648853690014,
+      "loss": 1.0758,
+      "step": 5462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26574617011082774,
+      "learning_rate": 0.00017655470838503834,
+      "loss": 1.0517,
+      "step": 5463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29380273713342564,
+      "learning_rate": 0.00017654452948524555,
+      "loss": 1.1355,
+      "step": 5464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26610600805445106,
+      "learning_rate": 0.00017653434866987783,
+      "loss": 1.156,
+      "step": 5465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.23226425198871892,
+      "learning_rate": 0.00017652416593918994,
+      "loss": 1.1446,
+      "step": 5466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3136944620672886,
+      "learning_rate": 0.00017651398129343667,
+      "loss": 1.1359,
+      "step": 5467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29970255604775625,
+      "learning_rate": 0.00017650379473287296,
+      "loss": 1.0718,
+      "step": 5468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2782035372010567,
+      "learning_rate": 0.0001764936062577537,
+      "loss": 1.0908,
+      "step": 5469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2616477257641661,
+      "learning_rate": 0.00017648341586833387,
+      "loss": 0.9107,
+      "step": 5470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2663523876553537,
+      "learning_rate": 0.00017647322356486848,
+      "loss": 0.9981,
+      "step": 5471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26092658264988006,
+      "learning_rate": 0.0001764630293476126,
+      "loss": 1.1115,
+      "step": 5472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27133782980711774,
+      "learning_rate": 0.0001764528332168214,
+      "loss": 1.1506,
+      "step": 5473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.29494184409148894,
+      "learning_rate": 0.00017644263517274997,
+      "loss": 1.0381,
+      "step": 5474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.293344464237028,
+      "learning_rate": 0.00017643243521565355,
+      "loss": 1.0955,
+      "step": 5475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3113070820581294,
+      "learning_rate": 0.0001764222333457874,
+      "loss": 0.9733,
+      "step": 5476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2686929294263848,
+      "learning_rate": 0.00017641202956340685,
+      "loss": 1.141,
+      "step": 5477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2963778388036239,
+      "learning_rate": 0.0001764018238687672,
+      "loss": 1.049,
+      "step": 5478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27712005782523735,
+      "learning_rate": 0.00017639161626212393,
+      "loss": 0.9666,
+      "step": 5479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.25074497020772385,
+      "learning_rate": 0.00017638140674373245,
+      "loss": 1.091,
+      "step": 5480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27481911625527383,
+      "learning_rate": 0.00017637119531384822,
+      "loss": 0.9804,
+      "step": 5481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.30326990661840897,
+      "learning_rate": 0.00017636098197272687,
+      "loss": 1.0196,
+      "step": 5482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2536248159519646,
+      "learning_rate": 0.00017635076672062395,
+      "loss": 1.0655,
+      "step": 5483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.24871071887818122,
+      "learning_rate": 0.0001763405495577951,
+      "loss": 1.0337,
+      "step": 5484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.27076559378833587,
+      "learning_rate": 0.00017633033048449607,
+      "loss": 1.0868,
+      "step": 5485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.26447743099546034,
+      "learning_rate": 0.00017632010950098247,
+      "loss": 1.1067,
+      "step": 5486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.2354573802397408,
+      "learning_rate": 0.00017630988660751018,
+      "loss": 0.9972,
+      "step": 5487
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31657477736419243,
+      "learning_rate": 0.00017629966180433503,
+      "loss": 1.1436,
+      "step": 5488
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31018143648114505,
+      "learning_rate": 0.0001762894350917129,
+      "loss": 1.1186,
+      "step": 5489
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2928001553899392,
+      "learning_rate": 0.00017627920646989971,
+      "loss": 1.1422,
+      "step": 5490
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.30032975451874694,
+      "learning_rate": 0.00017626897593915142,
+      "loss": 1.0623,
+      "step": 5491
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3019231833502392,
+      "learning_rate": 0.0001762587434997241,
+      "loss": 1.0632,
+      "step": 5492
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2647986946052283,
+      "learning_rate": 0.0001762485091518738,
+      "loss": 1.0093,
+      "step": 5493
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2490169986270707,
+      "learning_rate": 0.0001762382728958566,
+      "loss": 1.1866,
+      "step": 5494
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27623003094440113,
+      "learning_rate": 0.00017622803473192874,
+      "loss": 1.0377,
+      "step": 5495
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29379957830634423,
+      "learning_rate": 0.0001762177946603464,
+      "loss": 1.1129,
+      "step": 5496
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26006713944566145,
+      "learning_rate": 0.00017620755268136584,
+      "loss": 1.0786,
+      "step": 5497
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2984517115544308,
+      "learning_rate": 0.00017619730879524337,
+      "loss": 1.145,
+      "step": 5498
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2634404683614875,
+      "learning_rate": 0.00017618706300223536,
+      "loss": 1.0225,
+      "step": 5499
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2770749610740461,
+      "learning_rate": 0.00017617681530259822,
+      "loss": 1.0321,
+      "step": 5500
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29909952213865876,
+      "learning_rate": 0.00017616656569658843,
+      "loss": 0.9445,
+      "step": 5501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28919886120746485,
+      "learning_rate": 0.00017615631418446242,
+      "loss": 1.0648,
+      "step": 5502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28680714124802115,
+      "learning_rate": 0.00017614606076647683,
+      "loss": 1.1729,
+      "step": 5503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3018876386617487,
+      "learning_rate": 0.00017613580544288817,
+      "loss": 0.9817,
+      "step": 5504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28707084465144506,
+      "learning_rate": 0.00017612554821395314,
+      "loss": 1.1636,
+      "step": 5505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31719224150490105,
+      "learning_rate": 0.00017611528907992844,
+      "loss": 1.0808,
+      "step": 5506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2625161917732417,
+      "learning_rate": 0.00017610502804107082,
+      "loss": 1.1023,
+      "step": 5507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2872356902170121,
+      "learning_rate": 0.00017609476509763698,
+      "loss": 1.2038,
+      "step": 5508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25936712334845474,
+      "learning_rate": 0.00017608450024988382,
+      "loss": 0.9567,
+      "step": 5509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2791045831669703,
+      "learning_rate": 0.0001760742334980683,
+      "loss": 1.1003,
+      "step": 5510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2647336003050076,
+      "learning_rate": 0.00017606396484244721,
+      "loss": 1.1102,
+      "step": 5511
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27574681080006713,
+      "learning_rate": 0.00017605369428327761,
+      "loss": 1.048,
+      "step": 5512
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27994961262509854,
+      "learning_rate": 0.00017604342182081653,
+      "loss": 0.9866,
+      "step": 5513
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.261516937266442,
+      "learning_rate": 0.000176033147455321,
+      "loss": 1.0791,
+      "step": 5514
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2995061891654347,
+      "learning_rate": 0.0001760228711870482,
+      "loss": 1.1309,
+      "step": 5515
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2820497610571214,
+      "learning_rate": 0.00017601259301625524,
+      "loss": 1.0,
+      "step": 5516
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24255160815288854,
+      "learning_rate": 0.0001760023129431994,
+      "loss": 1.0015,
+      "step": 5517
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25094239441876454,
+      "learning_rate": 0.0001759920309681379,
+      "loss": 1.0133,
+      "step": 5518
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.283208208913356,
+      "learning_rate": 0.00017598174709132803,
+      "loss": 1.1402,
+      "step": 5519
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2826249410012862,
+      "learning_rate": 0.00017597146131302722,
+      "loss": 1.0725,
+      "step": 5520
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28680753301102535,
+      "learning_rate": 0.00017596117363349282,
+      "loss": 1.1058,
+      "step": 5521
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2915298750178331,
+      "learning_rate": 0.00017595088405298234,
+      "loss": 1.0234,
+      "step": 5522
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29543680871161176,
+      "learning_rate": 0.00017594059257175325,
+      "loss": 1.1357,
+      "step": 5523
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2734660068763002,
+      "learning_rate": 0.0001759302991900631,
+      "loss": 0.9803,
+      "step": 5524
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29698197284064065,
+      "learning_rate": 0.0001759200039081695,
+      "loss": 1.0425,
+      "step": 5525
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28542474538774965,
+      "learning_rate": 0.00017590970672633007,
+      "loss": 1.0379,
+      "step": 5526
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2891844662811193,
+      "learning_rate": 0.00017589940764480252,
+      "loss": 1.0073,
+      "step": 5527
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28262400994939874,
+      "learning_rate": 0.00017588910666384462,
+      "loss": 1.0782,
+      "step": 5528
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2780628600213389,
+      "learning_rate": 0.00017587880378371412,
+      "loss": 1.0688,
+      "step": 5529
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2884519639543894,
+      "learning_rate": 0.00017586849900466883,
+      "loss": 1.0724,
+      "step": 5530
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2815574093742316,
+      "learning_rate": 0.00017585819232696675,
+      "loss": 1.0518,
+      "step": 5531
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28038754616537154,
+      "learning_rate": 0.00017584788375086565,
+      "loss": 1.0102,
+      "step": 5532
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2525746581044452,
+      "learning_rate": 0.00017583757327662363,
+      "loss": 1.0554,
+      "step": 5533
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2922109980484516,
+      "learning_rate": 0.00017582726090449867,
+      "loss": 1.1027,
+      "step": 5534
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2751484254089623,
+      "learning_rate": 0.00017581694663474886,
+      "loss": 1.0159,
+      "step": 5535
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26863647079539626,
+      "learning_rate": 0.00017580663046763231,
+      "loss": 1.0621,
+      "step": 5536
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2985191792690513,
+      "learning_rate": 0.00017579631240340716,
+      "loss": 1.1758,
+      "step": 5537
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27614555889342746,
+      "learning_rate": 0.00017578599244233168,
+      "loss": 1.061,
+      "step": 5538
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25905510867782044,
+      "learning_rate": 0.00017577567058466414,
+      "loss": 1.1777,
+      "step": 5539
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3278506380359264,
+      "learning_rate": 0.00017576534683066278,
+      "loss": 1.1552,
+      "step": 5540
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.273031516195577,
+      "learning_rate": 0.000175755021180586,
+      "loss": 1.1249,
+      "step": 5541
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29918764475128873,
+      "learning_rate": 0.00017574469363469222,
+      "loss": 1.0937,
+      "step": 5542
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2753312938615492,
+      "learning_rate": 0.00017573436419323986,
+      "loss": 1.1123,
+      "step": 5543
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3076985822417243,
+      "learning_rate": 0.00017572403285648743,
+      "loss": 1.1392,
+      "step": 5544
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2858977995760822,
+      "learning_rate": 0.00017571369962469352,
+      "loss": 1.0683,
+      "step": 5545
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2826635982664086,
+      "learning_rate": 0.00017570336449811667,
+      "loss": 1.01,
+      "step": 5546
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25331688616473097,
+      "learning_rate": 0.00017569302747701558,
+      "loss": 1.0202,
+      "step": 5547
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25299991837835445,
+      "learning_rate": 0.00017568268856164886,
+      "loss": 0.9968,
+      "step": 5548
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26907253785105184,
+      "learning_rate": 0.00017567234775227533,
+      "loss": 1.0216,
+      "step": 5549
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2638154646751891,
+      "learning_rate": 0.0001756620050491537,
+      "loss": 1.169,
+      "step": 5550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2674914565002557,
+      "learning_rate": 0.0001756516604525429,
+      "loss": 1.0112,
+      "step": 5551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27204031327263634,
+      "learning_rate": 0.00017564131396270168,
+      "loss": 1.0392,
+      "step": 5552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2808788802646591,
+      "learning_rate": 0.0001756309655798891,
+      "loss": 1.1142,
+      "step": 5553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2737824930333972,
+      "learning_rate": 0.00017562061530436405,
+      "loss": 1.1227,
+      "step": 5554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28525600932665507,
+      "learning_rate": 0.00017561026313638557,
+      "loss": 1.0276,
+      "step": 5555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26176420613090035,
+      "learning_rate": 0.00017559990907621274,
+      "loss": 1.1253,
+      "step": 5556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25918345855645814,
+      "learning_rate": 0.00017558955312410468,
+      "loss": 1.1035,
+      "step": 5557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29529919222022755,
+      "learning_rate": 0.00017557919528032054,
+      "loss": 1.1018,
+      "step": 5558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28939151113442757,
+      "learning_rate": 0.00017556883554511953,
+      "loss": 1.0248,
+      "step": 5559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3092935541500706,
+      "learning_rate": 0.00017555847391876093,
+      "loss": 1.133,
+      "step": 5560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.268156083959369,
+      "learning_rate": 0.00017554811040150403,
+      "loss": 1.1213,
+      "step": 5561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2621845882914305,
+      "learning_rate": 0.0001755377449936082,
+      "loss": 1.0878,
+      "step": 5562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2661667850236094,
+      "learning_rate": 0.0001755273776953328,
+      "loss": 1.0683,
+      "step": 5563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25877418531921775,
+      "learning_rate": 0.00017551700850693732,
+      "loss": 1.1574,
+      "step": 5564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28288655943020663,
+      "learning_rate": 0.00017550663742868126,
+      "loss": 1.1328,
+      "step": 5565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26185751105253613,
+      "learning_rate": 0.00017549626446082412,
+      "loss": 0.973,
+      "step": 5566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.295683213989208,
+      "learning_rate": 0.0001754858896036255,
+      "loss": 0.9697,
+      "step": 5567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3264307309216949,
+      "learning_rate": 0.0001754755128573451,
+      "loss": 1.0159,
+      "step": 5568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2753013510908348,
+      "learning_rate": 0.00017546513422224253,
+      "loss": 1.0423,
+      "step": 5569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25968975647863907,
+      "learning_rate": 0.00017545475369857755,
+      "loss": 1.1648,
+      "step": 5570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2804203410262089,
+      "learning_rate": 0.00017544437128660993,
+      "loss": 1.0268,
+      "step": 5571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.31239158563885117,
+      "learning_rate": 0.0001754339869865995,
+      "loss": 1.1495,
+      "step": 5572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2733500389923679,
+      "learning_rate": 0.00017542360079880615,
+      "loss": 1.1128,
+      "step": 5573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24320614149769154,
+      "learning_rate": 0.00017541321272348978,
+      "loss": 1.0674,
+      "step": 5574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3181218743470043,
+      "learning_rate": 0.00017540282276091039,
+      "loss": 1.0091,
+      "step": 5575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3031447060986254,
+      "learning_rate": 0.00017539243091132793,
+      "loss": 1.1715,
+      "step": 5576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34380612538960487,
+      "learning_rate": 0.00017538203717500252,
+      "loss": 0.9964,
+      "step": 5577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2661056584426639,
+      "learning_rate": 0.00017537164155219428,
+      "loss": 1.0928,
+      "step": 5578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.26522368104967226,
+      "learning_rate": 0.0001753612440431633,
+      "loss": 1.0348,
+      "step": 5579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.28303690003405435,
+      "learning_rate": 0.00017535084464816985,
+      "loss": 1.0077,
+      "step": 5580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.29888676081156723,
+      "learning_rate": 0.00017534044336747418,
+      "loss": 0.9651,
+      "step": 5581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.23610813701308117,
+      "learning_rate": 0.00017533004020133653,
+      "loss": 1.0869,
+      "step": 5582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2473777150573363,
+      "learning_rate": 0.00017531963515001725,
+      "loss": 1.0581,
+      "step": 5583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2878283161585639,
+      "learning_rate": 0.00017530922821377683,
+      "loss": 1.1182,
+      "step": 5584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2895723267837339,
+      "learning_rate": 0.0001752988193928756,
+      "loss": 1.0193,
+      "step": 5585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.25924626472676504,
+      "learning_rate": 0.00017528840868757413,
+      "loss": 1.0764,
+      "step": 5586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2697046584787591,
+      "learning_rate": 0.00017527799609813287,
+      "loss": 1.2123,
+      "step": 5587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2779570564976789,
+      "learning_rate": 0.00017526758162481247,
+      "loss": 1.0463,
+      "step": 5588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.2659291393898864,
+      "learning_rate": 0.00017525716526787353,
+      "loss": 0.8657,
+      "step": 5589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.24940778956498255,
+      "learning_rate": 0.00017524674702757676,
+      "loss": 0.9349,
+      "step": 5590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3512612443921498,
+      "learning_rate": 0.00017523632690418281,
+      "loss": 1.0309,
+      "step": 5591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.27949443291870935,
+      "learning_rate": 0.0001752259048979525,
+      "loss": 1.0867,
+      "step": 5592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25241117168172916,
+      "learning_rate": 0.00017521548100914668,
+      "loss": 0.9595,
+      "step": 5593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31177379926003695,
+      "learning_rate": 0.00017520505523802615,
+      "loss": 1.1147,
+      "step": 5594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3035376452103113,
+      "learning_rate": 0.00017519462758485186,
+      "loss": 1.1091,
+      "step": 5595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28334338339825155,
+      "learning_rate": 0.00017518419804988473,
+      "loss": 1.0027,
+      "step": 5596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27635425273795633,
+      "learning_rate": 0.00017517376663338583,
+      "loss": 1.0832,
+      "step": 5597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27479092014996065,
+      "learning_rate": 0.00017516333333561615,
+      "loss": 1.0254,
+      "step": 5598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2923804540268794,
+      "learning_rate": 0.00017515289815683683,
+      "loss": 1.0939,
+      "step": 5599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2689789333872261,
+      "learning_rate": 0.000175142461097309,
+      "loss": 1.0924,
+      "step": 5600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2712229566662846,
+      "learning_rate": 0.00017513202215729384,
+      "loss": 1.1212,
+      "step": 5601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28499521398979466,
+      "learning_rate": 0.0001751215813370526,
+      "loss": 1.1271,
+      "step": 5602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.272591489298431,
+      "learning_rate": 0.00017511113863684662,
+      "loss": 1.1602,
+      "step": 5603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3008741005719719,
+      "learning_rate": 0.00017510069405693714,
+      "loss": 1.115,
+      "step": 5604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2531166951692093,
+      "learning_rate": 0.00017509024759758561,
+      "loss": 1.0802,
+      "step": 5605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24148412845313758,
+      "learning_rate": 0.00017507979925905347,
+      "loss": 0.953,
+      "step": 5606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2534357412942829,
+      "learning_rate": 0.00017506934904160213,
+      "loss": 1.0784,
+      "step": 5607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.270455450479634,
+      "learning_rate": 0.00017505889694549316,
+      "loss": 1.0495,
+      "step": 5608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2718743775512695,
+      "learning_rate": 0.00017504844297098812,
+      "loss": 1.1477,
+      "step": 5609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2627508977482193,
+      "learning_rate": 0.00017503798711834863,
+      "loss": 0.9724,
+      "step": 5610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28688718615510195,
+      "learning_rate": 0.00017502752938783637,
+      "loss": 1.0391,
+      "step": 5611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3088622269070572,
+      "learning_rate": 0.000175017069779713,
+      "loss": 1.07,
+      "step": 5612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2519490699405952,
+      "learning_rate": 0.00017500660829424035,
+      "loss": 0.9973,
+      "step": 5613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2593258598031295,
+      "learning_rate": 0.00017499614493168017,
+      "loss": 1.1488,
+      "step": 5614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28609091063527664,
+      "learning_rate": 0.00017498567969229432,
+      "loss": 1.1316,
+      "step": 5615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2876241038098421,
+      "learning_rate": 0.00017497521257634472,
+      "loss": 1.142,
+      "step": 5616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2699264999096545,
+      "learning_rate": 0.0001749647435840933,
+      "loss": 1.1133,
+      "step": 5617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23212212244543703,
+      "learning_rate": 0.00017495427271580207,
+      "loss": 1.0177,
+      "step": 5618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3191035435173965,
+      "learning_rate": 0.00017494379997173306,
+      "loss": 1.0753,
+      "step": 5619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28382503691739164,
+      "learning_rate": 0.00017493332535214835,
+      "loss": 1.0277,
+      "step": 5620
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2620982386743405,
+      "learning_rate": 0.00017492284885731006,
+      "loss": 1.1087,
+      "step": 5621
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24048768135309437,
+      "learning_rate": 0.00017491237048748042,
+      "loss": 1.0179,
+      "step": 5622
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29297621702120474,
+      "learning_rate": 0.00017490189024292157,
+      "loss": 1.031,
+      "step": 5623
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24169868093525804,
+      "learning_rate": 0.00017489140812389591,
+      "loss": 1.1275,
+      "step": 5624
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3265517682655214,
+      "learning_rate": 0.00017488092413066566,
+      "loss": 1.09,
+      "step": 5625
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2889070937315453,
+      "learning_rate": 0.00017487043826349324,
+      "loss": 1.0827,
+      "step": 5626
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2616726142921512,
+      "learning_rate": 0.00017485995052264107,
+      "loss": 1.1226,
+      "step": 5627
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2823144964930083,
+      "learning_rate": 0.00017484946090837153,
+      "loss": 0.9189,
+      "step": 5628
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26247896162782614,
+      "learning_rate": 0.0001748389694209472,
+      "loss": 1.0863,
+      "step": 5629
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29529068834973227,
+      "learning_rate": 0.0001748284760606307,
+      "loss": 1.1572,
+      "step": 5630
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2725314072487987,
+      "learning_rate": 0.00017481798082768447,
+      "loss": 1.0527,
+      "step": 5631
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3268266589015598,
+      "learning_rate": 0.0001748074837223713,
+      "loss": 1.1146,
+      "step": 5632
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.31412407097134504,
+      "learning_rate": 0.0001747969847449538,
+      "loss": 1.0571,
+      "step": 5633
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3129044080305532,
+      "learning_rate": 0.0001747864838956948,
+      "loss": 1.1354,
+      "step": 5634
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26484366612789245,
+      "learning_rate": 0.00017477598117485697,
+      "loss": 1.0219,
+      "step": 5635
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2881609277060793,
+      "learning_rate": 0.00017476547658270327,
+      "loss": 1.0661,
+      "step": 5636
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2785791649184835,
+      "learning_rate": 0.0001747549701194965,
+      "loss": 1.009,
+      "step": 5637
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2781943063291394,
+      "learning_rate": 0.00017474446178549963,
+      "loss": 1.1618,
+      "step": 5638
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.286700570393826,
+      "learning_rate": 0.00017473395158097566,
+      "loss": 1.1184,
+      "step": 5639
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.332805537003659,
+      "learning_rate": 0.00017472343950618755,
+      "loss": 1.0486,
+      "step": 5640
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27971090508186786,
+      "learning_rate": 0.0001747129255613984,
+      "loss": 1.1044,
+      "step": 5641
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28836371050426973,
+      "learning_rate": 0.00017470240974687133,
+      "loss": 1.0402,
+      "step": 5642
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30716094757551443,
+      "learning_rate": 0.00017469189206286952,
+      "loss": 1.113,
+      "step": 5643
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2743312373131173,
+      "learning_rate": 0.00017468137250965617,
+      "loss": 1.0299,
+      "step": 5644
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2609975564796487,
+      "learning_rate": 0.00017467085108749454,
+      "loss": 1.104,
+      "step": 5645
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29364822283678577,
+      "learning_rate": 0.0001746603277966479,
+      "loss": 1.0865,
+      "step": 5646
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2745578477787385,
+      "learning_rate": 0.00017464980263737968,
+      "loss": 1.1527,
+      "step": 5647
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26626720698396383,
+      "learning_rate": 0.00017463927560995321,
+      "loss": 1.0189,
+      "step": 5648
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27709851029020705,
+      "learning_rate": 0.000174628746714632,
+      "loss": 1.1241,
+      "step": 5649
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.306265199776124,
+      "learning_rate": 0.00017461821595167945,
+      "loss": 1.1816,
+      "step": 5650
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23094636122975018,
+      "learning_rate": 0.00017460768332135918,
+      "loss": 1.0002,
+      "step": 5651
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25739454457437366,
+      "learning_rate": 0.00017459714882393473,
+      "loss": 1.0748,
+      "step": 5652
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.24839067246041194,
+      "learning_rate": 0.00017458661245966974,
+      "loss": 1.1297,
+      "step": 5653
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2592947317889122,
+      "learning_rate": 0.0001745760742288279,
+      "loss": 1.1097,
+      "step": 5654
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29910161986761424,
+      "learning_rate": 0.00017456553413167293,
+      "loss": 1.0156,
+      "step": 5655
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2740550204065586,
+      "learning_rate": 0.00017455499216846864,
+      "loss": 1.1618,
+      "step": 5656
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2773348179265606,
+      "learning_rate": 0.00017454444833947877,
+      "loss": 1.0988,
+      "step": 5657
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2813065691260808,
+      "learning_rate": 0.00017453390264496728,
+      "loss": 0.9826,
+      "step": 5658
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2775051391562762,
+      "learning_rate": 0.000174523355085198,
+      "loss": 1.0356,
+      "step": 5659
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28015989674393643,
+      "learning_rate": 0.00017451280566043492,
+      "loss": 1.1507,
+      "step": 5660
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3011891640480555,
+      "learning_rate": 0.00017450225437094208,
+      "loss": 1.16,
+      "step": 5661
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.38100561539213623,
+      "learning_rate": 0.00017449170121698347,
+      "loss": 1.0236,
+      "step": 5662
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27806057374613613,
+      "learning_rate": 0.00017448114619882321,
+      "loss": 1.0843,
+      "step": 5663
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2851575165624332,
+      "learning_rate": 0.0001744705893167255,
+      "loss": 1.1458,
+      "step": 5664
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2913077381839934,
+      "learning_rate": 0.00017446003057095447,
+      "loss": 1.1039,
+      "step": 5665
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28664285006948,
+      "learning_rate": 0.00017444946996177433,
+      "loss": 1.1214,
+      "step": 5666
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2838055565232777,
+      "learning_rate": 0.00017443890748944946,
+      "loss": 1.0184,
+      "step": 5667
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2511465662279637,
+      "learning_rate": 0.00017442834315424416,
+      "loss": 1.0698,
+      "step": 5668
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27586967822568237,
+      "learning_rate": 0.0001744177769564228,
+      "loss": 1.0234,
+      "step": 5669
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23851868493807254,
+      "learning_rate": 0.00017440720889624978,
+      "loss": 1.109,
+      "step": 5670
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29280665470416317,
+      "learning_rate": 0.00017439663897398958,
+      "loss": 1.1473,
+      "step": 5671
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27293284903881204,
+      "learning_rate": 0.00017438606718990675,
+      "loss": 1.0855,
+      "step": 5672
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3102331601705705,
+      "learning_rate": 0.00017437549354426586,
+      "loss": 0.979,
+      "step": 5673
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2904424568361712,
+      "learning_rate": 0.00017436491803733147,
+      "loss": 1.0758,
+      "step": 5674
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3038807854764857,
+      "learning_rate": 0.00017435434066936828,
+      "loss": 1.0813,
+      "step": 5675
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2982170129842694,
+      "learning_rate": 0.00017434376144064096,
+      "loss": 0.9269,
+      "step": 5676
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.29066232686572985,
+      "learning_rate": 0.00017433318035141432,
+      "loss": 1.0761,
+      "step": 5677
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36384380111246045,
+      "learning_rate": 0.0001743225974019531,
+      "loss": 1.1277,
+      "step": 5678
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.28539850304066966,
+      "learning_rate": 0.00017431201259252222,
+      "loss": 1.1838,
+      "step": 5679
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3010319695376003,
+      "learning_rate": 0.00017430142592338648,
+      "loss": 0.9309,
+      "step": 5680
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2639994319325743,
+      "learning_rate": 0.00017429083739481087,
+      "loss": 1.1049,
+      "step": 5681
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2979159287960645,
+      "learning_rate": 0.00017428024700706036,
+      "loss": 1.1221,
+      "step": 5682
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2819099394302753,
+      "learning_rate": 0.0001742696547604,
+      "loss": 1.0916,
+      "step": 5683
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.260104363656028,
+      "learning_rate": 0.00017425906065509484,
+      "loss": 1.0034,
+      "step": 5684
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2970820007503617,
+      "learning_rate": 0.00017424846469141,
+      "loss": 1.1024,
+      "step": 5685
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2620883115817067,
+      "learning_rate": 0.0001742378668696107,
+      "loss": 1.0562,
+      "step": 5686
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2491078043946645,
+      "learning_rate": 0.0001742272671899621,
+      "loss": 1.0625,
+      "step": 5687
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30629114639440114,
+      "learning_rate": 0.00017421666565272948,
+      "loss": 1.202,
+      "step": 5688
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.30830018039821566,
+      "learning_rate": 0.0001742060622581782,
+      "loss": 1.0535,
+      "step": 5689
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2724222943145667,
+      "learning_rate": 0.00017419545700657354,
+      "loss": 1.1325,
+      "step": 5690
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.23708317781034122,
+      "learning_rate": 0.00017418484989818096,
+      "loss": 1.0638,
+      "step": 5691
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.26229220094670336,
+      "learning_rate": 0.00017417424093326588,
+      "loss": 0.9439,
+      "step": 5692
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.25017286913990483,
+      "learning_rate": 0.0001741636301120938,
+      "loss": 1.079,
+      "step": 5693
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2860341282853943,
+      "learning_rate": 0.00017415301743493026,
+      "loss": 1.109,
+      "step": 5694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.27000375628865514,
+      "learning_rate": 0.00017414240290204087,
+      "loss": 1.0366,
+      "step": 5695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.2944772274521784,
+      "learning_rate": 0.00017413178651369123,
+      "loss": 1.106,
+      "step": 5696
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29923112387635076,
+      "learning_rate": 0.00017412116827014707,
+      "loss": 1.118,
+      "step": 5697
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27744625628639163,
+      "learning_rate": 0.00017411054817167407,
+      "loss": 1.0582,
+      "step": 5698
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2872887689318743,
+      "learning_rate": 0.00017409992621853803,
+      "loss": 1.135,
+      "step": 5699
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2868590588865866,
+      "learning_rate": 0.00017408930241100476,
+      "loss": 1.1218,
+      "step": 5700
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28039756959308726,
+      "learning_rate": 0.00017407867674934014,
+      "loss": 1.0089,
+      "step": 5701
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26834033841073884,
+      "learning_rate": 0.00017406804923381008,
+      "loss": 1.0931,
+      "step": 5702
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2679055295777244,
+      "learning_rate": 0.00017405741986468054,
+      "loss": 1.0362,
+      "step": 5703
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31948114555469886,
+      "learning_rate": 0.00017404678864221752,
+      "loss": 1.1499,
+      "step": 5704
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.254735913314033,
+      "learning_rate": 0.00017403615556668708,
+      "loss": 1.0521,
+      "step": 5705
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2758568173458296,
+      "learning_rate": 0.00017402552063835533,
+      "loss": 1.0276,
+      "step": 5706
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2881056253518769,
+      "learning_rate": 0.0001740148838574884,
+      "loss": 0.9892,
+      "step": 5707
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28754810292102423,
+      "learning_rate": 0.00017400424522435247,
+      "loss": 1.0138,
+      "step": 5708
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26622897762881104,
+      "learning_rate": 0.0001739936047392138,
+      "loss": 1.0121,
+      "step": 5709
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3257799896467255,
+      "learning_rate": 0.00017398296240233866,
+      "loss": 1.0488,
+      "step": 5710
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.300245783416075,
+      "learning_rate": 0.0001739723182139934,
+      "loss": 1.199,
+      "step": 5711
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29793504189936215,
+      "learning_rate": 0.00017396167217444437,
+      "loss": 1.0326,
+      "step": 5712
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2895788092153529,
+      "learning_rate": 0.00017395102428395803,
+      "loss": 1.0282,
+      "step": 5713
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2759129672263007,
+      "learning_rate": 0.0001739403745428008,
+      "loss": 1.1353,
+      "step": 5714
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2736823382715656,
+      "learning_rate": 0.0001739297229512393,
+      "loss": 1.0918,
+      "step": 5715
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30836802392427065,
+      "learning_rate": 0.00017391906950953994,
+      "loss": 1.1101,
+      "step": 5716
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2923956322766849,
+      "learning_rate": 0.00017390841421796943,
+      "loss": 1.1005,
+      "step": 5717
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2582193114884755,
+      "learning_rate": 0.00017389775707679444,
+      "loss": 1.0708,
+      "step": 5718
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24558151292411884,
+      "learning_rate": 0.0001738870980862816,
+      "loss": 1.1303,
+      "step": 5719
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29322536081033934,
+      "learning_rate": 0.0001738764372466977,
+      "loss": 1.0927,
+      "step": 5720
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2659755712941869,
+      "learning_rate": 0.00017386577455830952,
+      "loss": 1.0032,
+      "step": 5721
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2635454724857395,
+      "learning_rate": 0.00017385511002138393,
+      "loss": 1.2808,
+      "step": 5722
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26786955193926654,
+      "learning_rate": 0.0001738444436361878,
+      "loss": 1.0475,
+      "step": 5723
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28425119186729847,
+      "learning_rate": 0.00017383377540298805,
+      "loss": 1.0817,
+      "step": 5724
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28386565821296494,
+      "learning_rate": 0.00017382310532205165,
+      "loss": 0.9009,
+      "step": 5725
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2455076247487801,
+      "learning_rate": 0.00017381243339364565,
+      "loss": 0.9972,
+      "step": 5726
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2949244070026697,
+      "learning_rate": 0.00017380175961803713,
+      "loss": 1.0042,
+      "step": 5727
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2806527868238312,
+      "learning_rate": 0.00017379108399549317,
+      "loss": 1.1932,
+      "step": 5728
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27785708544616866,
+      "learning_rate": 0.000173780406526281,
+      "loss": 1.0914,
+      "step": 5729
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28683282908042806,
+      "learning_rate": 0.00017376972721066776,
+      "loss": 1.1145,
+      "step": 5730
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2838770112878104,
+      "learning_rate": 0.00017375904604892073,
+      "loss": 1.0765,
+      "step": 5731
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2934314888524707,
+      "learning_rate": 0.0001737483630413072,
+      "loss": 1.0595,
+      "step": 5732
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25830945410699585,
+      "learning_rate": 0.00017373767818809456,
+      "loss": 0.9669,
+      "step": 5733
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2665953859888515,
+      "learning_rate": 0.00017372699148955018,
+      "loss": 1.0714,
+      "step": 5734
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2731137251463795,
+      "learning_rate": 0.0001737163029459415,
+      "loss": 1.0746,
+      "step": 5735
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25825981988483343,
+      "learning_rate": 0.00017370561255753602,
+      "loss": 0.9534,
+      "step": 5736
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30383684481721607,
+      "learning_rate": 0.00017369492032460123,
+      "loss": 1.1384,
+      "step": 5737
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31417342120569464,
+      "learning_rate": 0.00017368422624740478,
+      "loss": 1.0576,
+      "step": 5738
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2593381194717112,
+      "learning_rate": 0.00017367353032621426,
+      "loss": 1.0573,
+      "step": 5739
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2778893532357316,
+      "learning_rate": 0.00017366283256129732,
+      "loss": 0.9766,
+      "step": 5740
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3072153615495613,
+      "learning_rate": 0.0001736521329529217,
+      "loss": 1.1247,
+      "step": 5741
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28745857193672664,
+      "learning_rate": 0.00017364143150135517,
+      "loss": 1.1141,
+      "step": 5742
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2651127640673508,
+      "learning_rate": 0.00017363072820686552,
+      "loss": 1.0829,
+      "step": 5743
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2902298639371502,
+      "learning_rate": 0.00017362002306972065,
+      "loss": 0.987,
+      "step": 5744
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27655920423756686,
+      "learning_rate": 0.00017360931609018842,
+      "loss": 1.0076,
+      "step": 5745
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2567390152989073,
+      "learning_rate": 0.0001735986072685368,
+      "loss": 1.0375,
+      "step": 5746
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24515803399806585,
+      "learning_rate": 0.00017358789660503377,
+      "loss": 1.0745,
+      "step": 5747
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2575517841295073,
+      "learning_rate": 0.0001735771840999474,
+      "loss": 1.0153,
+      "step": 5748
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2764368481508909,
+      "learning_rate": 0.0001735664697535457,
+      "loss": 1.1584,
+      "step": 5749
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27351695694822586,
+      "learning_rate": 0.0001735557535660969,
+      "loss": 1.0281,
+      "step": 5750
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2667030418236546,
+      "learning_rate": 0.00017354503553786916,
+      "loss": 1.1321,
+      "step": 5751
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29279846320754066,
+      "learning_rate": 0.00017353431566913066,
+      "loss": 0.9984,
+      "step": 5752
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2415120154785472,
+      "learning_rate": 0.0001735235939601497,
+      "loss": 1.0611,
+      "step": 5753
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25955357361068576,
+      "learning_rate": 0.00017351287041119458,
+      "loss": 1.1294,
+      "step": 5754
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.294206732380647,
+      "learning_rate": 0.0001735021450225337,
+      "loss": 1.0812,
+      "step": 5755
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29031247550246536,
+      "learning_rate": 0.00017349141779443542,
+      "loss": 1.0576,
+      "step": 5756
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2672750801707316,
+      "learning_rate": 0.00017348068872716823,
+      "loss": 1.0851,
+      "step": 5757
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26957335518481657,
+      "learning_rate": 0.00017346995782100062,
+      "loss": 1.0858,
+      "step": 5758
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.32734612669273955,
+      "learning_rate": 0.00017345922507620116,
+      "loss": 1.0656,
+      "step": 5759
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24154451886063427,
+      "learning_rate": 0.00017344849049303842,
+      "loss": 0.9896,
+      "step": 5760
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27709567176869415,
+      "learning_rate": 0.00017343775407178104,
+      "loss": 1.0805,
+      "step": 5761
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.272244078427738,
+      "learning_rate": 0.0001734270158126977,
+      "loss": 1.0433,
+      "step": 5762
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2673872164348252,
+      "learning_rate": 0.00017341627571605716,
+      "loss": 1.1464,
+      "step": 5763
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27315997652853985,
+      "learning_rate": 0.00017340553378212816,
+      "loss": 1.1567,
+      "step": 5764
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.23474379585480734,
+      "learning_rate": 0.00017339479001117955,
+      "loss": 1.0089,
+      "step": 5765
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27652850547755903,
+      "learning_rate": 0.00017338404440348022,
+      "loss": 1.0461,
+      "step": 5766
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.31896644764441623,
+      "learning_rate": 0.00017337329695929902,
+      "loss": 1.0359,
+      "step": 5767
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.274574680140215,
+      "learning_rate": 0.00017336254767890498,
+      "loss": 1.0201,
+      "step": 5768
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2630831469195627,
+      "learning_rate": 0.00017335179656256705,
+      "loss": 1.0975,
+      "step": 5769
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2793728024955909,
+      "learning_rate": 0.00017334104361055436,
+      "loss": 1.1332,
+      "step": 5770
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27408839587852935,
+      "learning_rate": 0.0001733302888231359,
+      "loss": 0.939,
+      "step": 5771
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27297436472505615,
+      "learning_rate": 0.0001733195322005809,
+      "loss": 1.0168,
+      "step": 5772
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.30630256010686685,
+      "learning_rate": 0.00017330877374315855,
+      "loss": 1.1169,
+      "step": 5773
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2767879583144832,
+      "learning_rate": 0.00017329801345113802,
+      "loss": 1.0233,
+      "step": 5774
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29684472098320214,
+      "learning_rate": 0.0001732872513247887,
+      "loss": 1.0463,
+      "step": 5775
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2785233560114013,
+      "learning_rate": 0.00017327648736437977,
+      "loss": 1.142,
+      "step": 5776
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3208276011124166,
+      "learning_rate": 0.00017326572157018078,
+      "loss": 1.1932,
+      "step": 5777
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22684754778328245,
+      "learning_rate": 0.000173254953942461,
+      "loss": 1.0656,
+      "step": 5778
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3374242730404179,
+      "learning_rate": 0.00017324418448148998,
+      "loss": 1.049,
+      "step": 5779
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3428578883405557,
+      "learning_rate": 0.0001732334131875372,
+      "loss": 1.1952,
+      "step": 5780
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27472370192399703,
+      "learning_rate": 0.00017322264006087225,
+      "loss": 1.2441,
+      "step": 5781
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27491429642165643,
+      "learning_rate": 0.0001732118651017647,
+      "loss": 1.0994,
+      "step": 5782
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27459654102558345,
+      "learning_rate": 0.00017320108831048422,
+      "loss": 1.0496,
+      "step": 5783
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2699347475330186,
+      "learning_rate": 0.0001731903096873005,
+      "loss": 1.1214,
+      "step": 5784
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.296175698180158,
+      "learning_rate": 0.00017317952923248328,
+      "loss": 1.0308,
+      "step": 5785
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.29058979499652565,
+      "learning_rate": 0.00017316874694630236,
+      "loss": 1.0828,
+      "step": 5786
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2997585209236349,
+      "learning_rate": 0.00017315796282902753,
+      "loss": 1.1366,
+      "step": 5787
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27673867013826936,
+      "learning_rate": 0.00017314717688092873,
+      "loss": 0.945,
+      "step": 5788
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.23598291014298442,
+      "learning_rate": 0.00017313638910227585,
+      "loss": 0.9402,
+      "step": 5789
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26553154287144975,
+      "learning_rate": 0.00017312559949333886,
+      "loss": 1.1785,
+      "step": 5790
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.28452803788674824,
+      "learning_rate": 0.0001731148080543878,
+      "loss": 1.124,
+      "step": 5791
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.2705626232253567,
+      "learning_rate": 0.00017310401478569273,
+      "loss": 1.0191,
+      "step": 5792
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24817380433889646,
+      "learning_rate": 0.0001730932196875237,
+      "loss": 1.0482,
+      "step": 5793
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26652308262279306,
+      "learning_rate": 0.00017308242276015094,
+      "loss": 1.0738,
+      "step": 5794
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24818290945260654,
+      "learning_rate": 0.00017307162400384462,
+      "loss": 1.0175,
+      "step": 5795
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.26331611360600954,
+      "learning_rate": 0.000173060823418875,
+      "loss": 1.0788,
+      "step": 5796
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.25070740236687494,
+      "learning_rate": 0.00017305002100551233,
+      "loss": 0.9522,
+      "step": 5797
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.27552761246559027,
+      "learning_rate": 0.000173039216764027,
+      "loss": 1.1854,
+      "step": 5798
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24607203901766253,
+      "learning_rate": 0.00017302841069468934,
+      "loss": 1.0574,
+      "step": 5799
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.24789972386779166,
+      "learning_rate": 0.00017301760279776982,
+      "loss": 1.0585,
+      "step": 5800
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.22933484912929925,
+      "learning_rate": 0.00017300679307353888,
+      "loss": 1.0806,
+      "step": 5801
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27649671163848355,
+      "learning_rate": 0.0001729959815222671,
+      "loss": 1.1766,
+      "step": 5802
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24226545584694414,
+      "learning_rate": 0.00017298516814422498,
+      "loss": 0.9117,
+      "step": 5803
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26242930789541363,
+      "learning_rate": 0.00017297435293968315,
+      "loss": 1.1165,
+      "step": 5804
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3120420916073829,
+      "learning_rate": 0.0001729635359089123,
+      "loss": 1.0574,
+      "step": 5805
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2701256919492001,
+      "learning_rate": 0.00017295271705218307,
+      "loss": 1.1047,
+      "step": 5806
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29645770053433096,
+      "learning_rate": 0.0001729418963697663,
+      "loss": 1.1329,
+      "step": 5807
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24632237776935018,
+      "learning_rate": 0.0001729310738619327,
+      "loss": 1.1359,
+      "step": 5808
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27163467992709467,
+      "learning_rate": 0.00017292024952895313,
+      "loss": 1.1006,
+      "step": 5809
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26959200346069406,
+      "learning_rate": 0.0001729094233710985,
+      "loss": 1.0042,
+      "step": 5810
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.289959983695105,
+      "learning_rate": 0.00017289859538863973,
+      "loss": 1.1085,
+      "step": 5811
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3213875350320641,
+      "learning_rate": 0.0001728877655818478,
+      "loss": 1.1062,
+      "step": 5812
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3270439141750809,
+      "learning_rate": 0.0001728769339509937,
+      "loss": 1.0258,
+      "step": 5813
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3409072299007263,
+      "learning_rate": 0.00017286610049634856,
+      "loss": 0.9644,
+      "step": 5814
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24496972615719473,
+      "learning_rate": 0.00017285526521818346,
+      "loss": 1.1259,
+      "step": 5815
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3139089621120804,
+      "learning_rate": 0.00017284442811676953,
+      "loss": 1.0404,
+      "step": 5816
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26970603580675145,
+      "learning_rate": 0.00017283358919237802,
+      "loss": 1.0835,
+      "step": 5817
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2733732615441018,
+      "learning_rate": 0.0001728227484452802,
+      "loss": 1.0744,
+      "step": 5818
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28036462116083893,
+      "learning_rate": 0.00017281190587574728,
+      "loss": 1.1692,
+      "step": 5819
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2871660582969163,
+      "learning_rate": 0.0001728010614840507,
+      "loss": 1.0495,
+      "step": 5820
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.23884476735212962,
+      "learning_rate": 0.00017279021527046178,
+      "loss": 1.0443,
+      "step": 5821
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2873207540071599,
+      "learning_rate": 0.00017277936723525197,
+      "loss": 0.993,
+      "step": 5822
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30547644132922835,
+      "learning_rate": 0.00017276851737869274,
+      "loss": 1.0843,
+      "step": 5823
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3057414154299913,
+      "learning_rate": 0.00017275766570105567,
+      "loss": 1.0655,
+      "step": 5824
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2632962822442975,
+      "learning_rate": 0.00017274681220261226,
+      "loss": 1.0939,
+      "step": 5825
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30877886648124775,
+      "learning_rate": 0.00017273595688363416,
+      "loss": 1.1277,
+      "step": 5826
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26307234053148654,
+      "learning_rate": 0.00017272509974439304,
+      "loss": 1.0079,
+      "step": 5827
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2841767029229268,
+      "learning_rate": 0.00017271424078516055,
+      "loss": 1.1768,
+      "step": 5828
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2603214619363828,
+      "learning_rate": 0.00017270338000620856,
+      "loss": 1.0806,
+      "step": 5829
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2545067452234432,
+      "learning_rate": 0.00017269251740780874,
+      "loss": 1.0412,
+      "step": 5830
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27855295154779947,
+      "learning_rate": 0.000172681652990233,
+      "loss": 1.2206,
+      "step": 5831
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28261805590758377,
+      "learning_rate": 0.00017267078675375322,
+      "loss": 1.0033,
+      "step": 5832
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27667495248962876,
+      "learning_rate": 0.00017265991869864128,
+      "loss": 1.0831,
+      "step": 5833
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2714857639489485,
+      "learning_rate": 0.00017264904882516928,
+      "loss": 1.109,
+      "step": 5834
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27229604294119086,
+      "learning_rate": 0.00017263817713360915,
+      "loss": 1.0874,
+      "step": 5835
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.293920707567098,
+      "learning_rate": 0.00017262730362423297,
+      "loss": 1.0823,
+      "step": 5836
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2613726158263339,
+      "learning_rate": 0.00017261642829731287,
+      "loss": 1.0599,
+      "step": 5837
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2889492784232317,
+      "learning_rate": 0.00017260555115312104,
+      "loss": 1.1224,
+      "step": 5838
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25970536207303385,
+      "learning_rate": 0.00017259467219192968,
+      "loss": 1.1015,
+      "step": 5839
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28971995596673233,
+      "learning_rate": 0.00017258379141401098,
+      "loss": 1.0696,
+      "step": 5840
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30430404198918504,
+      "learning_rate": 0.00017257290881963732,
+      "loss": 1.0823,
+      "step": 5841
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2713987221161718,
+      "learning_rate": 0.00017256202440908095,
+      "loss": 1.0604,
+      "step": 5842
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3001859725452708,
+      "learning_rate": 0.00017255113818261437,
+      "loss": 0.8537,
+      "step": 5843
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24037659758164331,
+      "learning_rate": 0.00017254025014050995,
+      "loss": 1.0035,
+      "step": 5844
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29289291269851125,
+      "learning_rate": 0.00017252936028304015,
+      "loss": 0.9527,
+      "step": 5845
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28320339512377446,
+      "learning_rate": 0.00017251846861047755,
+      "loss": 1.0616,
+      "step": 5846
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2844954523450815,
+      "learning_rate": 0.0001725075751230947,
+      "loss": 1.1248,
+      "step": 5847
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25465891657756395,
+      "learning_rate": 0.0001724966798211642,
+      "loss": 1.2268,
+      "step": 5848
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.318243854381763,
+      "learning_rate": 0.00017248578270495873,
+      "loss": 1.1983,
+      "step": 5849
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.24392495708256506,
+      "learning_rate": 0.00017247488377475102,
+      "loss": 1.1131,
+      "step": 5850
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.339924875849168,
+      "learning_rate": 0.00017246398303081377,
+      "loss": 1.1255,
+      "step": 5851
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2824682986273372,
+      "learning_rate": 0.00017245308047341977,
+      "loss": 1.0928,
+      "step": 5852
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.294553844921709,
+      "learning_rate": 0.00017244217610284194,
+      "loss": 1.1941,
+      "step": 5853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3175139266139292,
+      "learning_rate": 0.0001724312699193531,
+      "loss": 1.0488,
+      "step": 5854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.26336543686584807,
+      "learning_rate": 0.0001724203619232262,
+      "loss": 1.0406,
+      "step": 5855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28698720470502975,
+      "learning_rate": 0.00017240945211473426,
+      "loss": 1.1673,
+      "step": 5856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2705696042656757,
+      "learning_rate": 0.0001723985404941503,
+      "loss": 1.0654,
+      "step": 5857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2609774023108237,
+      "learning_rate": 0.0001723876270617473,
+      "loss": 1.0166,
+      "step": 5858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2879084621496255,
+      "learning_rate": 0.0001723767118177985,
+      "loss": 1.1731,
+      "step": 5859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.25995995206356726,
+      "learning_rate": 0.00017236579476257694,
+      "loss": 1.1141,
+      "step": 5860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2832052274386346,
+      "learning_rate": 0.00017235487589635593,
+      "loss": 1.0356,
+      "step": 5861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3009633175427442,
+      "learning_rate": 0.00017234395521940866,
+      "loss": 1.0634,
+      "step": 5862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28724524890382835,
+      "learning_rate": 0.00017233303273200842,
+      "loss": 1.0978,
+      "step": 5863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30150303153924973,
+      "learning_rate": 0.0001723221084344286,
+      "loss": 1.0883,
+      "step": 5864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2602169228739236,
+      "learning_rate": 0.00017231118232694255,
+      "loss": 0.8809,
+      "step": 5865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2597086205288419,
+      "learning_rate": 0.00017230025440982373,
+      "loss": 1.0913,
+      "step": 5866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2627390421998769,
+      "learning_rate": 0.0001722893246833456,
+      "loss": 0.9786,
+      "step": 5867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29086276852649495,
+      "learning_rate": 0.0001722783931477817,
+      "loss": 1.1312,
+      "step": 5868
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2631914398523322,
+      "learning_rate": 0.00017226745980340556,
+      "loss": 1.1383,
+      "step": 5869
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.293675216189509,
+      "learning_rate": 0.00017225652465049086,
+      "loss": 1.037,
+      "step": 5870
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27269400481933986,
+      "learning_rate": 0.0001722455876893112,
+      "loss": 1.0594,
+      "step": 5871
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2913003315526418,
+      "learning_rate": 0.00017223464892014028,
+      "loss": 1.0343,
+      "step": 5872
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30749338727180847,
+      "learning_rate": 0.0001722237083432519,
+      "loss": 1.0609,
+      "step": 5873
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2780051476505012,
+      "learning_rate": 0.00017221276595891984,
+      "loss": 1.0403,
+      "step": 5874
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28351257017042697,
+      "learning_rate": 0.0001722018217674179,
+      "loss": 1.1222,
+      "step": 5875
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30556461707747595,
+      "learning_rate": 0.00017219087576902,
+      "loss": 1.1095,
+      "step": 5876
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2747330702712088,
+      "learning_rate": 0.00017217992796400005,
+      "loss": 1.0258,
+      "step": 5877
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3102546058358946,
+      "learning_rate": 0.00017216897835263209,
+      "loss": 0.9957,
+      "step": 5878
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3232197983937078,
+      "learning_rate": 0.00017215802693519003,
+      "loss": 1.0768,
+      "step": 5879
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.269670235437008,
+      "learning_rate": 0.00017214707371194802,
+      "loss": 1.1033,
+      "step": 5880
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2633897286188578,
+      "learning_rate": 0.00017213611868318015,
+      "loss": 1.0035,
+      "step": 5881
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3042144256612858,
+      "learning_rate": 0.00017212516184916056,
+      "loss": 1.0715,
+      "step": 5882
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.266078239126639,
+      "learning_rate": 0.0001721142032101635,
+      "loss": 1.1544,
+      "step": 5883
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30458594940724576,
+      "learning_rate": 0.00017210324276646316,
+      "loss": 1.1251,
+      "step": 5884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28395232279028726,
+      "learning_rate": 0.00017209228051833387,
+      "loss": 1.0925,
+      "step": 5885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29349557855153724,
+      "learning_rate": 0.00017208131646604993,
+      "loss": 1.1322,
+      "step": 5886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2729717369394539,
+      "learning_rate": 0.00017207035060988574,
+      "loss": 1.0009,
+      "step": 5887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.278418729052514,
+      "learning_rate": 0.00017205938295011575,
+      "loss": 1.0637,
+      "step": 5888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2689002291046586,
+      "learning_rate": 0.00017204841348701438,
+      "loss": 1.1421,
+      "step": 5889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2936863871934149,
+      "learning_rate": 0.00017203744222085623,
+      "loss": 1.078,
+      "step": 5890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.291267791636599,
+      "learning_rate": 0.00017202646915191578,
+      "loss": 1.2519,
+      "step": 5891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2717676026532944,
+      "learning_rate": 0.0001720154942804677,
+      "loss": 1.0844,
+      "step": 5892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2774147698904939,
+      "learning_rate": 0.0001720045176067866,
+      "loss": 1.0487,
+      "step": 5893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27887254178612914,
+      "learning_rate": 0.00017199353913114717,
+      "loss": 1.1334,
+      "step": 5894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.28978059963913083,
+      "learning_rate": 0.00017198255885382421,
+      "loss": 1.0868,
+      "step": 5895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2766397079264888,
+      "learning_rate": 0.00017197157677509246,
+      "loss": 1.1068,
+      "step": 5896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2851293309828098,
+      "learning_rate": 0.00017196059289522678,
+      "loss": 1.0717,
+      "step": 5897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2549062649316722,
+      "learning_rate": 0.00017194960721450206,
+      "loss": 1.0199,
+      "step": 5898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2461949098073852,
+      "learning_rate": 0.00017193861973319316,
+      "loss": 1.0461,
+      "step": 5899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27203801042895437,
+      "learning_rate": 0.0001719276304515751,
+      "loss": 1.1691,
+      "step": 5900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.27460790915022437,
+      "learning_rate": 0.00017191663936992288,
+      "loss": 1.105,
+      "step": 5901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29915013883951985,
+      "learning_rate": 0.0001719056464885116,
+      "loss": 0.9848,
+      "step": 5902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.29579312716715256,
+      "learning_rate": 0.00017189465180761628,
+      "loss": 1.1115,
+      "step": 5903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.30761711337398795,
+      "learning_rate": 0.00017188365532751213,
+      "loss": 1.164,
+      "step": 5904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.32410192655697073,
+      "learning_rate": 0.00017187265704847433,
+      "loss": 1.089,
+      "step": 5905
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2629249103887719,
+      "learning_rate": 0.00017186165697077809,
+      "loss": 0.9588,
+      "step": 5906
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28511065582542094,
+      "learning_rate": 0.00017185065509469876,
+      "loss": 1.1679,
+      "step": 5907
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2813371419019583,
+      "learning_rate": 0.00017183965142051163,
+      "loss": 1.0888,
+      "step": 5908
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27646089721769807,
+      "learning_rate": 0.00017182864594849205,
+      "loss": 1.0513,
+      "step": 5909
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30831778021556755,
+      "learning_rate": 0.00017181763867891547,
+      "loss": 1.1152,
+      "step": 5910
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31669929806320846,
+      "learning_rate": 0.00017180662961205733,
+      "loss": 1.0554,
+      "step": 5911
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2881244715767111,
+      "learning_rate": 0.00017179561874819318,
+      "loss": 1.0702,
+      "step": 5912
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28072579932667696,
+      "learning_rate": 0.00017178460608759853,
+      "loss": 0.9758,
+      "step": 5913
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2943649793795249,
+      "learning_rate": 0.00017177359163054903,
+      "loss": 1.0715,
+      "step": 5914
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24478814096705148,
+      "learning_rate": 0.00017176257537732025,
+      "loss": 1.0606,
+      "step": 5915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27922088346647145,
+      "learning_rate": 0.00017175155732818796,
+      "loss": 1.1677,
+      "step": 5916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.25994214062479126,
+      "learning_rate": 0.00017174053748342783,
+      "loss": 0.9742,
+      "step": 5917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.282588772331292,
+      "learning_rate": 0.00017172951584331565,
+      "loss": 0.9751,
+      "step": 5918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2878741945844007,
+      "learning_rate": 0.0001717184924081273,
+      "loss": 1.0041,
+      "step": 5919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3029923936191996,
+      "learning_rate": 0.00017170746717813854,
+      "loss": 1.1071,
+      "step": 5920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27290844629846106,
+      "learning_rate": 0.0001716964401536254,
+      "loss": 0.8786,
+      "step": 5921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31966494850943117,
+      "learning_rate": 0.00017168541133486377,
+      "loss": 1.1118,
+      "step": 5922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2689850435328001,
+      "learning_rate": 0.00017167438072212968,
+      "loss": 1.1024,
+      "step": 5923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27928731290969944,
+      "learning_rate": 0.00017166334831569916,
+      "loss": 1.1399,
+      "step": 5924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29591002661910865,
+      "learning_rate": 0.00017165231411584827,
+      "loss": 1.1074,
+      "step": 5925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3104516805825287,
+      "learning_rate": 0.00017164127812285324,
+      "loss": 1.0186,
+      "step": 5926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3329888234877537,
+      "learning_rate": 0.00017163024033699017,
+      "loss": 1.1095,
+      "step": 5927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27420800800981127,
+      "learning_rate": 0.00017161920075853534,
+      "loss": 1.1286,
+      "step": 5928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2687683336478087,
+      "learning_rate": 0.000171608159387765,
+      "loss": 1.1067,
+      "step": 5929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28848253158286347,
+      "learning_rate": 0.00017159711622495544,
+      "loss": 1.1874,
+      "step": 5930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28239325672690274,
+      "learning_rate": 0.0001715860712703831,
+      "loss": 1.1561,
+      "step": 5931
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2918884091102849,
+      "learning_rate": 0.00017157502452432429,
+      "loss": 1.0429,
+      "step": 5932
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24790386180695678,
+      "learning_rate": 0.00017156397598705548,
+      "loss": 0.9923,
+      "step": 5933
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24570072442193652,
+      "learning_rate": 0.00017155292565885328,
+      "loss": 1.0945,
+      "step": 5934
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.23125439331824077,
+      "learning_rate": 0.0001715418735399941,
+      "loss": 0.9957,
+      "step": 5935
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2554465249970479,
+      "learning_rate": 0.00017153081963075458,
+      "loss": 1.1284,
+      "step": 5936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2794800525821183,
+      "learning_rate": 0.00017151976393141132,
+      "loss": 0.9941,
+      "step": 5937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30723518305938063,
+      "learning_rate": 0.000171508706442241,
+      "loss": 1.1475,
+      "step": 5938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30203130938607553,
+      "learning_rate": 0.00017149764716352045,
+      "loss": 1.0853,
+      "step": 5939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.260328670983559,
+      "learning_rate": 0.00017148658609552627,
+      "loss": 1.1423,
+      "step": 5940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.293768400385368,
+      "learning_rate": 0.00017147552323853538,
+      "loss": 1.0503,
+      "step": 5941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26640544240980046,
+      "learning_rate": 0.00017146445859282457,
+      "loss": 1.0018,
+      "step": 5942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3159499546692293,
+      "learning_rate": 0.00017145339215867078,
+      "loss": 1.0303,
+      "step": 5943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.25980174549043206,
+      "learning_rate": 0.00017144232393635094,
+      "loss": 1.1472,
+      "step": 5944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2727359561908714,
+      "learning_rate": 0.00017143125392614207,
+      "loss": 1.1425,
+      "step": 5945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28270017496530925,
+      "learning_rate": 0.0001714201821283212,
+      "loss": 1.1023,
+      "step": 5946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.299277933501389,
+      "learning_rate": 0.0001714091085431653,
+      "loss": 1.1593,
+      "step": 5947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2762476973789334,
+      "learning_rate": 0.00017139803317095165,
+      "loss": 1.0942,
+      "step": 5948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3120883708584473,
+      "learning_rate": 0.00017138695601195733,
+      "loss": 1.0574,
+      "step": 5949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27024241469527543,
+      "learning_rate": 0.00017137587706645956,
+      "loss": 1.12,
+      "step": 5950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2584604551495038,
+      "learning_rate": 0.00017136479633473562,
+      "loss": 1.0682,
+      "step": 5951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2525657024101026,
+      "learning_rate": 0.0001713537138170628,
+      "loss": 1.1037,
+      "step": 5952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28274675021766404,
+      "learning_rate": 0.00017134262951371842,
+      "loss": 1.0135,
+      "step": 5953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2774062534984412,
+      "learning_rate": 0.00017133154342497995,
+      "loss": 0.9824,
+      "step": 5954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2832539855963225,
+      "learning_rate": 0.00017132045555112474,
+      "loss": 0.9708,
+      "step": 5955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.24641803171377838,
+      "learning_rate": 0.0001713093658924303,
+      "loss": 0.9826,
+      "step": 5956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29062056115142054,
+      "learning_rate": 0.0001712982744491742,
+      "loss": 0.9739,
+      "step": 5957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2922884188711339,
+      "learning_rate": 0.00017128718122163395,
+      "loss": 0.9959,
+      "step": 5958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26793724543335196,
+      "learning_rate": 0.00017127608621008718,
+      "loss": 1.0553,
+      "step": 5959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.276065664300244,
+      "learning_rate": 0.00017126498941481155,
+      "loss": 1.1004,
+      "step": 5960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26445107517178895,
+      "learning_rate": 0.00017125389083608479,
+      "loss": 1.0136,
+      "step": 5961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2724968692637002,
+      "learning_rate": 0.00017124279047418464,
+      "loss": 0.9554,
+      "step": 5962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2638896257757664,
+      "learning_rate": 0.00017123168832938886,
+      "loss": 1.0249,
+      "step": 5963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.25113449381499126,
+      "learning_rate": 0.00017122058440197533,
+      "loss": 1.0309,
+      "step": 5964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29069944989919566,
+      "learning_rate": 0.0001712094786922219,
+      "loss": 1.0919,
+      "step": 5965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28312314556833795,
+      "learning_rate": 0.00017119837120040652,
+      "loss": 1.0403,
+      "step": 5966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26755188643974503,
+      "learning_rate": 0.00017118726192680717,
+      "loss": 1.0553,
+      "step": 5967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26486022885030275,
+      "learning_rate": 0.0001711761508717018,
+      "loss": 1.0489,
+      "step": 5968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3020328536278636,
+      "learning_rate": 0.00017116503803536856,
+      "loss": 1.1584,
+      "step": 5969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34335564008695746,
+      "learning_rate": 0.00017115392341808555,
+      "loss": 1.1107,
+      "step": 5970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30126089791181193,
+      "learning_rate": 0.00017114280702013084,
+      "loss": 1.0434,
+      "step": 5971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28087942592247517,
+      "learning_rate": 0.00017113168884178267,
+      "loss": 1.0781,
+      "step": 5972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2728680461126566,
+      "learning_rate": 0.0001711205688833193,
+      "loss": 1.0257,
+      "step": 5973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2985709605948755,
+      "learning_rate": 0.000171109447145019,
+      "loss": 1.1728,
+      "step": 5974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3082359218373612,
+      "learning_rate": 0.0001710983236271601,
+      "loss": 1.0554,
+      "step": 5975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2795284645154645,
+      "learning_rate": 0.00017108719833002094,
+      "loss": 1.0948,
+      "step": 5976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28120362222029427,
+      "learning_rate": 0.00017107607125387998,
+      "loss": 0.9671,
+      "step": 5977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2823788055491324,
+      "learning_rate": 0.00017106494239901566,
+      "loss": 1.0073,
+      "step": 5978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31065181033843386,
+      "learning_rate": 0.00017105381176570652,
+      "loss": 1.1014,
+      "step": 5979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30900428726246143,
+      "learning_rate": 0.00017104267935423107,
+      "loss": 1.0998,
+      "step": 5980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30988350895081024,
+      "learning_rate": 0.0001710315451648679,
+      "loss": 1.0667,
+      "step": 5981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2981798835140071,
+      "learning_rate": 0.0001710204091978957,
+      "loss": 1.065,
+      "step": 5982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2732778988161561,
+      "learning_rate": 0.0001710092714535931,
+      "loss": 1.184,
+      "step": 5983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27164434920682934,
+      "learning_rate": 0.00017099813193223887,
+      "loss": 1.0524,
+      "step": 5984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29410498736707635,
+      "learning_rate": 0.00017098699063411178,
+      "loss": 1.0777,
+      "step": 5985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28163842201092487,
+      "learning_rate": 0.0001709758475594906,
+      "loss": 1.0853,
+      "step": 5986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27782202027455466,
+      "learning_rate": 0.00017096470270865427,
+      "loss": 1.0146,
+      "step": 5987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2768790776720804,
+      "learning_rate": 0.0001709535560818816,
+      "loss": 1.1312,
+      "step": 5988
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2824780756717249,
+      "learning_rate": 0.00017094240767945166,
+      "loss": 1.059,
+      "step": 5989
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27033728590320183,
+      "learning_rate": 0.00017093125750164333,
+      "loss": 1.197,
+      "step": 5990
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3100948675967715,
+      "learning_rate": 0.00017092010554873574,
+      "loss": 1.1042,
+      "step": 5991
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28820788597927405,
+      "learning_rate": 0.0001709089518210079,
+      "loss": 1.1635,
+      "step": 5992
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.28201539629588646,
+      "learning_rate": 0.00017089779631873904,
+      "loss": 0.9284,
+      "step": 5993
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27679231851615277,
+      "learning_rate": 0.0001708866390422082,
+      "loss": 1.0167,
+      "step": 5994
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27944884514951135,
+      "learning_rate": 0.0001708754799916947,
+      "loss": 1.139,
+      "step": 5995
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3002084383540107,
+      "learning_rate": 0.00017086431916747778,
+      "loss": 1.0454,
+      "step": 5996
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.26816351694504426,
+      "learning_rate": 0.0001708531565698367,
+      "loss": 1.0773,
+      "step": 5997
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30990517015210073,
+      "learning_rate": 0.0001708419921990509,
+      "loss": 1.1122,
+      "step": 5998
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.30574794780484754,
+      "learning_rate": 0.0001708308260553997,
+      "loss": 1.1214,
+      "step": 5999
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2690373353619774,
+      "learning_rate": 0.00017081965813916253,
+      "loss": 1.062,
+      "step": 6000
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27472519148333513,
+      "learning_rate": 0.00017080848845061892,
+      "loss": 1.0179,
+      "step": 6001
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.27470990890182256,
+      "learning_rate": 0.0001707973169900484,
+      "loss": 1.0747,
+      "step": 6002
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.29061144311513604,
+      "learning_rate": 0.00017078614375773052,
+      "loss": 1.1307,
+      "step": 6003
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2713055223809409,
+      "learning_rate": 0.00017077496875394493,
+      "loss": 1.0008,
+      "step": 6004
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2925514548076972,
+      "learning_rate": 0.00017076379197897122,
+      "loss": 1.1317,
+      "step": 6005
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.31101769602212137,
+      "learning_rate": 0.00017075261343308916,
+      "loss": 1.1037,
+      "step": 6006
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2647833297981797,
+      "learning_rate": 0.00017074143311657852,
+      "loss": 1.0076,
+      "step": 6007
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3013454403605899,
+      "learning_rate": 0.00017073025102971903,
+      "loss": 1.0792,
+      "step": 6008
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.263491300792746,
+      "learning_rate": 0.00017071906717279053,
+      "loss": 0.9797,
+      "step": 6009
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2724539339009908,
+      "learning_rate": 0.00017070788154607293,
+      "loss": 0.9805,
+      "step": 6010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.31840312141141264,
+      "learning_rate": 0.00017069669414984618,
+      "loss": 1.142,
+      "step": 6011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2983653164774024,
+      "learning_rate": 0.00017068550498439025,
+      "loss": 1.103,
+      "step": 6012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.282150969685538,
+      "learning_rate": 0.00017067431404998507,
+      "loss": 1.0682,
+      "step": 6013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2771901829735762,
+      "learning_rate": 0.00017066312134691083,
+      "loss": 1.009,
+      "step": 6014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2933084300533384,
+      "learning_rate": 0.00017065192687544753,
+      "loss": 1.102,
+      "step": 6015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27679497108587947,
+      "learning_rate": 0.00017064073063587535,
+      "loss": 1.1393,
+      "step": 6016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28895991711416996,
+      "learning_rate": 0.00017062953262847455,
+      "loss": 1.0694,
+      "step": 6017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27823308988900375,
+      "learning_rate": 0.00017061833285352527,
+      "loss": 1.0905,
+      "step": 6018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2748970724583933,
+      "learning_rate": 0.00017060713131130778,
+      "loss": 1.1278,
+      "step": 6019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2730226539296915,
+      "learning_rate": 0.00017059592800210252,
+      "loss": 1.0858,
+      "step": 6020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.271328807355675,
+      "learning_rate": 0.00017058472292618977,
+      "loss": 0.9972,
+      "step": 6021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27544958968420746,
+      "learning_rate": 0.00017057351608384995,
+      "loss": 1.1456,
+      "step": 6022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2833122301378086,
+      "learning_rate": 0.00017056230747536355,
+      "loss": 1.1727,
+      "step": 6023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2894626470697644,
+      "learning_rate": 0.00017055109710101108,
+      "loss": 1.2121,
+      "step": 6024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3306470002946546,
+      "learning_rate": 0.00017053988496107305,
+      "loss": 1.0776,
+      "step": 6025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27472508074504254,
+      "learning_rate": 0.00017052867105583005,
+      "loss": 0.973,
+      "step": 6026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3021983840922597,
+      "learning_rate": 0.00017051745538556278,
+      "loss": 1.0972,
+      "step": 6027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2970721247856769,
+      "learning_rate": 0.0001705062379505518,
+      "loss": 1.1562,
+      "step": 6028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2830455403468363,
+      "learning_rate": 0.00017049501875107795,
+      "loss": 0.9997,
+      "step": 6029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28710721053351523,
+      "learning_rate": 0.00017048379778742193,
+      "loss": 1.1453,
+      "step": 6030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27478245168556914,
+      "learning_rate": 0.00017047257505986457,
+      "loss": 1.0092,
+      "step": 6031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26896422201399867,
+      "learning_rate": 0.00017046135056868677,
+      "loss": 1.0855,
+      "step": 6032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24707636757742915,
+      "learning_rate": 0.00017045012431416936,
+      "loss": 1.0971,
+      "step": 6033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2493183820340796,
+      "learning_rate": 0.0001704388962965933,
+      "loss": 0.9823,
+      "step": 6034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2805800589880677,
+      "learning_rate": 0.00017042766651623962,
+      "loss": 1.0523,
+      "step": 6035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3038587968507171,
+      "learning_rate": 0.00017041643497338931,
+      "loss": 1.0162,
+      "step": 6036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30638016511222943,
+      "learning_rate": 0.00017040520166832344,
+      "loss": 1.0421,
+      "step": 6037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26320036877135733,
+      "learning_rate": 0.00017039396660132317,
+      "loss": 0.946,
+      "step": 6038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2754763891016706,
+      "learning_rate": 0.00017038272977266966,
+      "loss": 1.0652,
+      "step": 6039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26134384808972805,
+      "learning_rate": 0.0001703714911826441,
+      "loss": 1.1064,
+      "step": 6040
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2651686121986517,
+      "learning_rate": 0.0001703602508315277,
+      "loss": 1.1166,
+      "step": 6041
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25447003550809477,
+      "learning_rate": 0.00017034900871960184,
+      "loss": 1.0199,
+      "step": 6042
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28285744813232555,
+      "learning_rate": 0.0001703377648471478,
+      "loss": 1.0086,
+      "step": 6043
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2651737161168531,
+      "learning_rate": 0.000170326519214447,
+      "loss": 1.0729,
+      "step": 6044
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2855727067410992,
+      "learning_rate": 0.00017031527182178092,
+      "loss": 1.0919,
+      "step": 6045
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26177841865296425,
+      "learning_rate": 0.0001703040226694309,
+      "loss": 0.9408,
+      "step": 6046
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23691819299378247,
+      "learning_rate": 0.00017029277175767854,
+      "loss": 1.0849,
+      "step": 6047
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2511952002294269,
+      "learning_rate": 0.0001702815190868054,
+      "loss": 1.0779,
+      "step": 6048
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29110924767645496,
+      "learning_rate": 0.00017027026465709307,
+      "loss": 0.9933,
+      "step": 6049
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2830264573548182,
+      "learning_rate": 0.00017025900846882321,
+      "loss": 1.0192,
+      "step": 6050
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28455741758111286,
+      "learning_rate": 0.00017024775052227752,
+      "loss": 1.0588,
+      "step": 6051
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28932106749302483,
+      "learning_rate": 0.0001702364908177377,
+      "loss": 1.1211,
+      "step": 6052
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3307206565435221,
+      "learning_rate": 0.00017022522935548554,
+      "loss": 1.0975,
+      "step": 6053
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25352957645066343,
+      "learning_rate": 0.0001702139661358029,
+      "loss": 1.0298,
+      "step": 6054
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2572310440026188,
+      "learning_rate": 0.00017020270115897164,
+      "loss": 1.0728,
+      "step": 6055
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3172532579427463,
+      "learning_rate": 0.00017019143442527365,
+      "loss": 1.1023,
+      "step": 6056
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.282565292417829,
+      "learning_rate": 0.0001701801659349909,
+      "loss": 1.0488,
+      "step": 6057
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.275135543466168,
+      "learning_rate": 0.00017016889568840542,
+      "loss": 1.1721,
+      "step": 6058
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26182321188521385,
+      "learning_rate": 0.00017015762368579918,
+      "loss": 1.0598,
+      "step": 6059
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26125752561442744,
+      "learning_rate": 0.00017014634992745434,
+      "loss": 1.1684,
+      "step": 6060
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27290128374782624,
+      "learning_rate": 0.000170135074413653,
+      "loss": 0.9901,
+      "step": 6061
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2653240038158296,
+      "learning_rate": 0.00017012379714467736,
+      "loss": 1.1086,
+      "step": 6062
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3293682851680674,
+      "learning_rate": 0.0001701125181208096,
+      "loss": 1.117,
+      "step": 6063
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2329956724588023,
+      "learning_rate": 0.00017010123734233204,
+      "loss": 1.0925,
+      "step": 6064
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2755426850023404,
+      "learning_rate": 0.00017008995480952694,
+      "loss": 1.1292,
+      "step": 6065
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29650880006845154,
+      "learning_rate": 0.00017007867052267666,
+      "loss": 1.0334,
+      "step": 6066
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2803458685525201,
+      "learning_rate": 0.00017006738448206363,
+      "loss": 1.0168,
+      "step": 6067
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2916272901141261,
+      "learning_rate": 0.00017005609668797024,
+      "loss": 1.0039,
+      "step": 6068
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2691648334575106,
+      "learning_rate": 0.00017004480714067903,
+      "loss": 1.0727,
+      "step": 6069
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3270372768288688,
+      "learning_rate": 0.00017003351584047249,
+      "loss": 1.1879,
+      "step": 6070
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2782777828019017,
+      "learning_rate": 0.0001700222227876332,
+      "loss": 1.0006,
+      "step": 6071
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2661246430524472,
+      "learning_rate": 0.00017001092798244377,
+      "loss": 1.0393,
+      "step": 6072
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.281715885539005,
+      "learning_rate": 0.00016999963142518687,
+      "loss": 1.1322,
+      "step": 6073
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2986783823949029,
+      "learning_rate": 0.0001699883331161452,
+      "loss": 0.9984,
+      "step": 6074
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28065012498362923,
+      "learning_rate": 0.00016997703305560153,
+      "loss": 1.06,
+      "step": 6075
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3292647162053623,
+      "learning_rate": 0.00016996573124383862,
+      "loss": 1.075,
+      "step": 6076
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3073957042018125,
+      "learning_rate": 0.0001699544276811393,
+      "loss": 1.1219,
+      "step": 6077
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3076254806022189,
+      "learning_rate": 0.00016994312236778646,
+      "loss": 1.1214,
+      "step": 6078
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24563100316558975,
+      "learning_rate": 0.00016993181530406304,
+      "loss": 1.0868,
+      "step": 6079
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.32781029130768446,
+      "learning_rate": 0.00016992050649025197,
+      "loss": 1.0481,
+      "step": 6080
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27854015122577225,
+      "learning_rate": 0.0001699091959266363,
+      "loss": 1.0749,
+      "step": 6081
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2844546319680473,
+      "learning_rate": 0.00016989788361349906,
+      "loss": 1.0692,
+      "step": 6082
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26692237610439323,
+      "learning_rate": 0.00016988656955112337,
+      "loss": 1.1151,
+      "step": 6083
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26269204364410026,
+      "learning_rate": 0.00016987525373979233,
+      "loss": 0.9348,
+      "step": 6084
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27306822357867006,
+      "learning_rate": 0.00016986393617978918,
+      "loss": 1.2159,
+      "step": 6085
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2827921168255787,
+      "learning_rate": 0.0001698526168713971,
+      "loss": 0.9376,
+      "step": 6086
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.23947335215371623,
+      "learning_rate": 0.00016984129581489935,
+      "loss": 1.0817,
+      "step": 6087
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3343810073022071,
+      "learning_rate": 0.0001698299730105793,
+      "loss": 1.1555,
+      "step": 6088
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26515752282559285,
+      "learning_rate": 0.00016981864845872033,
+      "loss": 1.1123,
+      "step": 6089
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.27681937200219026,
+      "learning_rate": 0.00016980732215960575,
+      "loss": 1.0367,
+      "step": 6090
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2519010649566014,
+      "learning_rate": 0.0001697959941135191,
+      "loss": 1.0861,
+      "step": 6091
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2704443985177932,
+      "learning_rate": 0.00016978466432074381,
+      "loss": 1.0398,
+      "step": 6092
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28884051221369256,
+      "learning_rate": 0.00016977333278156347,
+      "loss": 1.0501,
+      "step": 6093
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2749924386950333,
+      "learning_rate": 0.0001697619994962616,
+      "loss": 1.0743,
+      "step": 6094
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.279690239046547,
+      "learning_rate": 0.00016975066446512185,
+      "loss": 1.1504,
+      "step": 6095
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24616630543118756,
+      "learning_rate": 0.00016973932768842787,
+      "loss": 1.1121,
+      "step": 6096
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.25087990953142,
+      "learning_rate": 0.00016972798916646336,
+      "loss": 1.0454,
+      "step": 6097
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2766189199572125,
+      "learning_rate": 0.00016971664889951215,
+      "loss": 1.0582,
+      "step": 6098
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28980752883287725,
+      "learning_rate": 0.00016970530688785798,
+      "loss": 1.1664,
+      "step": 6099
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.334844857943314,
+      "learning_rate": 0.0001696939631317847,
+      "loss": 1.1833,
+      "step": 6100
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2923718680275688,
+      "learning_rate": 0.0001696826176315762,
+      "loss": 1.063,
+      "step": 6101
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2927999981743822,
+      "learning_rate": 0.00016967127038751637,
+      "loss": 1.0568,
+      "step": 6102
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30725008382724595,
+      "learning_rate": 0.0001696599213998892,
+      "loss": 1.113,
+      "step": 6103
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2864392247154817,
+      "learning_rate": 0.00016964857066897876,
+      "loss": 1.0959,
+      "step": 6104
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29730179190891975,
+      "learning_rate": 0.00016963721819506904,
+      "loss": 1.0681,
+      "step": 6105
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2338409227151746,
+      "learning_rate": 0.00016962586397844417,
+      "loss": 0.9203,
+      "step": 6106
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.30140296809951767,
+      "learning_rate": 0.0001696145080193883,
+      "loss": 1.1501,
+      "step": 6107
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.29957908772250824,
+      "learning_rate": 0.00016960315031818563,
+      "loss": 0.9797,
+      "step": 6108
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2703387299476374,
+      "learning_rate": 0.00016959179087512038,
+      "loss": 1.1028,
+      "step": 6109
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.24345511227668115,
+      "learning_rate": 0.0001695804296904768,
+      "loss": 1.1393,
+      "step": 6110
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.28466641760021855,
+      "learning_rate": 0.00016956906676453927,
+      "loss": 1.0691,
+      "step": 6111
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.26942292513037813,
+      "learning_rate": 0.00016955770209759206,
+      "loss": 1.1351,
+      "step": 6112
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2432370590626464,
+      "learning_rate": 0.0001695463356899197,
+      "loss": 1.111,
+      "step": 6113
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.2768776116799371,
+      "learning_rate": 0.00016953496754180657,
+      "loss": 1.0402,
+      "step": 6114
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2782542666733751,
+      "learning_rate": 0.00016952359765353716,
+      "loss": 1.016,
+      "step": 6115
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2606113670203369,
+      "learning_rate": 0.00016951222602539604,
+      "loss": 1.1145,
+      "step": 6116
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28152618216858916,
+      "learning_rate": 0.00016950085265766775,
+      "loss": 1.0144,
+      "step": 6117
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2762557373966403,
+      "learning_rate": 0.000169489477550637,
+      "loss": 1.0188,
+      "step": 6118
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27464396847770955,
+      "learning_rate": 0.00016947810070458836,
+      "loss": 1.0246,
+      "step": 6119
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2553714383261195,
+      "learning_rate": 0.00016946672211980656,
+      "loss": 1.0676,
+      "step": 6120
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2663697635305504,
+      "learning_rate": 0.00016945534179657642,
+      "loss": 1.0607,
+      "step": 6121
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2724085363177121,
+      "learning_rate": 0.00016944395973518273,
+      "loss": 1.0022,
+      "step": 6122
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2830680768080686,
+      "learning_rate": 0.00016943257593591025,
+      "loss": 1.1051,
+      "step": 6123
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28224872460306244,
+      "learning_rate": 0.00016942119039904392,
+      "loss": 1.066,
+      "step": 6124
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26377009917280125,
+      "learning_rate": 0.0001694098031248687,
+      "loss": 1.0324,
+      "step": 6125
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2718917219979864,
+      "learning_rate": 0.0001693984141136695,
+      "loss": 1.0499,
+      "step": 6126
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.259809210248362,
+      "learning_rate": 0.0001693870233657314,
+      "loss": 1.0218,
+      "step": 6127
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2757737836543942,
+      "learning_rate": 0.00016937563088133942,
+      "loss": 1.0728,
+      "step": 6128
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3068744212914902,
+      "learning_rate": 0.0001693642366607787,
+      "loss": 1.0689,
+      "step": 6129
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2559474725346445,
+      "learning_rate": 0.00016935284070433436,
+      "loss": 0.9259,
+      "step": 6130
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2889397480153887,
+      "learning_rate": 0.00016934144301229155,
+      "loss": 1.1747,
+      "step": 6131
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3088905871334589,
+      "learning_rate": 0.0001693300435849356,
+      "loss": 1.172,
+      "step": 6132
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30471384991519385,
+      "learning_rate": 0.00016931864242255171,
+      "loss": 1.0735,
+      "step": 6133
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5872459177211959,
+      "learning_rate": 0.00016930723952542523,
+      "loss": 0.9714,
+      "step": 6134
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26880571869976677,
+      "learning_rate": 0.0001692958348938415,
+      "loss": 1.0637,
+      "step": 6135
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2842892019706417,
+      "learning_rate": 0.000169284428528086,
+      "loss": 1.0986,
+      "step": 6136
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3139992368614553,
+      "learning_rate": 0.0001692730204284441,
+      "loss": 1.0542,
+      "step": 6137
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2939973190730885,
+      "learning_rate": 0.00016926161059520133,
+      "loss": 1.018,
+      "step": 6138
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2745235711764705,
+      "learning_rate": 0.00016925019902864325,
+      "loss": 1.0936,
+      "step": 6139
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.295538764046495,
+      "learning_rate": 0.0001692387857290554,
+      "loss": 1.238,
+      "step": 6140
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2918293376304929,
+      "learning_rate": 0.00016922737069672344,
+      "loss": 1.0324,
+      "step": 6141
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2661937879474406,
+      "learning_rate": 0.000169215953931933,
+      "loss": 1.1461,
+      "step": 6142
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29698439715428343,
+      "learning_rate": 0.0001692045354349698,
+      "loss": 1.0072,
+      "step": 6143
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2841385747639344,
+      "learning_rate": 0.0001691931152061196,
+      "loss": 1.0169,
+      "step": 6144
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2525826472239577,
+      "learning_rate": 0.00016918169324566825,
+      "loss": 1.0821,
+      "step": 6145
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30426659332063666,
+      "learning_rate": 0.0001691702695539015,
+      "loss": 1.0161,
+      "step": 6146
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3247256246142582,
+      "learning_rate": 0.00016915884413110536,
+      "loss": 1.0768,
+      "step": 6147
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28154171353151247,
+      "learning_rate": 0.0001691474169775656,
+      "loss": 1.0462,
+      "step": 6148
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32206989451215157,
+      "learning_rate": 0.0001691359880935683,
+      "loss": 1.0335,
+      "step": 6149
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29975094798814433,
+      "learning_rate": 0.00016912455747939946,
+      "loss": 1.08,
+      "step": 6150
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2940671992246637,
+      "learning_rate": 0.0001691131251353451,
+      "loss": 1.0279,
+      "step": 6151
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30973218098094907,
+      "learning_rate": 0.00016910169106169138,
+      "loss": 1.1113,
+      "step": 6152
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2985525692876313,
+      "learning_rate": 0.0001690902552587244,
+      "loss": 1.0637,
+      "step": 6153
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2539451576750145,
+      "learning_rate": 0.00016907881772673032,
+      "loss": 1.0424,
+      "step": 6154
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2687758329509323,
+      "learning_rate": 0.00016906737846599548,
+      "loss": 1.0618,
+      "step": 6155
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27971131955326073,
+      "learning_rate": 0.00016905593747680602,
+      "loss": 1.0893,
+      "step": 6156
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2543125383692707,
+      "learning_rate": 0.00016904449475944837,
+      "loss": 1.1021,
+      "step": 6157
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2789446404233202,
+      "learning_rate": 0.00016903305031420885,
+      "loss": 1.1103,
+      "step": 6158
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28334066222200344,
+      "learning_rate": 0.00016902160414137383,
+      "loss": 1.065,
+      "step": 6159
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2358577232682755,
+      "learning_rate": 0.0001690101562412298,
+      "loss": 1.021,
+      "step": 6160
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27559840264421875,
+      "learning_rate": 0.00016899870661406325,
+      "loss": 1.1527,
+      "step": 6161
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30571401380369834,
+      "learning_rate": 0.0001689872552601607,
+      "loss": 1.1118,
+      "step": 6162
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29488378034589946,
+      "learning_rate": 0.00016897580217980872,
+      "loss": 1.1922,
+      "step": 6163
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31253654893817673,
+      "learning_rate": 0.00016896434737329394,
+      "loss": 0.9932,
+      "step": 6164
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.271362147934014,
+      "learning_rate": 0.00016895289084090304,
+      "loss": 1.1036,
+      "step": 6165
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2703212529224629,
+      "learning_rate": 0.0001689414325829227,
+      "loss": 1.1416,
+      "step": 6166
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2824957796659181,
+      "learning_rate": 0.0001689299725996397,
+      "loss": 1.0145,
+      "step": 6167
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2976711651462754,
+      "learning_rate": 0.00016891851089134079,
+      "loss": 1.0386,
+      "step": 6168
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26975195101683597,
+      "learning_rate": 0.00016890704745831282,
+      "loss": 1.0879,
+      "step": 6169
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2643048862785792,
+      "learning_rate": 0.00016889558230084273,
+      "loss": 1.0786,
+      "step": 6170
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2717495195222049,
+      "learning_rate": 0.00016888411541921735,
+      "loss": 1.1159,
+      "step": 6171
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2833620530227191,
+      "learning_rate": 0.0001688726468137237,
+      "loss": 1.097,
+      "step": 6172
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25257033358758985,
+      "learning_rate": 0.00016886117648464878,
+      "loss": 1.0967,
+      "step": 6173
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27530661316906846,
+      "learning_rate": 0.00016884970443227965,
+      "loss": 1.0395,
+      "step": 6174
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2527678120212698,
+      "learning_rate": 0.00016883823065690337,
+      "loss": 1.0305,
+      "step": 6175
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2574184683768036,
+      "learning_rate": 0.00016882675515880714,
+      "loss": 1.0878,
+      "step": 6176
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25761118308699854,
+      "learning_rate": 0.00016881527793827808,
+      "loss": 0.9574,
+      "step": 6177
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27107127416113136,
+      "learning_rate": 0.00016880379899560346,
+      "loss": 1.1092,
+      "step": 6178
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2842811642803488,
+      "learning_rate": 0.0001687923183310705,
+      "loss": 1.1637,
+      "step": 6179
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2728440826452907,
+      "learning_rate": 0.00016878083594496656,
+      "loss": 0.9771,
+      "step": 6180
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2886790568351847,
+      "learning_rate": 0.000168769351837579,
+      "loss": 0.9462,
+      "step": 6181
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.25378061719229456,
+      "learning_rate": 0.00016875786600919514,
+      "loss": 1.0195,
+      "step": 6182
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2942815181656749,
+      "learning_rate": 0.00016874637846010248,
+      "loss": 1.1533,
+      "step": 6183
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.26947359910559543,
+      "learning_rate": 0.00016873488919058854,
+      "loss": 0.9764,
+      "step": 6184
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2997412184794112,
+      "learning_rate": 0.00016872339820094074,
+      "loss": 1.0704,
+      "step": 6185
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.241116336668137,
+      "learning_rate": 0.00016871190549144673,
+      "loss": 1.0213,
+      "step": 6186
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29798499768309145,
+      "learning_rate": 0.00016870041106239412,
+      "loss": 1.0051,
+      "step": 6187
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28286155546938635,
+      "learning_rate": 0.00016868891491407054,
+      "loss": 1.1061,
+      "step": 6188
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24398969837876033,
+      "learning_rate": 0.00016867741704676368,
+      "loss": 1.0401,
+      "step": 6189
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28740259957474623,
+      "learning_rate": 0.0001686659174607613,
+      "loss": 1.1156,
+      "step": 6190
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2892058554716867,
+      "learning_rate": 0.0001686544161563512,
+      "loss": 1.0922,
+      "step": 6191
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31955410366097675,
+      "learning_rate": 0.00016864291313382115,
+      "loss": 1.0641,
+      "step": 6192
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2794578951295337,
+      "learning_rate": 0.00016863140839345908,
+      "loss": 1.0945,
+      "step": 6193
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30665149624277827,
+      "learning_rate": 0.00016861990193555292,
+      "loss": 1.0357,
+      "step": 6194
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27440411091642547,
+      "learning_rate": 0.00016860839376039053,
+      "loss": 1.2011,
+      "step": 6195
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2817725535852179,
+      "learning_rate": 0.00016859688386826,
+      "loss": 1.0482,
+      "step": 6196
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3011982191792621,
+      "learning_rate": 0.0001685853722594493,
+      "loss": 1.0853,
+      "step": 6197
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29959655201744034,
+      "learning_rate": 0.00016857385893424658,
+      "loss": 1.0111,
+      "step": 6198
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2802735304530325,
+      "learning_rate": 0.00016856234389293995,
+      "loss": 1.1118,
+      "step": 6199
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30399590914216984,
+      "learning_rate": 0.00016855082713581758,
+      "loss": 1.0576,
+      "step": 6200
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28089634625081406,
+      "learning_rate": 0.00016853930866316766,
+      "loss": 1.1154,
+      "step": 6201
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33064903413133756,
+      "learning_rate": 0.00016852778847527847,
+      "loss": 1.0004,
+      "step": 6202
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28719188673845863,
+      "learning_rate": 0.00016851626657243834,
+      "loss": 1.0599,
+      "step": 6203
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2562638224677485,
+      "learning_rate": 0.0001685047429549355,
+      "loss": 1.0726,
+      "step": 6204
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.27769483083836394,
+      "learning_rate": 0.00016849321762305846,
+      "loss": 1.0452,
+      "step": 6205
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.311376243880295,
+      "learning_rate": 0.00016848169057709562,
+      "loss": 1.1139,
+      "step": 6206
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.33186169126683296,
+      "learning_rate": 0.00016847016181733543,
+      "loss": 1.1708,
+      "step": 6207
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31561407579512624,
+      "learning_rate": 0.0001684586313440664,
+      "loss": 1.0678,
+      "step": 6208
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2907024924078061,
+      "learning_rate": 0.00016844709915757707,
+      "loss": 1.0885,
+      "step": 6209
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.29061404968369564,
+      "learning_rate": 0.0001684355652581561,
+      "loss": 1.0829,
+      "step": 6210
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2379708775530534,
+      "learning_rate": 0.00016842402964609209,
+      "loss": 1.0688,
+      "step": 6211
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.278600564728756,
+      "learning_rate": 0.00016841249232167372,
+      "loss": 0.975,
+      "step": 6212
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2995894305662367,
+      "learning_rate": 0.00016840095328518975,
+      "loss": 1.0657,
+      "step": 6213
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.28752497877593164,
+      "learning_rate": 0.00016838941253692895,
+      "loss": 1.1441,
+      "step": 6214
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.24322263909794603,
+      "learning_rate": 0.00016837787007718008,
+      "loss": 0.9605,
+      "step": 6215
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2706035311294752,
+      "learning_rate": 0.00016836632590623208,
+      "loss": 1.0975,
+      "step": 6216
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32174695455051877,
+      "learning_rate": 0.0001683547800243738,
+      "loss": 1.0976,
+      "step": 6217
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2886586498798772,
+      "learning_rate": 0.00016834323243189415,
+      "loss": 1.1534,
+      "step": 6218
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.2855976937236902,
+      "learning_rate": 0.00016833168312908222,
+      "loss": 1.1865,
+      "step": 6219
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2667954790754751,
+      "learning_rate": 0.00016832013211622694,
+      "loss": 1.2281,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24933114672059709,
+      "learning_rate": 0.00016830857939361738,
+      "loss": 0.9391,
+      "step": 6221
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27072321221050405,
+      "learning_rate": 0.0001682970249615427,
+      "loss": 1.0534,
+      "step": 6222
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28347062120739175,
+      "learning_rate": 0.0001682854688202921,
+      "loss": 1.0556,
+      "step": 6223
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31128711878834986,
+      "learning_rate": 0.00016827391097015473,
+      "loss": 1.0196,
+      "step": 6224
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29466538688649235,
+      "learning_rate": 0.00016826235141141976,
+      "loss": 1.1365,
+      "step": 6225
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25144131922524754,
+      "learning_rate": 0.0001682507901443766,
+      "loss": 1.01,
+      "step": 6226
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2613185058399168,
+      "learning_rate": 0.00016823922716931451,
+      "loss": 1.0745,
+      "step": 6227
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27346829355051,
+      "learning_rate": 0.00016822766248652288,
+      "loss": 1.0676,
+      "step": 6228
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28753894574718736,
+      "learning_rate": 0.0001682160960962911,
+      "loss": 1.1616,
+      "step": 6229
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27774064585558034,
+      "learning_rate": 0.00016820452799890865,
+      "loss": 1.0945,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29844620149160067,
+      "learning_rate": 0.00016819295819466503,
+      "loss": 1.1999,
+      "step": 6231
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.23278252939346147,
+      "learning_rate": 0.00016818138668384976,
+      "loss": 1.1572,
+      "step": 6232
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2910036401832354,
+      "learning_rate": 0.00016816981346675242,
+      "loss": 1.0418,
+      "step": 6233
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2679309567861235,
+      "learning_rate": 0.00016815823854366267,
+      "loss": 1.1591,
+      "step": 6234
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28959484717417816,
+      "learning_rate": 0.0001681466619148702,
+      "loss": 1.0724,
+      "step": 6235
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24969815430510078,
+      "learning_rate": 0.00016813508358066466,
+      "loss": 1.0834,
+      "step": 6236
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2817944152182526,
+      "learning_rate": 0.00016812350354133583,
+      "loss": 1.0752,
+      "step": 6237
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3095644924541878,
+      "learning_rate": 0.00016811192179717353,
+      "loss": 1.0517,
+      "step": 6238
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2826064816728548,
+      "learning_rate": 0.00016810033834846754,
+      "loss": 0.9918,
+      "step": 6239
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2670350115993338,
+      "learning_rate": 0.0001680887531955078,
+      "loss": 1.0787,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2786656526566983,
+      "learning_rate": 0.00016807716633858425,
+      "loss": 1.1872,
+      "step": 6241
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24633827043340034,
+      "learning_rate": 0.00016806557777798676,
+      "loss": 1.1754,
+      "step": 6242
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2510571689731006,
+      "learning_rate": 0.00016805398751400548,
+      "loss": 1.0483,
+      "step": 6243
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25726341692210536,
+      "learning_rate": 0.00016804239554693036,
+      "loss": 1.0626,
+      "step": 6244
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2973987113917484,
+      "learning_rate": 0.00016803080187705152,
+      "loss": 1.1194,
+      "step": 6245
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24985679408993092,
+      "learning_rate": 0.00016801920650465912,
+      "loss": 0.9716,
+      "step": 6246
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2373094805089608,
+      "learning_rate": 0.00016800760943004334,
+      "loss": 0.9422,
+      "step": 6247
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30043718466849206,
+      "learning_rate": 0.0001679960106534944,
+      "loss": 1.0484,
+      "step": 6248
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.286122704125855,
+      "learning_rate": 0.0001679844101753025,
+      "loss": 1.1008,
+      "step": 6249
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.262959271095953,
+      "learning_rate": 0.00016797280799575804,
+      "loss": 1.1192,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2715165644061917,
+      "learning_rate": 0.00016796120411515138,
+      "loss": 1.0921,
+      "step": 6251
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28638836988275007,
+      "learning_rate": 0.00016794959853377284,
+      "loss": 1.1773,
+      "step": 6252
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2776266911958128,
+      "learning_rate": 0.00016793799125191288,
+      "loss": 1.0659,
+      "step": 6253
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28766018953162986,
+      "learning_rate": 0.00016792638226986202,
+      "loss": 1.0234,
+      "step": 6254
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24147276451994776,
+      "learning_rate": 0.00016791477158791077,
+      "loss": 1.0404,
+      "step": 6255
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28431529367278485,
+      "learning_rate": 0.00016790315920634964,
+      "loss": 1.1169,
+      "step": 6256
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.297855580080018,
+      "learning_rate": 0.00016789154512546927,
+      "loss": 1.1081,
+      "step": 6257
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.26389657019849716,
+      "learning_rate": 0.00016787992934556032,
+      "loss": 1.0314,
+      "step": 6258
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2928163841306302,
+      "learning_rate": 0.0001678683118669135,
+      "loss": 1.2123,
+      "step": 6259
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2975129278676374,
+      "learning_rate": 0.00016785669268981949,
+      "loss": 1.1446,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2853505641305865,
+      "learning_rate": 0.00016784507181456912,
+      "loss": 0.9765,
+      "step": 6261
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29085919322220416,
+      "learning_rate": 0.0001678334492414532,
+      "loss": 1.0566,
+      "step": 6262
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2951008224142174,
+      "learning_rate": 0.00016782182497076257,
+      "loss": 1.0795,
+      "step": 6263
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2866385453022363,
+      "learning_rate": 0.00016781019900278813,
+      "loss": 1.043,
+      "step": 6264
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28440677949059723,
+      "learning_rate": 0.00016779857133782087,
+      "loss": 1.0659,
+      "step": 6265
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2513177187771401,
+      "learning_rate": 0.0001677869419761517,
+      "loss": 1.0256,
+      "step": 6266
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28595770700712264,
+      "learning_rate": 0.00016777531091807175,
+      "loss": 1.1266,
+      "step": 6267
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2514182299644808,
+      "learning_rate": 0.00016776367816387206,
+      "loss": 1.118,
+      "step": 6268
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2766744455914722,
+      "learning_rate": 0.00016775204371384373,
+      "loss": 1.0138,
+      "step": 6269
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30206625962707667,
+      "learning_rate": 0.0001677404075682779,
+      "loss": 1.194,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2631889376384602,
+      "learning_rate": 0.0001677287697274658,
+      "loss": 0.9856,
+      "step": 6271
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25406693794632235,
+      "learning_rate": 0.0001677171301916987,
+      "loss": 1.0063,
+      "step": 6272
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28950694169287894,
+      "learning_rate": 0.00016770548896126783,
+      "loss": 1.0609,
+      "step": 6273
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2761692174626653,
+      "learning_rate": 0.00016769384603646455,
+      "loss": 1.0876,
+      "step": 6274
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.27628317931169183,
+      "learning_rate": 0.00016768220141758023,
+      "loss": 1.0243,
+      "step": 6275
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2660866756756718,
+      "learning_rate": 0.0001676705551049063,
+      "loss": 1.0615,
+      "step": 6276
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24048261615675012,
+      "learning_rate": 0.0001676589070987342,
+      "loss": 1.152,
+      "step": 6277
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30207736888758835,
+      "learning_rate": 0.0001676472573993554,
+      "loss": 1.1271,
+      "step": 6278
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28419649601292135,
+      "learning_rate": 0.0001676356060070615,
+      "loss": 1.0366,
+      "step": 6279
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25939328583829085,
+      "learning_rate": 0.00016762395292214404,
+      "loss": 1.1239,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24732667648488688,
+      "learning_rate": 0.00016761229814489466,
+      "loss": 1.0374,
+      "step": 6281
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28033363094981284,
+      "learning_rate": 0.00016760064167560502,
+      "loss": 1.1033,
+      "step": 6282
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2837131506670996,
+      "learning_rate": 0.00016758898351456683,
+      "loss": 1.1314,
+      "step": 6283
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3341961682280646,
+      "learning_rate": 0.00016757732366207186,
+      "loss": 1.1564,
+      "step": 6284
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2789710186742297,
+      "learning_rate": 0.00016756566211841188,
+      "loss": 1.0238,
+      "step": 6285
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3018369450049017,
+      "learning_rate": 0.00016755399888387874,
+      "loss": 1.0962,
+      "step": 6286
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.265390590494802,
+      "learning_rate": 0.00016754233395876439,
+      "loss": 1.0181,
+      "step": 6287
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2849229079192597,
+      "learning_rate": 0.0001675306673433606,
+      "loss": 1.0362,
+      "step": 6288
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.279551424881973,
+      "learning_rate": 0.00016751899903795947,
+      "loss": 1.0863,
+      "step": 6289
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30798802423123717,
+      "learning_rate": 0.00016750732904285292,
+      "loss": 1.0984,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28386059483142134,
+      "learning_rate": 0.00016749565735833306,
+      "loss": 1.0513,
+      "step": 6291
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2971213923065991,
+      "learning_rate": 0.0001674839839846919,
+      "loss": 1.0989,
+      "step": 6292
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.279028774197727,
+      "learning_rate": 0.0001674723089222217,
+      "loss": 0.9628,
+      "step": 6293
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3005236299374139,
+      "learning_rate": 0.00016746063217121452,
+      "loss": 0.9983,
+      "step": 6294
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2556549845006676,
+      "learning_rate": 0.00016744895373196265,
+      "loss": 1.0838,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2754893169400004,
+      "learning_rate": 0.00016743727360475833,
+      "loss": 1.0528,
+      "step": 6296
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2525035036290481,
+      "learning_rate": 0.00016742559178989383,
+      "loss": 1.1438,
+      "step": 6297
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2994082116871912,
+      "learning_rate": 0.00016741390828766152,
+      "loss": 1.0641,
+      "step": 6298
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2915966725527121,
+      "learning_rate": 0.00016740222309835382,
+      "loss": 1.1256,
+      "step": 6299
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2976305302047498,
+      "learning_rate": 0.00016739053622226305,
+      "loss": 1.0256,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30003197401374204,
+      "learning_rate": 0.00016737884765968185,
+      "loss": 1.1071,
+      "step": 6301
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25781935696696723,
+      "learning_rate": 0.0001673671574109026,
+      "loss": 1.0172,
+      "step": 6302
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2854845428390189,
+      "learning_rate": 0.00016735546547621787,
+      "loss": 1.1589,
+      "step": 6303
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30166239723579663,
+      "learning_rate": 0.00016734377185592032,
+      "loss": 1.1384,
+      "step": 6304
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28301402595792946,
+      "learning_rate": 0.00016733207655030254,
+      "loss": 1.0882,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25169386672382876,
+      "learning_rate": 0.00016732037955965724,
+      "loss": 1.1254,
+      "step": 6306
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28028422765627997,
+      "learning_rate": 0.00016730868088427712,
+      "loss": 1.0061,
+      "step": 6307
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29537665905438676,
+      "learning_rate": 0.00016729698052445497,
+      "loss": 1.0539,
+      "step": 6308
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28211755903503166,
+      "learning_rate": 0.0001672852784804836,
+      "loss": 1.0818,
+      "step": 6309
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2685003161595143,
+      "learning_rate": 0.00016727357475265582,
+      "loss": 0.9727,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.271237791328004,
+      "learning_rate": 0.00016726186934126457,
+      "loss": 1.0693,
+      "step": 6311
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29234966306787097,
+      "learning_rate": 0.00016725016224660274,
+      "loss": 1.0985,
+      "step": 6312
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2415128764029434,
+      "learning_rate": 0.00016723845346896336,
+      "loss": 1.0949,
+      "step": 6313
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28487373231440377,
+      "learning_rate": 0.00016722674300863942,
+      "loss": 1.0031,
+      "step": 6314
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.24380319070580195,
+      "learning_rate": 0.00016721503086592398,
+      "loss": 0.9791,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2854973044582919,
+      "learning_rate": 0.00016720331704111015,
+      "loss": 1.0771,
+      "step": 6316
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2926702434021319,
+      "learning_rate": 0.0001671916015344911,
+      "loss": 1.0796,
+      "step": 6317
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2625594361259241,
+      "learning_rate": 0.00016717988434636,
+      "loss": 1.0848,
+      "step": 6318
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.272123940034111,
+      "learning_rate": 0.00016716816547701003,
+      "loss": 1.0751,
+      "step": 6319
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.25026070785323234,
+      "learning_rate": 0.00016715644492673452,
+      "loss": 1.0925,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.28111670498694746,
+      "learning_rate": 0.00016714472269582678,
+      "loss": 1.1395,
+      "step": 6321
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.30210515943962146,
+      "learning_rate": 0.00016713299878458012,
+      "loss": 1.117,
+      "step": 6322
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2717616248041951,
+      "learning_rate": 0.00016712127319328803,
+      "loss": 1.0637,
+      "step": 6323
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24801900424048062,
+      "learning_rate": 0.00016710954592224386,
+      "loss": 1.1057,
+      "step": 6324
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2898349287109403,
+      "learning_rate": 0.00016709781697174113,
+      "loss": 1.0904,
+      "step": 6325
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2615926237179176,
+      "learning_rate": 0.00016708608634207338,
+      "loss": 1.013,
+      "step": 6326
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25404922506608696,
+      "learning_rate": 0.00016707435403353412,
+      "loss": 1.0322,
+      "step": 6327
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2993573106392043,
+      "learning_rate": 0.000167062620046417,
+      "loss": 1.1574,
+      "step": 6328
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27922342474327827,
+      "learning_rate": 0.0001670508843810157,
+      "loss": 1.0711,
+      "step": 6329
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3011854686091883,
+      "learning_rate": 0.00016703914703762387,
+      "loss": 1.054,
+      "step": 6330
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25394776998572083,
+      "learning_rate": 0.00016702740801653523,
+      "loss": 1.0363,
+      "step": 6331
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32790248402799294,
+      "learning_rate": 0.00016701566731804358,
+      "loss": 1.1621,
+      "step": 6332
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2825346749177906,
+      "learning_rate": 0.00016700392494244277,
+      "loss": 0.9924,
+      "step": 6333
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2615222834120971,
+      "learning_rate": 0.0001669921808900266,
+      "loss": 1.1129,
+      "step": 6334
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2685877115229659,
+      "learning_rate": 0.000166980435161089,
+      "loss": 1.0624,
+      "step": 6335
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2782305195102996,
+      "learning_rate": 0.00016696868775592394,
+      "loss": 1.1125,
+      "step": 6336
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26429721749726304,
+      "learning_rate": 0.00016695693867482535,
+      "loss": 1.0898,
+      "step": 6337
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3006719307055734,
+      "learning_rate": 0.0001669451879180873,
+      "loss": 1.1202,
+      "step": 6338
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2546442290098055,
+      "learning_rate": 0.00016693343548600386,
+      "loss": 1.0205,
+      "step": 6339
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3163145075653169,
+      "learning_rate": 0.00016692168137886912,
+      "loss": 1.0082,
+      "step": 6340
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26767591471900337,
+      "learning_rate": 0.00016690992559697726,
+      "loss": 1.0872,
+      "step": 6341
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3092574347023085,
+      "learning_rate": 0.00016689816814062245,
+      "loss": 1.0457,
+      "step": 6342
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2637784366965473,
+      "learning_rate": 0.00016688640901009894,
+      "loss": 1.0477,
+      "step": 6343
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30076256747451274,
+      "learning_rate": 0.000166874648205701,
+      "loss": 1.1079,
+      "step": 6344
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2923545840530909,
+      "learning_rate": 0.00016686288572772295,
+      "loss": 1.2222,
+      "step": 6345
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3153084026273102,
+      "learning_rate": 0.00016685112157645916,
+      "loss": 0.9666,
+      "step": 6346
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2963713357046184,
+      "learning_rate": 0.00016683935575220407,
+      "loss": 1.0839,
+      "step": 6347
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2737775027067931,
+      "learning_rate": 0.00016682758825525208,
+      "loss": 1.1275,
+      "step": 6348
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2800942528032241,
+      "learning_rate": 0.00016681581908589772,
+      "loss": 0.9692,
+      "step": 6349
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3109943800000079,
+      "learning_rate": 0.00016680404824443546,
+      "loss": 1.0421,
+      "step": 6350
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27462352106623983,
+      "learning_rate": 0.0001667922757311599,
+      "loss": 1.1845,
+      "step": 6351
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32025726456616344,
+      "learning_rate": 0.0001667805015463657,
+      "loss": 1.0136,
+      "step": 6352
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.301660874521527,
+      "learning_rate": 0.00016676872569034744,
+      "loss": 1.1865,
+      "step": 6353
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29842518248408173,
+      "learning_rate": 0.00016675694816339987,
+      "loss": 1.2503,
+      "step": 6354
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26028217660743824,
+      "learning_rate": 0.00016674516896581773,
+      "loss": 0.9257,
+      "step": 6355
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.290277459931021,
+      "learning_rate": 0.00016673338809789577,
+      "loss": 1.0985,
+      "step": 6356
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27210815585379544,
+      "learning_rate": 0.00016672160555992885,
+      "loss": 1.1111,
+      "step": 6357
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27064221588352777,
+      "learning_rate": 0.0001667098213522118,
+      "loss": 0.9628,
+      "step": 6358
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31149953692608723,
+      "learning_rate": 0.00016669803547503958,
+      "loss": 1.0404,
+      "step": 6359
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2859751349658579,
+      "learning_rate": 0.0001666862479287071,
+      "loss": 0.9679,
+      "step": 6360
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29734676588988057,
+      "learning_rate": 0.0001666744587135093,
+      "loss": 1.1466,
+      "step": 6361
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2695100397906395,
+      "learning_rate": 0.00016666266782974133,
+      "loss": 1.0106,
+      "step": 6362
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.33059150738359744,
+      "learning_rate": 0.00016665087527769815,
+      "loss": 1.1697,
+      "step": 6363
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2783320081837106,
+      "learning_rate": 0.00016663908105767495,
+      "loss": 1.0822,
+      "step": 6364
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31565170424334077,
+      "learning_rate": 0.00016662728516996688,
+      "loss": 1.1806,
+      "step": 6365
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2709155225707643,
+      "learning_rate": 0.0001666154876148691,
+      "loss": 1.0718,
+      "step": 6366
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2743956028640576,
+      "learning_rate": 0.00016660368839267693,
+      "loss": 1.1101,
+      "step": 6367
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3083906678441462,
+      "learning_rate": 0.00016659188750368554,
+      "loss": 1.1813,
+      "step": 6368
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27005425077077816,
+      "learning_rate": 0.00016658008494819032,
+      "loss": 1.01,
+      "step": 6369
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3190770015107477,
+      "learning_rate": 0.00016656828072648665,
+      "loss": 1.0918,
+      "step": 6370
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2786920758144284,
+      "learning_rate": 0.00016655647483886988,
+      "loss": 0.9536,
+      "step": 6371
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28799801927513924,
+      "learning_rate": 0.00016654466728563557,
+      "loss": 1.1653,
+      "step": 6372
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2740543474851488,
+      "learning_rate": 0.00016653285806707908,
+      "loss": 1.0493,
+      "step": 6373
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.283012370411115,
+      "learning_rate": 0.000166521047183496,
+      "loss": 1.0719,
+      "step": 6374
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2866968650229277,
+      "learning_rate": 0.00016650923463518196,
+      "loss": 0.9907,
+      "step": 6375
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2802944367538119,
+      "learning_rate": 0.00016649742042243248,
+      "loss": 1.0706,
+      "step": 6376
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27154248970374756,
+      "learning_rate": 0.00016648560454554328,
+      "loss": 1.1223,
+      "step": 6377
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2838775500172997,
+      "learning_rate": 0.00016647378700481005,
+      "loss": 1.0581,
+      "step": 6378
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2920529073936615,
+      "learning_rate": 0.00016646196780052848,
+      "loss": 1.0481,
+      "step": 6379
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2670726827629129,
+      "learning_rate": 0.00016645014693299442,
+      "loss": 1.0974,
+      "step": 6380
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24730959097979163,
+      "learning_rate": 0.00016643832440250367,
+      "loss": 1.1209,
+      "step": 6381
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2947114263680111,
+      "learning_rate": 0.00016642650020935214,
+      "loss": 0.9936,
+      "step": 6382
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3233060078932617,
+      "learning_rate": 0.00016641467435383564,
+      "loss": 1.1597,
+      "step": 6383
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26002181675498415,
+      "learning_rate": 0.00016640284683625017,
+      "loss": 0.9699,
+      "step": 6384
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26273440181552443,
+      "learning_rate": 0.0001663910176568918,
+      "loss": 1.1475,
+      "step": 6385
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.24674523967677028,
+      "learning_rate": 0.00016637918681605639,
+      "loss": 1.0134,
+      "step": 6386
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2618701917902359,
+      "learning_rate": 0.00016636735431404019,
+      "loss": 1.1473,
+      "step": 6387
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28468586915692684,
+      "learning_rate": 0.00016635552015113918,
+      "loss": 1.0526,
+      "step": 6388
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3028992395965799,
+      "learning_rate": 0.0001663436843276496,
+      "loss": 1.0536,
+      "step": 6389
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2789358949175681,
+      "learning_rate": 0.00016633184684386763,
+      "loss": 1.1303,
+      "step": 6390
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29126579487824283,
+      "learning_rate": 0.00016632000770008947,
+      "loss": 1.0763,
+      "step": 6391
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27542318323495574,
+      "learning_rate": 0.0001663081668966115,
+      "loss": 1.1802,
+      "step": 6392
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30182241732147397,
+      "learning_rate": 0.00016629632443372993,
+      "loss": 1.054,
+      "step": 6393
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2819317111614903,
+      "learning_rate": 0.0001662844803117412,
+      "loss": 1.0247,
+      "step": 6394
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2813661907175226,
+      "learning_rate": 0.00016627263453094168,
+      "loss": 1.153,
+      "step": 6395
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3156855165232155,
+      "learning_rate": 0.00016626078709162782,
+      "loss": 1.1216,
+      "step": 6396
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2762036811217373,
+      "learning_rate": 0.00016624893799409613,
+      "loss": 1.0578,
+      "step": 6397
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25813174874980066,
+      "learning_rate": 0.00016623708723864314,
+      "loss": 1.0028,
+      "step": 6398
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2547111721851036,
+      "learning_rate": 0.0001662252348255654,
+      "loss": 1.0322,
+      "step": 6399
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2670666010121307,
+      "learning_rate": 0.00016621338075515954,
+      "loss": 1.0951,
+      "step": 6400
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25941920255451006,
+      "learning_rate": 0.00016620152502772224,
+      "loss": 1.009,
+      "step": 6401
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32953249149853675,
+      "learning_rate": 0.00016618966764355016,
+      "loss": 1.1824,
+      "step": 6402
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3400097911397109,
+      "learning_rate": 0.00016617780860294002,
+      "loss": 1.0785,
+      "step": 6403
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2606374235637558,
+      "learning_rate": 0.00016616594790618865,
+      "loss": 1.0066,
+      "step": 6404
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.28689829208173906,
+      "learning_rate": 0.00016615408555359284,
+      "loss": 1.1033,
+      "step": 6405
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.275071242493895,
+      "learning_rate": 0.00016614222154544948,
+      "loss": 1.0145,
+      "step": 6406
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3036143858185991,
+      "learning_rate": 0.00016613035588205542,
+      "loss": 0.976,
+      "step": 6407
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2739250339275323,
+      "learning_rate": 0.00016611848856370768,
+      "loss": 0.9842,
+      "step": 6408
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.269718907859172,
+      "learning_rate": 0.0001661066195907032,
+      "loss": 1.0573,
+      "step": 6409
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27815112294798416,
+      "learning_rate": 0.000166094748963339,
+      "loss": 1.1291,
+      "step": 6410
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2810699578192083,
+      "learning_rate": 0.0001660828766819122,
+      "loss": 0.9918,
+      "step": 6411
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.25506946682445086,
+      "learning_rate": 0.00016607100274671982,
+      "loss": 1.1449,
+      "step": 6412
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2890524163750954,
+      "learning_rate": 0.00016605912715805915,
+      "loss": 1.124,
+      "step": 6413
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29021337271256664,
+      "learning_rate": 0.00016604724991622726,
+      "loss": 1.0194,
+      "step": 6414
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27539899144845825,
+      "learning_rate": 0.00016603537102152145,
+      "loss": 1.077,
+      "step": 6415
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29031381020521185,
+      "learning_rate": 0.00016602349047423895,
+      "loss": 1.068,
+      "step": 6416
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.259955978828572,
+      "learning_rate": 0.00016601160827467713,
+      "loss": 0.9664,
+      "step": 6417
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29036829398239994,
+      "learning_rate": 0.00016599972442313333,
+      "loss": 1.0306,
+      "step": 6418
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2771649514160956,
+      "learning_rate": 0.00016598783891990496,
+      "loss": 1.0726,
+      "step": 6419
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2597924047332126,
+      "learning_rate": 0.00016597595176528942,
+      "loss": 1.0015,
+      "step": 6420
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.291533841923736,
+      "learning_rate": 0.00016596406295958421,
+      "loss": 1.0385,
+      "step": 6421
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.27196336627994283,
+      "learning_rate": 0.0001659521725030869,
+      "loss": 1.0915,
+      "step": 6422
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3091516771574349,
+      "learning_rate": 0.00016594028039609504,
+      "loss": 1.1498,
+      "step": 6423
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.269385742381281,
+      "learning_rate": 0.00016592838663890617,
+      "loss": 1.1685,
+      "step": 6424
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31387068153422903,
+      "learning_rate": 0.00016591649123181803,
+      "loss": 1.0578,
+      "step": 6425
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2715541694948404,
+      "learning_rate": 0.00016590459417512824,
+      "loss": 1.0411,
+      "step": 6426
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.31668645833456616,
+      "learning_rate": 0.00016589269546913457,
+      "loss": 1.045,
+      "step": 6427
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.26537674929552535,
+      "learning_rate": 0.0001658807951141348,
+      "loss": 0.9741,
+      "step": 6428
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30617980217717927,
+      "learning_rate": 0.00016586889311042674,
+      "loss": 1.0323,
+      "step": 6429
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2893643055407271,
+      "learning_rate": 0.00016585698945830818,
+      "loss": 0.9841,
+      "step": 6430
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26428036894894746,
+      "learning_rate": 0.00016584508415807712,
+      "loss": 1.0233,
+      "step": 6431
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25987613015099514,
+      "learning_rate": 0.00016583317721003142,
+      "loss": 0.9113,
+      "step": 6432
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2676269631291612,
+      "learning_rate": 0.0001658212686144691,
+      "loss": 1.0322,
+      "step": 6433
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27792570272364814,
+      "learning_rate": 0.00016580935837168817,
+      "loss": 1.1305,
+      "step": 6434
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3054138054002042,
+      "learning_rate": 0.00016579744648198666,
+      "loss": 0.9909,
+      "step": 6435
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28755696638880873,
+      "learning_rate": 0.0001657855329456627,
+      "loss": 1.1987,
+      "step": 6436
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2937260358743029,
+      "learning_rate": 0.0001657736177630145,
+      "loss": 1.175,
+      "step": 6437
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25543914091501174,
+      "learning_rate": 0.00016576170093434008,
+      "loss": 0.9054,
+      "step": 6438
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2358502828421239,
+      "learning_rate": 0.00016574978245993783,
+      "loss": 1.1184,
+      "step": 6439
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.32159479122906715,
+      "learning_rate": 0.00016573786234010593,
+      "loss": 1.0697,
+      "step": 6440
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29597420637835464,
+      "learning_rate": 0.0001657259405751427,
+      "loss": 1.0147,
+      "step": 6441
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27502750185112845,
+      "learning_rate": 0.0001657140171653465,
+      "loss": 1.1519,
+      "step": 6442
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28880378692907466,
+      "learning_rate": 0.00016570209211101578,
+      "loss": 1.0251,
+      "step": 6443
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29201502579906263,
+      "learning_rate": 0.00016569016541244884,
+      "loss": 1.1337,
+      "step": 6444
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2399540511362018,
+      "learning_rate": 0.00016567823706994426,
+      "loss": 1.0998,
+      "step": 6445
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3268134405014831,
+      "learning_rate": 0.00016566630708380052,
+      "loss": 0.9888,
+      "step": 6446
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3103957574422285,
+      "learning_rate": 0.00016565437545431618,
+      "loss": 1.1474,
+      "step": 6447
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2718259809863342,
+      "learning_rate": 0.0001656424421817898,
+      "loss": 0.9704,
+      "step": 6448
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27353989936463624,
+      "learning_rate": 0.00016563050726652007,
+      "loss": 1.0992,
+      "step": 6449
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2856249195760054,
+      "learning_rate": 0.00016561857070880565,
+      "loss": 1.1137,
+      "step": 6450
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.273725036502166,
+      "learning_rate": 0.00016560663250894526,
+      "loss": 0.9277,
+      "step": 6451
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2763856735936746,
+      "learning_rate": 0.00016559469266723767,
+      "loss": 1.0204,
+      "step": 6452
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28085668794591295,
+      "learning_rate": 0.00016558275118398164,
+      "loss": 0.9921,
+      "step": 6453
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2735774703202355,
+      "learning_rate": 0.00016557080805947605,
+      "loss": 0.9692,
+      "step": 6454
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26616410789129574,
+      "learning_rate": 0.0001655588632940198,
+      "loss": 0.9541,
+      "step": 6455
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2890311486765243,
+      "learning_rate": 0.0001655469168879118,
+      "loss": 1.0578,
+      "step": 6456
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29609104767351785,
+      "learning_rate": 0.00016553496884145097,
+      "loss": 1.0883,
+      "step": 6457
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2608637836134015,
+      "learning_rate": 0.0001655230191549364,
+      "loss": 1.0304,
+      "step": 6458
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27959926336448077,
+      "learning_rate": 0.00016551106782866705,
+      "loss": 1.0061,
+      "step": 6459
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2771253997843709,
+      "learning_rate": 0.0001654991148629421,
+      "loss": 1.1194,
+      "step": 6460
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28449814950864316,
+      "learning_rate": 0.00016548716025806062,
+      "loss": 1.088,
+      "step": 6461
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2691654812600947,
+      "learning_rate": 0.0001654752040143218,
+      "loss": 1.0968,
+      "step": 6462
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31954023097003986,
+      "learning_rate": 0.00016546324613202483,
+      "loss": 0.9794,
+      "step": 6463
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2649303795932421,
+      "learning_rate": 0.000165451286611469,
+      "loss": 1.0693,
+      "step": 6464
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2630512812762983,
+      "learning_rate": 0.0001654393254529536,
+      "loss": 1.0332,
+      "step": 6465
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2938590102565347,
+      "learning_rate": 0.00016542736265677795,
+      "loss": 1.1377,
+      "step": 6466
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29423857954744015,
+      "learning_rate": 0.0001654153982232414,
+      "loss": 1.0742,
+      "step": 6467
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2770519185899091,
+      "learning_rate": 0.00016540343215264342,
+      "loss": 0.9763,
+      "step": 6468
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27149015096394086,
+      "learning_rate": 0.00016539146444528345,
+      "loss": 1.2458,
+      "step": 6469
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2723835734113916,
+      "learning_rate": 0.00016537949510146097,
+      "loss": 1.0588,
+      "step": 6470
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2860060696336331,
+      "learning_rate": 0.00016536752412147555,
+      "loss": 0.9471,
+      "step": 6471
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.256259821327159,
+      "learning_rate": 0.0001653555515056268,
+      "loss": 1.0662,
+      "step": 6472
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26590917435909617,
+      "learning_rate": 0.00016534357725421422,
+      "loss": 0.9938,
+      "step": 6473
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30036253247255307,
+      "learning_rate": 0.0001653316013675376,
+      "loss": 0.9861,
+      "step": 6474
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24447569581162235,
+      "learning_rate": 0.00016531962384589655,
+      "loss": 1.0813,
+      "step": 6475
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31129076026873814,
+      "learning_rate": 0.0001653076446895909,
+      "loss": 1.0898,
+      "step": 6476
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2755133042122456,
+      "learning_rate": 0.00016529566389892039,
+      "loss": 1.0924,
+      "step": 6477
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24784520601298288,
+      "learning_rate": 0.00016528368147418485,
+      "loss": 1.2024,
+      "step": 6478
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2869637235166629,
+      "learning_rate": 0.00016527169741568416,
+      "loss": 1.0971,
+      "step": 6479
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.25871213333736603,
+      "learning_rate": 0.00016525971172371822,
+      "loss": 0.993,
+      "step": 6480
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2604087934206153,
+      "learning_rate": 0.00016524772439858694,
+      "loss": 1.1067,
+      "step": 6481
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2816319137305283,
+      "learning_rate": 0.0001652357354405904,
+      "loss": 1.0467,
+      "step": 6482
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27702751178720125,
+      "learning_rate": 0.0001652237448500286,
+      "loss": 1.0449,
+      "step": 6483
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27403004840211326,
+      "learning_rate": 0.00016521175262720154,
+      "loss": 1.1382,
+      "step": 6484
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27012660723215115,
+      "learning_rate": 0.00016519975877240942,
+      "loss": 1.1069,
+      "step": 6485
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.262756289576385,
+      "learning_rate": 0.00016518776328595234,
+      "loss": 1.0257,
+      "step": 6486
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2898435765096172,
+      "learning_rate": 0.0001651757661681305,
+      "loss": 1.2361,
+      "step": 6487
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2778059074062683,
+      "learning_rate": 0.0001651637674192442,
+      "loss": 1.116,
+      "step": 6488
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29747157960661386,
+      "learning_rate": 0.00016515176703959364,
+      "loss": 1.0449,
+      "step": 6489
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2758337700493432,
+      "learning_rate": 0.00016513976502947913,
+      "loss": 0.9708,
+      "step": 6490
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29558292743334763,
+      "learning_rate": 0.00016512776138920108,
+      "loss": 1.0372,
+      "step": 6491
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26433384453275566,
+      "learning_rate": 0.0001651157561190599,
+      "loss": 1.1413,
+      "step": 6492
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3164936697696587,
+      "learning_rate": 0.00016510374921935598,
+      "loss": 1.1305,
+      "step": 6493
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2680448222970725,
+      "learning_rate": 0.00016509174069038985,
+      "loss": 1.2003,
+      "step": 6494
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26701918502526756,
+      "learning_rate": 0.00016507973053246197,
+      "loss": 0.9809,
+      "step": 6495
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28698630137463055,
+      "learning_rate": 0.00016506771874587296,
+      "loss": 1.1306,
+      "step": 6496
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2861081221048606,
+      "learning_rate": 0.00016505570533092333,
+      "loss": 1.2112,
+      "step": 6497
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27462328764365,
+      "learning_rate": 0.00016504369028791382,
+      "loss": 1.0667,
+      "step": 6498
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2542616461897349,
+      "learning_rate": 0.0001650316736171451,
+      "loss": 1.1647,
+      "step": 6499
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.268648077124173,
+      "learning_rate": 0.00016501965531891786,
+      "loss": 1.023,
+      "step": 6500
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30932903309172555,
+      "learning_rate": 0.0001650076353935329,
+      "loss": 1.0771,
+      "step": 6501
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2711465953776904,
+      "learning_rate": 0.000164995613841291,
+      "loss": 1.023,
+      "step": 6502
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2488690322666522,
+      "learning_rate": 0.000164983590662493,
+      "loss": 1.1685,
+      "step": 6503
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2447593866202204,
+      "learning_rate": 0.00016497156585743982,
+      "loss": 1.0553,
+      "step": 6504
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28996537913588744,
+      "learning_rate": 0.00016495953942643237,
+      "loss": 1.0453,
+      "step": 6505
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3100091579089353,
+      "learning_rate": 0.00016494751136977165,
+      "loss": 1.0363,
+      "step": 6506
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26438432513334514,
+      "learning_rate": 0.0001649354816877586,
+      "loss": 1.0239,
+      "step": 6507
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3136841961736258,
+      "learning_rate": 0.0001649234503806943,
+      "loss": 1.1726,
+      "step": 6508
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2793775958515165,
+      "learning_rate": 0.0001649114174488799,
+      "loss": 1.0953,
+      "step": 6509
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2991111217534248,
+      "learning_rate": 0.0001648993828926164,
+      "loss": 0.9635,
+      "step": 6510
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33123211260741653,
+      "learning_rate": 0.00016488734671220512,
+      "loss": 1.09,
+      "step": 6511
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3071684112575216,
+      "learning_rate": 0.0001648753089079472,
+      "loss": 1.112,
+      "step": 6512
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2919753257761195,
+      "learning_rate": 0.0001648632694801439,
+      "loss": 1.2142,
+      "step": 6513
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3047462341213951,
+      "learning_rate": 0.00016485122842909653,
+      "loss": 1.0747,
+      "step": 6514
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2561697275636995,
+      "learning_rate": 0.00016483918575510638,
+      "loss": 1.0719,
+      "step": 6515
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29441141071756,
+      "learning_rate": 0.00016482714145847488,
+      "loss": 1.1003,
+      "step": 6516
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2507495981732362,
+      "learning_rate": 0.0001648150955395034,
+      "loss": 1.0107,
+      "step": 6517
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2745896765591778,
+      "learning_rate": 0.00016480304799849343,
+      "loss": 1.0221,
+      "step": 6518
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2956074131649605,
+      "learning_rate": 0.00016479099883574648,
+      "loss": 1.0595,
+      "step": 6519
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2755124978473152,
+      "learning_rate": 0.00016477894805156404,
+      "loss": 0.9841,
+      "step": 6520
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2653332774638079,
+      "learning_rate": 0.00016476689564624773,
+      "loss": 1.1523,
+      "step": 6521
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28336035642857826,
+      "learning_rate": 0.00016475484162009913,
+      "loss": 1.0458,
+      "step": 6522
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.27133324752026045,
+      "learning_rate": 0.00016474278597341995,
+      "loss": 1.0068,
+      "step": 6523
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.254789334192467,
+      "learning_rate": 0.00016473072870651183,
+      "loss": 1.0911,
+      "step": 6524
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2645683646629639,
+      "learning_rate": 0.00016471866981967654,
+      "loss": 1.0735,
+      "step": 6525
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.29514733454194775,
+      "learning_rate": 0.0001647066093132159,
+      "loss": 1.1258,
+      "step": 6526
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30773726775104837,
+      "learning_rate": 0.00016469454718743166,
+      "loss": 1.2085,
+      "step": 6527
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28753719268726,
+      "learning_rate": 0.00016468248344262575,
+      "loss": 1.1034,
+      "step": 6528
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.28935088461393205,
+      "learning_rate": 0.00016467041807910002,
+      "loss": 0.9435,
+      "step": 6529
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2910181115331406,
+      "learning_rate": 0.00016465835109715643,
+      "loss": 0.8991,
+      "step": 6530
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.26222425373411046,
+      "learning_rate": 0.00016464628249709699,
+      "loss": 1.1253,
+      "step": 6531
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.24656232690052832,
+      "learning_rate": 0.00016463421227922367,
+      "loss": 0.979,
+      "step": 6532
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2997042750535666,
+      "learning_rate": 0.0001646221404438386,
+      "loss": 1.0947,
+      "step": 6533
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26285239590686826,
+      "learning_rate": 0.0001646100669912438,
+      "loss": 0.9987,
+      "step": 6534
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2810386117270015,
+      "learning_rate": 0.00016459799192174152,
+      "loss": 1.08,
+      "step": 6535
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2899490777892802,
+      "learning_rate": 0.0001645859152356339,
+      "loss": 1.0636,
+      "step": 6536
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32728289531293414,
+      "learning_rate": 0.00016457383693322314,
+      "loss": 1.1029,
+      "step": 6537
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2873444747610267,
+      "learning_rate": 0.0001645617570148115,
+      "loss": 1.1222,
+      "step": 6538
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26266152625659334,
+      "learning_rate": 0.00016454967548070135,
+      "loss": 1.1339,
+      "step": 6539
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2803672145152364,
+      "learning_rate": 0.00016453759233119503,
+      "loss": 1.094,
+      "step": 6540
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27084575822983264,
+      "learning_rate": 0.00016452550756659482,
+      "loss": 0.937,
+      "step": 6541
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2701911111109992,
+      "learning_rate": 0.00016451342118720328,
+      "loss": 1.029,
+      "step": 6542
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2716711794064477,
+      "learning_rate": 0.00016450133319332282,
+      "loss": 0.8668,
+      "step": 6543
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2794432093629184,
+      "learning_rate": 0.00016448924358525595,
+      "loss": 1.0109,
+      "step": 6544
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28134829606945316,
+      "learning_rate": 0.00016447715236330524,
+      "loss": 1.191,
+      "step": 6545
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.22932974730919156,
+      "learning_rate": 0.0001644650595277733,
+      "loss": 1.0067,
+      "step": 6546
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2627137036762324,
+      "learning_rate": 0.00016445296507896267,
+      "loss": 1.0458,
+      "step": 6547
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31351571153357294,
+      "learning_rate": 0.00016444086901717614,
+      "loss": 1.0699,
+      "step": 6548
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3113263872681726,
+      "learning_rate": 0.00016442877134271633,
+      "loss": 1.1383,
+      "step": 6549
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24271597078080498,
+      "learning_rate": 0.00016441667205588603,
+      "loss": 0.9184,
+      "step": 6550
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27790009137065697,
+      "learning_rate": 0.00016440457115698802,
+      "loss": 1.2613,
+      "step": 6551
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27765509036577246,
+      "learning_rate": 0.0001643924686463252,
+      "loss": 1.1003,
+      "step": 6552
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29484287527906927,
+      "learning_rate": 0.00016438036452420032,
+      "loss": 1.0934,
+      "step": 6553
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29018553943805164,
+      "learning_rate": 0.00016436825879091635,
+      "loss": 1.0676,
+      "step": 6554
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28130026896984306,
+      "learning_rate": 0.00016435615144677629,
+      "loss": 1.1587,
+      "step": 6555
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2878221364125203,
+      "learning_rate": 0.00016434404249208306,
+      "loss": 1.0381,
+      "step": 6556
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2777652669244412,
+      "learning_rate": 0.00016433193192713974,
+      "loss": 1.0429,
+      "step": 6557
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26453569833758483,
+      "learning_rate": 0.00016431981975224938,
+      "loss": 1.1111,
+      "step": 6558
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28622435501200705,
+      "learning_rate": 0.00016430770596771512,
+      "loss": 1.0154,
+      "step": 6559
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2847988417285998,
+      "learning_rate": 0.00016429559057384011,
+      "loss": 0.9996,
+      "step": 6560
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26384397601020537,
+      "learning_rate": 0.00016428347357092755,
+      "loss": 0.9437,
+      "step": 6561
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2775557302168176,
+      "learning_rate": 0.00016427135495928062,
+      "loss": 1.1182,
+      "step": 6562
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30234120956824934,
+      "learning_rate": 0.00016425923473920267,
+      "loss": 1.0162,
+      "step": 6563
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2667444994608358,
+      "learning_rate": 0.000164247112910997,
+      "loss": 0.9534,
+      "step": 6564
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.295387882239055,
+      "learning_rate": 0.0001642349894749669,
+      "loss": 1.1292,
+      "step": 6565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25364552624174114,
+      "learning_rate": 0.00016422286443141585,
+      "loss": 1.0281,
+      "step": 6566
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30870163058659145,
+      "learning_rate": 0.00016421073778064726,
+      "loss": 0.9925,
+      "step": 6567
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31418016446029906,
+      "learning_rate": 0.0001641986095229646,
+      "loss": 1.0183,
+      "step": 6568
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30568890062391635,
+      "learning_rate": 0.0001641864796586714,
+      "loss": 1.1813,
+      "step": 6569
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28403262150412845,
+      "learning_rate": 0.00016417434818807118,
+      "loss": 1.1134,
+      "step": 6570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26397330303772965,
+      "learning_rate": 0.00016416221511146757,
+      "loss": 1.113,
+      "step": 6571
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2863559947419068,
+      "learning_rate": 0.0001641500804291642,
+      "loss": 1.1201,
+      "step": 6572
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24406593687365424,
+      "learning_rate": 0.00016413794414146476,
+      "loss": 1.0542,
+      "step": 6573
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27139228105504454,
+      "learning_rate": 0.00016412580624867299,
+      "loss": 1.012,
+      "step": 6574
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27171095999953015,
+      "learning_rate": 0.00016411366675109256,
+      "loss": 0.9942,
+      "step": 6575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2692825756718052,
+      "learning_rate": 0.00016410152564902734,
+      "loss": 1.0619,
+      "step": 6576
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2694897408553471,
+      "learning_rate": 0.00016408938294278118,
+      "loss": 1.1153,
+      "step": 6577
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2946068069084391,
+      "learning_rate": 0.0001640772386326579,
+      "loss": 0.9959,
+      "step": 6578
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2732041052214766,
+      "learning_rate": 0.0001640650927189615,
+      "loss": 1.1325,
+      "step": 6579
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25828674961643106,
+      "learning_rate": 0.00016405294520199586,
+      "loss": 1.0311,
+      "step": 6580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2528935519865577,
+      "learning_rate": 0.000164040796082065,
+      "loss": 1.0198,
+      "step": 6581
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24788820972140796,
+      "learning_rate": 0.00016402864535947298,
+      "loss": 1.0777,
+      "step": 6582
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27027895661109375,
+      "learning_rate": 0.00016401649303452386,
+      "loss": 0.9911,
+      "step": 6583
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29776619082303163,
+      "learning_rate": 0.0001640043391075218,
+      "loss": 0.995,
+      "step": 6584
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25164405759102015,
+      "learning_rate": 0.0001639921835787709,
+      "loss": 1.0057,
+      "step": 6585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28442402581138315,
+      "learning_rate": 0.00016398002644857538,
+      "loss": 1.0983,
+      "step": 6586
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27597236003640674,
+      "learning_rate": 0.00016396786771723953,
+      "loss": 1.0816,
+      "step": 6587
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3036495741618543,
+      "learning_rate": 0.00016395570738506754,
+      "loss": 1.0313,
+      "step": 6588
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2983086346865117,
+      "learning_rate": 0.0001639435454523638,
+      "loss": 1.1382,
+      "step": 6589
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.24415345911957764,
+      "learning_rate": 0.00016393138191943266,
+      "loss": 0.9764,
+      "step": 6590
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26401568236315054,
+      "learning_rate": 0.0001639192167865785,
+      "loss": 1.0753,
+      "step": 6591
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30616078316140916,
+      "learning_rate": 0.00016390705005410577,
+      "loss": 1.1654,
+      "step": 6592
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.268584602967286,
+      "learning_rate": 0.00016389488172231895,
+      "loss": 1.1281,
+      "step": 6593
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3042044038945393,
+      "learning_rate": 0.00016388271179152255,
+      "loss": 1.1451,
+      "step": 6594
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30865017852012816,
+      "learning_rate": 0.00016387054026202114,
+      "loss": 1.0497,
+      "step": 6595
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2981587654841934,
+      "learning_rate": 0.00016385836713411932,
+      "loss": 1.0117,
+      "step": 6596
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30634610666151335,
+      "learning_rate": 0.00016384619240812173,
+      "loss": 1.0624,
+      "step": 6597
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31330552508402865,
+      "learning_rate": 0.00016383401608433305,
+      "loss": 1.0021,
+      "step": 6598
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2746478862766659,
+      "learning_rate": 0.00016382183816305798,
+      "loss": 1.092,
+      "step": 6599
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.31666878049663144,
+      "learning_rate": 0.00016380965864460135,
+      "loss": 1.1224,
+      "step": 6600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2805806328462776,
+      "learning_rate": 0.00016379747752926787,
+      "loss": 1.0326,
+      "step": 6601
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27859538132326334,
+      "learning_rate": 0.00016378529481736242,
+      "loss": 1.0901,
+      "step": 6602
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27971190411521635,
+      "learning_rate": 0.00016377311050918989,
+      "loss": 1.0349,
+      "step": 6603
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.35928083318694953,
+      "learning_rate": 0.0001637609246050552,
+      "loss": 1.0325,
+      "step": 6604
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2791809789319585,
+      "learning_rate": 0.00016374873710526327,
+      "loss": 1.101,
+      "step": 6605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.25414723944354833,
+      "learning_rate": 0.00016373654801011913,
+      "loss": 0.9845,
+      "step": 6606
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29501863717394433,
+      "learning_rate": 0.00016372435731992784,
+      "loss": 1.0297,
+      "step": 6607
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27739025744682017,
+      "learning_rate": 0.00016371216503499443,
+      "loss": 1.1649,
+      "step": 6608
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27992773939480853,
+      "learning_rate": 0.0001636999711556241,
+      "loss": 1.1623,
+      "step": 6609
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.256055609384181,
+      "learning_rate": 0.00016368777568212192,
+      "loss": 1.0349,
+      "step": 6610
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27352283442582775,
+      "learning_rate": 0.00016367557861479316,
+      "loss": 1.081,
+      "step": 6611
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2878898287968566,
+      "learning_rate": 0.00016366337995394296,
+      "loss": 1.0915,
+      "step": 6612
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29458815084779255,
+      "learning_rate": 0.0001636511796998767,
+      "loss": 1.0093,
+      "step": 6613
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2833302474424508,
+      "learning_rate": 0.0001636389778528997,
+      "loss": 1.1375,
+      "step": 6614
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2827351352279611,
+      "learning_rate": 0.00016362677441331727,
+      "loss": 1.133,
+      "step": 6615
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2776267400325992,
+      "learning_rate": 0.0001636145693814348,
+      "loss": 1.082,
+      "step": 6616
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29210767304951707,
+      "learning_rate": 0.00016360236275755777,
+      "loss": 1.0961,
+      "step": 6617
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3111465892344082,
+      "learning_rate": 0.00016359015454199161,
+      "loss": 1.0517,
+      "step": 6618
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30230280360893014,
+      "learning_rate": 0.0001635779447350419,
+      "loss": 1.0257,
+      "step": 6619
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2640103618422073,
+      "learning_rate": 0.00016356573333701414,
+      "loss": 0.9745,
+      "step": 6620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.27301899027846943,
+      "learning_rate": 0.00016355352034821396,
+      "loss": 0.974,
+      "step": 6621
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2738570363795273,
+      "learning_rate": 0.00016354130576894698,
+      "loss": 0.995,
+      "step": 6622
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29413026629702177,
+      "learning_rate": 0.00016352908959951892,
+      "loss": 1.16,
+      "step": 6623
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30276135434523793,
+      "learning_rate": 0.00016351687184023547,
+      "loss": 1.0801,
+      "step": 6624
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.26212091458025155,
+      "learning_rate": 0.00016350465249140235,
+      "loss": 1.1354,
+      "step": 6625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.295200448114549,
+      "learning_rate": 0.0001634924315533254,
+      "loss": 1.1294,
+      "step": 6626
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2719331912386029,
+      "learning_rate": 0.00016348020902631047,
+      "loss": 1.0,
+      "step": 6627
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2801548078187952,
+      "learning_rate": 0.0001634679849106634,
+      "loss": 1.0472,
+      "step": 6628
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28890264685867756,
+      "learning_rate": 0.0001634557592066901,
+      "loss": 1.159,
+      "step": 6629
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2972148378840613,
+      "learning_rate": 0.00016344353191469657,
+      "loss": 1.0649,
+      "step": 6630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30972438758134824,
+      "learning_rate": 0.00016343130303498877,
+      "loss": 1.1313,
+      "step": 6631
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2787430037836057,
+      "learning_rate": 0.00016341907256787273,
+      "loss": 1.0461,
+      "step": 6632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.2509205560032734,
+      "learning_rate": 0.00016340684051365458,
+      "loss": 1.1465,
+      "step": 6633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3371469012042158,
+      "learning_rate": 0.00016339460687264039,
+      "loss": 0.9917,
+      "step": 6634
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28031780574967835,
+      "learning_rate": 0.0001633823716451363,
+      "loss": 1.0859,
+      "step": 6635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.32014706377626473,
+      "learning_rate": 0.00016337013483144853,
+      "loss": 1.0125,
+      "step": 6636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.28048230557063836,
+      "learning_rate": 0.00016335789643188333,
+      "loss": 1.119,
+      "step": 6637
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27484347493922334,
+      "learning_rate": 0.0001633456564467469,
+      "loss": 1.001,
+      "step": 6638
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26589912392685283,
+      "learning_rate": 0.00016333341487634567,
+      "loss": 1.2356,
+      "step": 6639
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3261860350977622,
+      "learning_rate": 0.0001633211717209859,
+      "loss": 1.1768,
+      "step": 6640
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30705986319442513,
+      "learning_rate": 0.000163308926980974,
+      "loss": 1.072,
+      "step": 6641
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3098762264842706,
+      "learning_rate": 0.00016329668065661644,
+      "loss": 1.088,
+      "step": 6642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24610152808075564,
+      "learning_rate": 0.00016328443274821964,
+      "loss": 1.0679,
+      "step": 6643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.293387945611742,
+      "learning_rate": 0.00016327218325609018,
+      "loss": 1.0764,
+      "step": 6644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24880390773011782,
+      "learning_rate": 0.0001632599321805345,
+      "loss": 1.1687,
+      "step": 6645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2871610507138743,
+      "learning_rate": 0.00016324767952185932,
+      "loss": 1.0554,
+      "step": 6646
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33306363001968703,
+      "learning_rate": 0.00016323542528037116,
+      "loss": 1.066,
+      "step": 6647
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2601315042428321,
+      "learning_rate": 0.0001632231694563768,
+      "loss": 1.062,
+      "step": 6648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3122750901555906,
+      "learning_rate": 0.00016321091205018283,
+      "loss": 0.988,
+      "step": 6649
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2530794057553161,
+      "learning_rate": 0.0001631986530620961,
+      "loss": 0.9536,
+      "step": 6650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25248623075991256,
+      "learning_rate": 0.00016318639249242336,
+      "loss": 1.0528,
+      "step": 6651
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25241303157165856,
+      "learning_rate": 0.00016317413034147143,
+      "loss": 0.9887,
+      "step": 6652
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25151161516493253,
+      "learning_rate": 0.00016316186660954716,
+      "loss": 1.0605,
+      "step": 6653
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2537064242578609,
+      "learning_rate": 0.0001631496012969575,
+      "loss": 1.0014,
+      "step": 6654
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26974830206999884,
+      "learning_rate": 0.00016313733440400941,
+      "loss": 1.1359,
+      "step": 6655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28423801348421174,
+      "learning_rate": 0.0001631250659310098,
+      "loss": 1.0911,
+      "step": 6656
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3067572434414906,
+      "learning_rate": 0.00016311279587826575,
+      "loss": 1.1511,
+      "step": 6657
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30349651329317273,
+      "learning_rate": 0.00016310052424608435,
+      "loss": 1.0717,
+      "step": 6658
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.31106706237441717,
+      "learning_rate": 0.00016308825103477262,
+      "loss": 1.0138,
+      "step": 6659
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30512529577510444,
+      "learning_rate": 0.0001630759762446378,
+      "loss": 1.1175,
+      "step": 6660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3156400341146579,
+      "learning_rate": 0.00016306369987598705,
+      "loss": 1.0594,
+      "step": 6661
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3024624555405739,
+      "learning_rate": 0.00016305142192912754,
+      "loss": 1.1203,
+      "step": 6662
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24676397866245015,
+      "learning_rate": 0.00016303914240436656,
+      "loss": 1.0936,
+      "step": 6663
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2585771707617533,
+      "learning_rate": 0.00016302686130201144,
+      "loss": 1.1232,
+      "step": 6664
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2726696313152824,
+      "learning_rate": 0.00016301457862236954,
+      "loss": 1.0913,
+      "step": 6665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2760746274104927,
+      "learning_rate": 0.00016300229436574815,
+      "loss": 0.9746,
+      "step": 6666
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2755395263384695,
+      "learning_rate": 0.00016299000853245475,
+      "loss": 1.0183,
+      "step": 6667
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27852069729003387,
+      "learning_rate": 0.00016297772112279683,
+      "loss": 1.0574,
+      "step": 6668
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27979974439621436,
+      "learning_rate": 0.00016296543213708184,
+      "loss": 0.9499,
+      "step": 6669
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2960060531441859,
+      "learning_rate": 0.00016295314157561736,
+      "loss": 1.0549,
+      "step": 6670
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2660000695711491,
+      "learning_rate": 0.00016294084943871092,
+      "loss": 1.1072,
+      "step": 6671
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2564269859230904,
+      "learning_rate": 0.0001629285557266702,
+      "loss": 1.1333,
+      "step": 6672
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27504533401306863,
+      "learning_rate": 0.00016291626043980282,
+      "loss": 1.0034,
+      "step": 6673
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3099324729512832,
+      "learning_rate": 0.00016290396357841646,
+      "loss": 1.0459,
+      "step": 6674
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2844820883790574,
+      "learning_rate": 0.00016289166514281888,
+      "loss": 0.9184,
+      "step": 6675
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27114882440422344,
+      "learning_rate": 0.00016287936513331787,
+      "loss": 1.1727,
+      "step": 6676
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2720012225716546,
+      "learning_rate": 0.00016286706355022118,
+      "loss": 1.0654,
+      "step": 6677
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2916372120440236,
+      "learning_rate": 0.00016285476039383675,
+      "loss": 1.0744,
+      "step": 6678
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27052138249686447,
+      "learning_rate": 0.00016284245566447245,
+      "loss": 1.1014,
+      "step": 6679
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2652236414824706,
+      "learning_rate": 0.0001628301493624362,
+      "loss": 1.1672,
+      "step": 6680
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27950873661043685,
+      "learning_rate": 0.00016281784148803596,
+      "loss": 1.0014,
+      "step": 6681
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2827281504818738,
+      "learning_rate": 0.0001628055320415798,
+      "loss": 0.9676,
+      "step": 6682
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2609854109436044,
+      "learning_rate": 0.00016279322102337565,
+      "loss": 0.8978,
+      "step": 6683
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24726688155868715,
+      "learning_rate": 0.00016278090843373173,
+      "loss": 1.0843,
+      "step": 6684
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3193189601858733,
+      "learning_rate": 0.00016276859427295613,
+      "loss": 1.028,
+      "step": 6685
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2845484123594936,
+      "learning_rate": 0.00016275627854135698,
+      "loss": 0.9256,
+      "step": 6686
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29028230142099054,
+      "learning_rate": 0.00016274396123924252,
+      "loss": 1.0812,
+      "step": 6687
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32773446539660916,
+      "learning_rate": 0.000162731642366921,
+      "loss": 1.0608,
+      "step": 6688
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29790785005575593,
+      "learning_rate": 0.00016271932192470074,
+      "loss": 1.0306,
+      "step": 6689
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27883966783277836,
+      "learning_rate": 0.00016270699991289,
+      "loss": 1.0423,
+      "step": 6690
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28349188815486853,
+      "learning_rate": 0.0001626946763317972,
+      "loss": 1.0897,
+      "step": 6691
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3054195899080486,
+      "learning_rate": 0.00016268235118173068,
+      "loss": 1.0126,
+      "step": 6692
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.297957213586038,
+      "learning_rate": 0.00016267002446299891,
+      "loss": 1.0832,
+      "step": 6693
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2786420902687923,
+      "learning_rate": 0.00016265769617591046,
+      "loss": 1.1095,
+      "step": 6694
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28426733069435445,
+      "learning_rate": 0.00016264536632077376,
+      "loss": 1.0612,
+      "step": 6695
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27433860519495945,
+      "learning_rate": 0.0001626330348978974,
+      "loss": 1.0349,
+      "step": 6696
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.317775496735832,
+      "learning_rate": 0.00016262070190758995,
+      "loss": 1.1712,
+      "step": 6697
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2568267188069798,
+      "learning_rate": 0.00016260836735016012,
+      "loss": 0.9826,
+      "step": 6698
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2852886758776049,
+      "learning_rate": 0.00016259603122591653,
+      "loss": 1.0588,
+      "step": 6699
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28264369508139864,
+      "learning_rate": 0.0001625836935351679,
+      "loss": 1.139,
+      "step": 6700
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3102222867034462,
+      "learning_rate": 0.00016257135427822302,
+      "loss": 1.0697,
+      "step": 6701
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.268465641442644,
+      "learning_rate": 0.00016255901345539072,
+      "loss": 1.0887,
+      "step": 6702
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30026142767107994,
+      "learning_rate": 0.00016254667106697972,
+      "loss": 1.1509,
+      "step": 6703
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2706314630840685,
+      "learning_rate": 0.000162534327113299,
+      "loss": 1.0657,
+      "step": 6704
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26773703311582037,
+      "learning_rate": 0.00016252198159465744,
+      "loss": 1.1287,
+      "step": 6705
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2957605324588231,
+      "learning_rate": 0.000162509634511364,
+      "loss": 1.2115,
+      "step": 6706
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29017697603422365,
+      "learning_rate": 0.00016249728586372765,
+      "loss": 1.0255,
+      "step": 6707
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27496495480231525,
+      "learning_rate": 0.0001624849356520575,
+      "loss": 1.0261,
+      "step": 6708
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26386368126415394,
+      "learning_rate": 0.0001624725838766625,
+      "loss": 1.1855,
+      "step": 6709
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27191856479446785,
+      "learning_rate": 0.00016246023053785184,
+      "loss": 1.033,
+      "step": 6710
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26761649181678265,
+      "learning_rate": 0.0001624478756359347,
+      "loss": 1.137,
+      "step": 6711
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26810717439455456,
+      "learning_rate": 0.00016243551917122017,
+      "loss": 1.169,
+      "step": 6712
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.23716095057095146,
+      "learning_rate": 0.00016242316114401754,
+      "loss": 0.9461,
+      "step": 6713
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2820129032334972,
+      "learning_rate": 0.00016241080155463613,
+      "loss": 1.028,
+      "step": 6714
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29187354158124584,
+      "learning_rate": 0.00016239844040338513,
+      "loss": 1.1075,
+      "step": 6715
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2707882691645488,
+      "learning_rate": 0.00016238607769057396,
+      "loss": 1.1043,
+      "step": 6716
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.27825828988627205,
+      "learning_rate": 0.00016237371341651198,
+      "loss": 0.9317,
+      "step": 6717
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24456292837088042,
+      "learning_rate": 0.00016236134758150863,
+      "loss": 1.115,
+      "step": 6718
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30642644995469454,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 1.0855,
+      "step": 6719
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2536012594797928,
+      "learning_rate": 0.00016233661122991568,
+      "loss": 1.0347,
+      "step": 6720
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28201291743191076,
+      "learning_rate": 0.00016232424071394513,
+      "loss": 0.923,
+      "step": 6721
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26500412097642506,
+      "learning_rate": 0.00016231186863827128,
+      "loss": 1.1197,
+      "step": 6722
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28336108524806514,
+      "learning_rate": 0.00016229949500320376,
+      "loss": 0.9663,
+      "step": 6723
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.30535153183981684,
+      "learning_rate": 0.00016228711980905222,
+      "loss": 1.0546,
+      "step": 6724
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2738980528128915,
+      "learning_rate": 0.00016227474305612635,
+      "loss": 1.1652,
+      "step": 6725
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29696172073189514,
+      "learning_rate": 0.00016226236474473592,
+      "loss": 1.1307,
+      "step": 6726
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2575129900484908,
+      "learning_rate": 0.00016224998487519065,
+      "loss": 1.0553,
+      "step": 6727
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25508951971135796,
+      "learning_rate": 0.0001622376034478004,
+      "loss": 0.9931,
+      "step": 6728
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3026293198793625,
+      "learning_rate": 0.00016222522046287506,
+      "loss": 1.0979,
+      "step": 6729
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2598437183575833,
+      "learning_rate": 0.00016221283592072442,
+      "loss": 1.0128,
+      "step": 6730
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3163844838276975,
+      "learning_rate": 0.00016220044982165845,
+      "loss": 1.1538,
+      "step": 6731
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2362891477357362,
+      "learning_rate": 0.00016218806216598713,
+      "loss": 1.0506,
+      "step": 6732
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.26662591447108,
+      "learning_rate": 0.00016217567295402052,
+      "loss": 0.962,
+      "step": 6733
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29725496528185136,
+      "learning_rate": 0.00016216328218606856,
+      "loss": 1.0977,
+      "step": 6734
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2571941755837014,
+      "learning_rate": 0.00016215088986244145,
+      "loss": 1.0301,
+      "step": 6735
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.24442116258690824,
+      "learning_rate": 0.00016213849598344923,
+      "loss": 1.0068,
+      "step": 6736
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2747700753479199,
+      "learning_rate": 0.0001621261005494021,
+      "loss": 0.984,
+      "step": 6737
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.2947494451070546,
+      "learning_rate": 0.00016211370356061024,
+      "loss": 0.9723,
+      "step": 6738
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.28377559231615324,
+      "learning_rate": 0.00016210130501738393,
+      "loss": 1.1093,
+      "step": 6739
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.281938530095994,
+      "learning_rate": 0.00016208890492003345,
+      "loss": 1.0411,
+      "step": 6740
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.25782869247051643,
+      "learning_rate": 0.00016207650326886908,
+      "loss": 1.0448,
+      "step": 6741
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27863659241136324,
+      "learning_rate": 0.0001620641000642012,
+      "loss": 1.0766,
+      "step": 6742
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3034063935548644,
+      "learning_rate": 0.00016205169530634022,
+      "loss": 1.0668,
+      "step": 6743
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29522301302575077,
+      "learning_rate": 0.00016203928899559655,
+      "loss": 1.0437,
+      "step": 6744
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2407728356207695,
+      "learning_rate": 0.0001620268811322807,
+      "loss": 1.1312,
+      "step": 6745
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26324055091027326,
+      "learning_rate": 0.0001620144717167032,
+      "loss": 1.0093,
+      "step": 6746
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2804316768014678,
+      "learning_rate": 0.0001620020607491745,
+      "loss": 1.0229,
+      "step": 6747
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2753277098551558,
+      "learning_rate": 0.00016198964823000531,
+      "loss": 1.0634,
+      "step": 6748
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2800366285769209,
+      "learning_rate": 0.00016197723415950618,
+      "loss": 1.0365,
+      "step": 6749
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.302146197809392,
+      "learning_rate": 0.00016196481853798783,
+      "loss": 0.9998,
+      "step": 6750
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3568635660133519,
+      "learning_rate": 0.00016195240136576098,
+      "loss": 1.1658,
+      "step": 6751
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2716917931898554,
+      "learning_rate": 0.00016193998264313632,
+      "loss": 1.0812,
+      "step": 6752
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27290621018790484,
+      "learning_rate": 0.0001619275623704247,
+      "loss": 1.094,
+      "step": 6753
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2682620960621536,
+      "learning_rate": 0.00016191514054793687,
+      "loss": 1.0381,
+      "step": 6754
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2845594233710903,
+      "learning_rate": 0.00016190271717598376,
+      "loss": 1.1278,
+      "step": 6755
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27355020118683776,
+      "learning_rate": 0.0001618902922548762,
+      "loss": 0.9993,
+      "step": 6756
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.287166856164538,
+      "learning_rate": 0.00016187786578492527,
+      "loss": 1.1079,
+      "step": 6757
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31367583134610894,
+      "learning_rate": 0.00016186543776644177,
+      "loss": 1.0392,
+      "step": 6758
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26476803361212475,
+      "learning_rate": 0.00016185300819973687,
+      "loss": 1.0339,
+      "step": 6759
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2869935686558382,
+      "learning_rate": 0.00016184057708512156,
+      "loss": 1.0353,
+      "step": 6760
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2607335209317506,
+      "learning_rate": 0.0001618281444229069,
+      "loss": 1.0328,
+      "step": 6761
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2819258447169869,
+      "learning_rate": 0.0001618157102134041,
+      "loss": 1.0694,
+      "step": 6762
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2486602423602287,
+      "learning_rate": 0.0001618032744569243,
+      "loss": 1.0398,
+      "step": 6763
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2964468238045624,
+      "learning_rate": 0.0001617908371537787,
+      "loss": 1.1674,
+      "step": 6764
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28968276628283657,
+      "learning_rate": 0.00016177839830427862,
+      "loss": 1.0923,
+      "step": 6765
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30259410481968824,
+      "learning_rate": 0.00016176595790873526,
+      "loss": 1.1047,
+      "step": 6766
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.299883854383291,
+      "learning_rate": 0.00016175351596745997,
+      "loss": 1.0987,
+      "step": 6767
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28742268599750015,
+      "learning_rate": 0.00016174107248076414,
+      "loss": 1.1337,
+      "step": 6768
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27078457130344585,
+      "learning_rate": 0.00016172862744895917,
+      "loss": 1.1273,
+      "step": 6769
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28329813420365574,
+      "learning_rate": 0.00016171618087235652,
+      "loss": 1.042,
+      "step": 6770
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2835411240856413,
+      "learning_rate": 0.00016170373275126761,
+      "loss": 1.1055,
+      "step": 6771
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2697682905187557,
+      "learning_rate": 0.00016169128308600404,
+      "loss": 1.1592,
+      "step": 6772
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2606901412284456,
+      "learning_rate": 0.00016167883187687737,
+      "loss": 1.0796,
+      "step": 6773
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2718079210220933,
+      "learning_rate": 0.0001616663791241991,
+      "loss": 0.9178,
+      "step": 6774
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2992698573782557,
+      "learning_rate": 0.00016165392482828098,
+      "loss": 0.9155,
+      "step": 6775
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2820273589437464,
+      "learning_rate": 0.00016164146898943463,
+      "loss": 1.1096,
+      "step": 6776
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27964729949513284,
+      "learning_rate": 0.00016162901160797182,
+      "loss": 1.0301,
+      "step": 6777
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30035527749172686,
+      "learning_rate": 0.0001616165526842042,
+      "loss": 1.0112,
+      "step": 6778
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26367756968433076,
+      "learning_rate": 0.0001616040922184437,
+      "loss": 1.062,
+      "step": 6779
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30454178782431157,
+      "learning_rate": 0.000161591630211002,
+      "loss": 1.0908,
+      "step": 6780
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2736709005007848,
+      "learning_rate": 0.0001615791666621911,
+      "loss": 1.0642,
+      "step": 6781
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28955239699459784,
+      "learning_rate": 0.00016156670157232278,
+      "loss": 1.038,
+      "step": 6782
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3055984799372455,
+      "learning_rate": 0.00016155423494170913,
+      "loss": 1.083,
+      "step": 6783
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28914856789930005,
+      "learning_rate": 0.00016154176677066204,
+      "loss": 1.0251,
+      "step": 6784
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2937920207039406,
+      "learning_rate": 0.00016152929705949356,
+      "loss": 1.0395,
+      "step": 6785
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26681380130500015,
+      "learning_rate": 0.00016151682580851576,
+      "loss": 1.1342,
+      "step": 6786
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31513605238395026,
+      "learning_rate": 0.00016150435301804072,
+      "loss": 1.1459,
+      "step": 6787
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2962395662354257,
+      "learning_rate": 0.0001614918786883806,
+      "loss": 0.9866,
+      "step": 6788
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27715689822063977,
+      "learning_rate": 0.00016147940281984754,
+      "loss": 1.0869,
+      "step": 6789
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.261319217896763,
+      "learning_rate": 0.00016146692541275383,
+      "loss": 0.9879,
+      "step": 6790
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29020817288316764,
+      "learning_rate": 0.00016145444646741166,
+      "loss": 1.1018,
+      "step": 6791
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2564090714223762,
+      "learning_rate": 0.00016144196598413336,
+      "loss": 1.0812,
+      "step": 6792
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.252997662859385,
+      "learning_rate": 0.00016142948396323124,
+      "loss": 1.0051,
+      "step": 6793
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27979442491585477,
+      "learning_rate": 0.00016141700040501767,
+      "loss": 1.0166,
+      "step": 6794
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2723199327047408,
+      "learning_rate": 0.00016140451530980503,
+      "loss": 1.0498,
+      "step": 6795
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24997068906233713,
+      "learning_rate": 0.00016139202867790586,
+      "loss": 1.0254,
+      "step": 6796
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25606449185707403,
+      "learning_rate": 0.00016137954050963256,
+      "loss": 1.0375,
+      "step": 6797
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27072931875871203,
+      "learning_rate": 0.0001613670508052977,
+      "loss": 1.1278,
+      "step": 6798
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2861760154233243,
+      "learning_rate": 0.00016135455956521383,
+      "loss": 1.0276,
+      "step": 6799
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.281890088050873,
+      "learning_rate": 0.00016134206678969351,
+      "loss": 1.151,
+      "step": 6800
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2561578106187753,
+      "learning_rate": 0.00016132957247904948,
+      "loss": 1.0579,
+      "step": 6801
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.296476813043644,
+      "learning_rate": 0.0001613170766335943,
+      "loss": 1.0815,
+      "step": 6802
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24047219584398974,
+      "learning_rate": 0.00016130457925364074,
+      "loss": 1.0674,
+      "step": 6803
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28537999657913954,
+      "learning_rate": 0.00016129208033950157,
+      "loss": 1.1703,
+      "step": 6804
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3071076656903495,
+      "learning_rate": 0.00016127957989148958,
+      "loss": 1.1097,
+      "step": 6805
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2912482911905167,
+      "learning_rate": 0.00016126707790991757,
+      "loss": 1.0033,
+      "step": 6806
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2873587122615141,
+      "learning_rate": 0.00016125457439509843,
+      "loss": 1.1402,
+      "step": 6807
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2822651267978888,
+      "learning_rate": 0.00016124206934734509,
+      "loss": 1.0684,
+      "step": 6808
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30569024152993257,
+      "learning_rate": 0.0001612295627669705,
+      "loss": 1.0356,
+      "step": 6809
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2552189424981213,
+      "learning_rate": 0.00016121705465428756,
+      "loss": 1.1015,
+      "step": 6810
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2780062595426105,
+      "learning_rate": 0.0001612045450096094,
+      "loss": 1.0252,
+      "step": 6811
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2958626948496943,
+      "learning_rate": 0.000161192033833249,
+      "loss": 1.047,
+      "step": 6812
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2547715653655437,
+      "learning_rate": 0.0001611795211255195,
+      "loss": 0.9724,
+      "step": 6813
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.26184844172876026,
+      "learning_rate": 0.00016116700688673406,
+      "loss": 0.9989,
+      "step": 6814
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2903838433558316,
+      "learning_rate": 0.0001611544911172058,
+      "loss": 1.0365,
+      "step": 6815
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2787578244449706,
+      "learning_rate": 0.00016114197381724798,
+      "loss": 1.1122,
+      "step": 6816
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2890262361588652,
+      "learning_rate": 0.00016112945498717384,
+      "loss": 0.9743,
+      "step": 6817
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25957088901852887,
+      "learning_rate": 0.00016111693462729666,
+      "loss": 1.0411,
+      "step": 6818
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29141337706844284,
+      "learning_rate": 0.0001611044127379298,
+      "loss": 1.0738,
+      "step": 6819
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.27868532305453414,
+      "learning_rate": 0.00016109188931938658,
+      "loss": 1.0309,
+      "step": 6820
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2865127928357969,
+      "learning_rate": 0.00016107936437198048,
+      "loss": 1.0432,
+      "step": 6821
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28901415788858237,
+      "learning_rate": 0.00016106683789602485,
+      "loss": 1.0913,
+      "step": 6822
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31087668233226007,
+      "learning_rate": 0.00016105430989183324,
+      "loss": 1.0044,
+      "step": 6823
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2907385138727252,
+      "learning_rate": 0.0001610417803597192,
+      "loss": 0.9812,
+      "step": 6824
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2955311073850929,
+      "learning_rate": 0.00016102924929999618,
+      "loss": 0.9718,
+      "step": 6825
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25938433521785725,
+      "learning_rate": 0.00016101671671297786,
+      "loss": 1.0653,
+      "step": 6826
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29835875083842295,
+      "learning_rate": 0.00016100418259897787,
+      "loss": 1.1136,
+      "step": 6827
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28475427911325607,
+      "learning_rate": 0.00016099164695830987,
+      "loss": 1.0817,
+      "step": 6828
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29712133149339015,
+      "learning_rate": 0.00016097910979128756,
+      "loss": 1.2519,
+      "step": 6829
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30942764161840036,
+      "learning_rate": 0.00016096657109822472,
+      "loss": 1.153,
+      "step": 6830
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.24877924990251216,
+      "learning_rate": 0.0001609540308794351,
+      "loss": 1.0787,
+      "step": 6831
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2491230221886881,
+      "learning_rate": 0.00016094148913523254,
+      "loss": 1.0693,
+      "step": 6832
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.33875565455907103,
+      "learning_rate": 0.00016092894586593098,
+      "loss": 1.0437,
+      "step": 6833
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.25683830659317614,
+      "learning_rate": 0.00016091640107184418,
+      "loss": 1.0761,
+      "step": 6834
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28029893510677445,
+      "learning_rate": 0.00016090385475328616,
+      "loss": 1.0066,
+      "step": 6835
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2638298088456665,
+      "learning_rate": 0.00016089130691057096,
+      "loss": 1.0488,
+      "step": 6836
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2732308385667981,
+      "learning_rate": 0.0001608787575440125,
+      "loss": 1.0224,
+      "step": 6837
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2652051822786465,
+      "learning_rate": 0.0001608662066539249,
+      "loss": 1.1503,
+      "step": 6838
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3030626041580578,
+      "learning_rate": 0.00016085365424062218,
+      "loss": 1.0871,
+      "step": 6839
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.30008149751746244,
+      "learning_rate": 0.00016084110030441853,
+      "loss": 1.0668,
+      "step": 6840
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2888340632480941,
+      "learning_rate": 0.00016082854484562813,
+      "loss": 1.0374,
+      "step": 6841
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2862699272392478,
+      "learning_rate": 0.00016081598786456516,
+      "loss": 1.1416,
+      "step": 6842
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2717722410816119,
+      "learning_rate": 0.00016080342936154388,
+      "loss": 1.0422,
+      "step": 6843
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2726242149316908,
+      "learning_rate": 0.00016079086933687854,
+      "loss": 1.0457,
+      "step": 6844
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2531509697020992,
+      "learning_rate": 0.0001607783077908835,
+      "loss": 1.0804,
+      "step": 6845
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2574041172829116,
+      "learning_rate": 0.0001607657447238731,
+      "loss": 1.0583,
+      "step": 6846
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28649411980651573,
+      "learning_rate": 0.00016075318013616174,
+      "loss": 1.0341,
+      "step": 6847
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32154960312770714,
+      "learning_rate": 0.0001607406140280639,
+      "loss": 1.0304,
+      "step": 6848
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27521017806337644,
+      "learning_rate": 0.000160728046399894,
+      "loss": 1.0798,
+      "step": 6849
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27096865352974614,
+      "learning_rate": 0.00016071547725196657,
+      "loss": 1.1019,
+      "step": 6850
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2827906410490978,
+      "learning_rate": 0.0001607029065845962,
+      "loss": 1.0481,
+      "step": 6851
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28755989114991476,
+      "learning_rate": 0.00016069033439809738,
+      "loss": 1.221,
+      "step": 6852
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2849156830204197,
+      "learning_rate": 0.00016067776069278485,
+      "loss": 1.0805,
+      "step": 6853
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2738501723030775,
+      "learning_rate": 0.0001606651854689732,
+      "loss": 1.0744,
+      "step": 6854
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30408775836991736,
+      "learning_rate": 0.00016065260872697717,
+      "loss": 1.0593,
+      "step": 6855
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2742033230168456,
+      "learning_rate": 0.00016064003046711148,
+      "loss": 1.1384,
+      "step": 6856
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2623388811156543,
+      "learning_rate": 0.00016062745068969088,
+      "loss": 1.0076,
+      "step": 6857
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2830978029989236,
+      "learning_rate": 0.00016061486939503028,
+      "loss": 1.0956,
+      "step": 6858
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2825549299962127,
+      "learning_rate": 0.00016060228658344445,
+      "loss": 0.9938,
+      "step": 6859
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2787473530770543,
+      "learning_rate": 0.00016058970225524833,
+      "loss": 1.108,
+      "step": 6860
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2766509594000378,
+      "learning_rate": 0.00016057711641075684,
+      "loss": 1.0104,
+      "step": 6861
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2835424799524376,
+      "learning_rate": 0.00016056452905028492,
+      "loss": 1.0538,
+      "step": 6862
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2718902303042608,
+      "learning_rate": 0.0001605519401741476,
+      "loss": 1.1156,
+      "step": 6863
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31396026069305516,
+      "learning_rate": 0.0001605393497826599,
+      "loss": 1.0302,
+      "step": 6864
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3105439907019263,
+      "learning_rate": 0.00016052675787613696,
+      "loss": 1.0684,
+      "step": 6865
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2728613957688195,
+      "learning_rate": 0.00016051416445489385,
+      "loss": 1.0522,
+      "step": 6866
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26407320682849367,
+      "learning_rate": 0.00016050156951924574,
+      "loss": 0.9828,
+      "step": 6867
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.286389653868074,
+      "learning_rate": 0.00016048897306950784,
+      "loss": 1.0011,
+      "step": 6868
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.20462171644432642,
+      "learning_rate": 0.00016047637510599534,
+      "loss": 0.9521,
+      "step": 6869
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28123560210233106,
+      "learning_rate": 0.00016046377562902356,
+      "loss": 0.9788,
+      "step": 6870
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30001191282665973,
+      "learning_rate": 0.0001604511746389078,
+      "loss": 1.0776,
+      "step": 6871
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24622530091185105,
+      "learning_rate": 0.00016043857213596344,
+      "loss": 1.0529,
+      "step": 6872
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27492099325556013,
+      "learning_rate": 0.00016042596812050576,
+      "loss": 1.0378,
+      "step": 6873
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2918899294834224,
+      "learning_rate": 0.00016041336259285031,
+      "loss": 1.0596,
+      "step": 6874
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2623677600554193,
+      "learning_rate": 0.00016040075555331246,
+      "loss": 1.1632,
+      "step": 6875
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26087552461867863,
+      "learning_rate": 0.00016038814700220777,
+      "loss": 1.0231,
+      "step": 6876
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27308513799182704,
+      "learning_rate": 0.00016037553693985172,
+      "loss": 1.0093,
+      "step": 6877
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3112434726283997,
+      "learning_rate": 0.00016036292536655993,
+      "loss": 1.1635,
+      "step": 6878
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2884405944631099,
+      "learning_rate": 0.00016035031228264798,
+      "loss": 1.1504,
+      "step": 6879
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25052783602254847,
+      "learning_rate": 0.00016033769768843153,
+      "loss": 1.0535,
+      "step": 6880
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2952430681334669,
+      "learning_rate": 0.00016032508158422633,
+      "loss": 1.1563,
+      "step": 6881
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2775844498309543,
+      "learning_rate": 0.00016031246397034797,
+      "loss": 1.063,
+      "step": 6882
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2641737504846404,
+      "learning_rate": 0.00016029984484711233,
+      "loss": 0.9881,
+      "step": 6883
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.25954131255382207,
+      "learning_rate": 0.0001602872242148352,
+      "loss": 0.9942,
+      "step": 6884
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.235184004978054,
+      "learning_rate": 0.00016027460207383238,
+      "loss": 1.0592,
+      "step": 6885
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3208319103526141,
+      "learning_rate": 0.00016026197842441975,
+      "loss": 0.9071,
+      "step": 6886
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.38310981570001407,
+      "learning_rate": 0.00016024935326691323,
+      "loss": 1.0084,
+      "step": 6887
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2736848488804497,
+      "learning_rate": 0.00016023672660162881,
+      "loss": 1.1093,
+      "step": 6888
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30255083039667124,
+      "learning_rate": 0.00016022409842888244,
+      "loss": 1.0453,
+      "step": 6889
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29616770728129344,
+      "learning_rate": 0.00016021146874899015,
+      "loss": 1.1753,
+      "step": 6890
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26891944823488517,
+      "learning_rate": 0.000160198837562268,
+      "loss": 0.998,
+      "step": 6891
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3069851887194386,
+      "learning_rate": 0.00016018620486903213,
+      "loss": 0.9517,
+      "step": 6892
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2878707069041881,
+      "learning_rate": 0.00016017357066959863,
+      "loss": 1.0137,
+      "step": 6893
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2804636505710241,
+      "learning_rate": 0.0001601609349642837,
+      "loss": 1.0739,
+      "step": 6894
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.272015497269986,
+      "learning_rate": 0.00016014829775340362,
+      "loss": 1.1176,
+      "step": 6895
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24507472885852935,
+      "learning_rate": 0.00016013565903727454,
+      "loss": 1.1288,
+      "step": 6896
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30186275659636463,
+      "learning_rate": 0.00016012301881621283,
+      "loss": 1.0384,
+      "step": 6897
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3163308512937789,
+      "learning_rate": 0.00016011037709053478,
+      "loss": 1.0186,
+      "step": 6898
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30810804171123124,
+      "learning_rate": 0.00016009773386055676,
+      "loss": 1.1093,
+      "step": 6899
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2816814074951005,
+      "learning_rate": 0.00016008508912659518,
+      "loss": 1.0513,
+      "step": 6900
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30201609837419807,
+      "learning_rate": 0.00016007244288896645,
+      "loss": 0.9661,
+      "step": 6901
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28829302388205935,
+      "learning_rate": 0.00016005979514798713,
+      "loss": 1.0704,
+      "step": 6902
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26515326393964955,
+      "learning_rate": 0.00016004714590397366,
+      "loss": 1.1436,
+      "step": 6903
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2971122785850026,
+      "learning_rate": 0.00016003449515724263,
+      "loss": 1.1088,
+      "step": 6904
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3097599416745419,
+      "learning_rate": 0.00016002184290811065,
+      "loss": 1.0972,
+      "step": 6905
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2522652174175764,
+      "learning_rate": 0.00016000918915689432,
+      "loss": 1.0747,
+      "step": 6906
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3182119441827633,
+      "learning_rate": 0.0001599965339039103,
+      "loss": 1.027,
+      "step": 6907
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2783524829429487,
+      "learning_rate": 0.0001599838771494753,
+      "loss": 1.0525,
+      "step": 6908
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26729480846780146,
+      "learning_rate": 0.0001599712188939061,
+      "loss": 1.02,
+      "step": 6909
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27005772910227627,
+      "learning_rate": 0.00015995855913751946,
+      "loss": 1.1131,
+      "step": 6910
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.23819686657027173,
+      "learning_rate": 0.00015994589788063222,
+      "loss": 1.0546,
+      "step": 6911
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2937203478865386,
+      "learning_rate": 0.00015993323512356118,
+      "loss": 1.1432,
+      "step": 6912
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2788834474966195,
+      "learning_rate": 0.00015992057086662323,
+      "loss": 1.2207,
+      "step": 6913
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26680578605709754,
+      "learning_rate": 0.0001599079051101354,
+      "loss": 1.0555,
+      "step": 6914
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2829489557203681,
+      "learning_rate": 0.00015989523785441456,
+      "loss": 0.9449,
+      "step": 6915
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29363875973797865,
+      "learning_rate": 0.00015988256909977777,
+      "loss": 0.9878,
+      "step": 6916
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2543061502185059,
+      "learning_rate": 0.00015986989884654202,
+      "loss": 1.1064,
+      "step": 6917
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26512820084084004,
+      "learning_rate": 0.00015985722709502444,
+      "loss": 1.036,
+      "step": 6918
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28652859367622907,
+      "learning_rate": 0.00015984455384554215,
+      "loss": 1.1668,
+      "step": 6919
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31723374235235535,
+      "learning_rate": 0.00015983187909841226,
+      "loss": 1.0463,
+      "step": 6920
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29129953499102984,
+      "learning_rate": 0.00015981920285395202,
+      "loss": 1.058,
+      "step": 6921
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29993610209681537,
+      "learning_rate": 0.0001598065251124786,
+      "loss": 1.04,
+      "step": 6922
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2947422459230847,
+      "learning_rate": 0.00015979384587430935,
+      "loss": 1.0852,
+      "step": 6923
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2714231587486264,
+      "learning_rate": 0.00015978116513976152,
+      "loss": 0.9586,
+      "step": 6924
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.31441338998821106,
+      "learning_rate": 0.00015976848290915244,
+      "loss": 1.0261,
+      "step": 6925
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3020176096145915,
+      "learning_rate": 0.0001597557991827995,
+      "loss": 1.0744,
+      "step": 6926
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2941393454482125,
+      "learning_rate": 0.00015974311396102015,
+      "loss": 1.0624,
+      "step": 6927
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29719508910446507,
+      "learning_rate": 0.00015973042724413183,
+      "loss": 1.18,
+      "step": 6928
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.28110919244257476,
+      "learning_rate": 0.00015971773903245202,
+      "loss": 0.9661,
+      "step": 6929
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2803124705765725,
+      "learning_rate": 0.00015970504932629823,
+      "loss": 1.0396,
+      "step": 6930
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29208189180807126,
+      "learning_rate": 0.0001596923581259881,
+      "loss": 1.1863,
+      "step": 6931
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29689595292251414,
+      "learning_rate": 0.0001596796654318392,
+      "loss": 1.0811,
+      "step": 6932
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3077207175340659,
+      "learning_rate": 0.00015966697124416914,
+      "loss": 1.0538,
+      "step": 6933
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30686176921995806,
+      "learning_rate": 0.0001596542755632956,
+      "loss": 1.0504,
+      "step": 6934
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24712629310974943,
+      "learning_rate": 0.00015964157838953638,
+      "loss": 1.0084,
+      "step": 6935
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26002780535214853,
+      "learning_rate": 0.00015962887972320914,
+      "loss": 1.0972,
+      "step": 6936
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29963814827197655,
+      "learning_rate": 0.00015961617956463173,
+      "loss": 1.1617,
+      "step": 6937
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29022102343364803,
+      "learning_rate": 0.00015960347791412196,
+      "loss": 0.9665,
+      "step": 6938
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30836910286066316,
+      "learning_rate": 0.00015959077477199765,
+      "loss": 1.0885,
+      "step": 6939
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2836035210572582,
+      "learning_rate": 0.0001595780701385768,
+      "loss": 1.0539,
+      "step": 6940
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.30435356184547685,
+      "learning_rate": 0.0001595653640141773,
+      "loss": 0.976,
+      "step": 6941
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.26960506796384626,
+      "learning_rate": 0.00015955265639911711,
+      "loss": 1.1245,
+      "step": 6942
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36290217128036273,
+      "learning_rate": 0.00015953994729371427,
+      "loss": 1.0396,
+      "step": 6943
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32401610966187167,
+      "learning_rate": 0.00015952723669828683,
+      "loss": 1.0519,
+      "step": 6944
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.27984623740696507,
+      "learning_rate": 0.00015951452461315292,
+      "loss": 1.0498,
+      "step": 6945
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.24996113180352433,
+      "learning_rate": 0.00015950181103863056,
+      "loss": 1.0642,
+      "step": 6946
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2587685668722977,
+      "learning_rate": 0.000159489095975038,
+      "loss": 1.1062,
+      "step": 6947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32763762154630816,
+      "learning_rate": 0.00015947637942269343,
+      "loss": 1.1287,
+      "step": 6948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2585215704752277,
+      "learning_rate": 0.0001594636613819151,
+      "loss": 1.0892,
+      "step": 6949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2885043831452831,
+      "learning_rate": 0.0001594509418530213,
+      "loss": 1.1353,
+      "step": 6950
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26515145631234566,
+      "learning_rate": 0.00015943822083633026,
+      "loss": 0.9613,
+      "step": 6951
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2832070363192226,
+      "learning_rate": 0.00015942549833216043,
+      "loss": 1.047,
+      "step": 6952
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2851606867131872,
+      "learning_rate": 0.00015941277434083014,
+      "loss": 1.1776,
+      "step": 6953
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26970800321487326,
+      "learning_rate": 0.00015940004886265781,
+      "loss": 1.0471,
+      "step": 6954
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30024373875561594,
+      "learning_rate": 0.00015938732189796196,
+      "loss": 1.1125,
+      "step": 6955
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29870880904188896,
+      "learning_rate": 0.00015937459344706105,
+      "loss": 1.106,
+      "step": 6956
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2990216369851441,
+      "learning_rate": 0.0001593618635102736,
+      "loss": 1.0811,
+      "step": 6957
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2848511215046961,
+      "learning_rate": 0.00015934913208791825,
+      "loss": 1.0252,
+      "step": 6958
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26962263498375855,
+      "learning_rate": 0.00015933639918031353,
+      "loss": 1.1282,
+      "step": 6959
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3226088976764872,
+      "learning_rate": 0.00015932366478777816,
+      "loss": 1.1156,
+      "step": 6960
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2775279024141733,
+      "learning_rate": 0.00015931092891063078,
+      "loss": 1.0649,
+      "step": 6961
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29817813856951386,
+      "learning_rate": 0.0001592981915491901,
+      "loss": 1.0693,
+      "step": 6962
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2685645767583313,
+      "learning_rate": 0.00015928545270377494,
+      "loss": 1.019,
+      "step": 6963
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2950719704962928,
+      "learning_rate": 0.00015927271237470408,
+      "loss": 1.0845,
+      "step": 6964
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2834261349957536,
+      "learning_rate": 0.0001592599705622963,
+      "loss": 1.153,
+      "step": 6965
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2703350081406132,
+      "learning_rate": 0.00015924722726687058,
+      "loss": 1.1079,
+      "step": 6966
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2573244560178056,
+      "learning_rate": 0.0001592344824887457,
+      "loss": 1.0111,
+      "step": 6967
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24815316601701018,
+      "learning_rate": 0.0001592217362282407,
+      "loss": 1.0481,
+      "step": 6968
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30891957392030694,
+      "learning_rate": 0.0001592089884856745,
+      "loss": 1.0926,
+      "step": 6969
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30606523617707265,
+      "learning_rate": 0.00015919623926136618,
+      "loss": 1.099,
+      "step": 6970
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28728857946398434,
+      "learning_rate": 0.00015918348855563477,
+      "loss": 1.1097,
+      "step": 6971
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26905395320578945,
+      "learning_rate": 0.00015917073636879936,
+      "loss": 1.0316,
+      "step": 6972
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2791301663055386,
+      "learning_rate": 0.00015915798270117905,
+      "loss": 1.078,
+      "step": 6973
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27771895204228286,
+      "learning_rate": 0.0001591452275530931,
+      "loss": 0.9038,
+      "step": 6974
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27933418779286445,
+      "learning_rate": 0.0001591324709248606,
+      "loss": 1.0519,
+      "step": 6975
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2816750044645727,
+      "learning_rate": 0.00015911971281680088,
+      "loss": 0.9544,
+      "step": 6976
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.285922508579133,
+      "learning_rate": 0.0001591069532292332,
+      "loss": 1.0725,
+      "step": 6977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.33215001802712657,
+      "learning_rate": 0.00015909419216247688,
+      "loss": 1.0101,
+      "step": 6978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23393732531173045,
+      "learning_rate": 0.00015908142961685125,
+      "loss": 1.0705,
+      "step": 6979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.257596551482316,
+      "learning_rate": 0.0001590686655926757,
+      "loss": 1.1013,
+      "step": 6980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3135249829190419,
+      "learning_rate": 0.00015905590009026967,
+      "loss": 1.0304,
+      "step": 6981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2856466968173445,
+      "learning_rate": 0.00015904313310995263,
+      "loss": 0.9811,
+      "step": 6982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2829987935572402,
+      "learning_rate": 0.00015903036465204407,
+      "loss": 1.1061,
+      "step": 6983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2780452591589329,
+      "learning_rate": 0.00015901759471686358,
+      "loss": 1.1667,
+      "step": 6984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27814002322278586,
+      "learning_rate": 0.00015900482330473062,
+      "loss": 1.1414,
+      "step": 6985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2860266253901477,
+      "learning_rate": 0.0001589920504159649,
+      "loss": 0.9647,
+      "step": 6986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2760940301311592,
+      "learning_rate": 0.0001589792760508861,
+      "loss": 1.0185,
+      "step": 6987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2706769903514577,
+      "learning_rate": 0.00015896650020981378,
+      "loss": 1.2201,
+      "step": 6988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2622304954412599,
+      "learning_rate": 0.00015895372289306776,
+      "loss": 1.0844,
+      "step": 6989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30180292321521496,
+      "learning_rate": 0.00015894094410096775,
+      "loss": 0.9252,
+      "step": 6990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.280964051345383,
+      "learning_rate": 0.0001589281638338336,
+      "loss": 1.0206,
+      "step": 6991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2722800997770079,
+      "learning_rate": 0.0001589153820919851,
+      "loss": 0.9916,
+      "step": 6992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27884975208948837,
+      "learning_rate": 0.00015890259887574215,
+      "loss": 0.9978,
+      "step": 6993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24598496860742308,
+      "learning_rate": 0.00015888981418542462,
+      "loss": 1.0904,
+      "step": 6994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2961053801063102,
+      "learning_rate": 0.00015887702802135252,
+      "loss": 0.9764,
+      "step": 6995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26389220997829704,
+      "learning_rate": 0.00015886424038384577,
+      "loss": 1.0735,
+      "step": 6996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3058819655409103,
+      "learning_rate": 0.00015885145127322438,
+      "loss": 1.1001,
+      "step": 6997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.279716453620408,
+      "learning_rate": 0.00015883866068980846,
+      "loss": 1.1435,
+      "step": 6998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2938126610887859,
+      "learning_rate": 0.00015882586863391807,
+      "loss": 1.0411,
+      "step": 6999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.24676335418681838,
+      "learning_rate": 0.00015881307510587337,
+      "loss": 0.9959,
+      "step": 7000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2653940896900372,
+      "learning_rate": 0.0001588002801059945,
+      "loss": 1.0835,
+      "step": 7001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3109595788932631,
+      "learning_rate": 0.00015878748363460163,
+      "loss": 1.0282,
+      "step": 7002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30081836243424015,
+      "learning_rate": 0.00015877468569201506,
+      "loss": 0.9415,
+      "step": 7003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31448590635210893,
+      "learning_rate": 0.00015876188627855507,
+      "loss": 1.074,
+      "step": 7004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3017891643305516,
+      "learning_rate": 0.00015874908539454188,
+      "loss": 1.0724,
+      "step": 7005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3277818013492165,
+      "learning_rate": 0.00015873628304029596,
+      "loss": 1.1018,
+      "step": 7006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2802892779964618,
+      "learning_rate": 0.00015872347921613763,
+      "loss": 1.0991,
+      "step": 7007
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28604449999798276,
+      "learning_rate": 0.0001587106739223873,
+      "loss": 0.9591,
+      "step": 7008
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2974499942407421,
+      "learning_rate": 0.0001586978671593655,
+      "loss": 1.0818,
+      "step": 7009
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26346530008044805,
+      "learning_rate": 0.00015868505892739266,
+      "loss": 1.0774,
+      "step": 7010
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30282010918915797,
+      "learning_rate": 0.00015867224922678933,
+      "loss": 1.0961,
+      "step": 7011
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31016719288564637,
+      "learning_rate": 0.0001586594380578761,
+      "loss": 1.1236,
+      "step": 7012
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28052033504795476,
+      "learning_rate": 0.00015864662542097358,
+      "loss": 1.0947,
+      "step": 7013
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.23638403426307897,
+      "learning_rate": 0.00015863381131640236,
+      "loss": 1.0231,
+      "step": 7014
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.291740553126471,
+      "learning_rate": 0.00015862099574448317,
+      "loss": 0.9802,
+      "step": 7015
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3029857623248768,
+      "learning_rate": 0.00015860817870553677,
+      "loss": 1.12,
+      "step": 7016
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2630855945498341,
+      "learning_rate": 0.00015859536019988384,
+      "loss": 0.9507,
+      "step": 7017
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2758349794123403,
+      "learning_rate": 0.00015858254022784515,
+      "loss": 1.1074,
+      "step": 7018
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27661938873325614,
+      "learning_rate": 0.00015856971878974163,
+      "loss": 1.0374,
+      "step": 7019
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2714412044951736,
+      "learning_rate": 0.00015855689588589405,
+      "loss": 1.052,
+      "step": 7020
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2677737703288316,
+      "learning_rate": 0.00015854407151662337,
+      "loss": 1.0147,
+      "step": 7021
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2927994644038343,
+      "learning_rate": 0.0001585312456822505,
+      "loss": 1.0889,
+      "step": 7022
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2790949416690381,
+      "learning_rate": 0.0001585184183830964,
+      "loss": 1.1484,
+      "step": 7023
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28034621779286634,
+      "learning_rate": 0.00015850558961948217,
+      "loss": 1.0267,
+      "step": 7024
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2745504280573277,
+      "learning_rate": 0.00015849275939172874,
+      "loss": 1.1027,
+      "step": 7025
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26978925236279233,
+      "learning_rate": 0.00015847992770015725,
+      "loss": 0.9182,
+      "step": 7026
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29839939566941587,
+      "learning_rate": 0.00015846709454508883,
+      "loss": 1.0897,
+      "step": 7027
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26196842338250126,
+      "learning_rate": 0.0001584542599268446,
+      "loss": 1.0644,
+      "step": 7028
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28373187399028676,
+      "learning_rate": 0.0001584414238457458,
+      "loss": 1.0633,
+      "step": 7029
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2537841362135921,
+      "learning_rate": 0.0001584285863021136,
+      "loss": 1.0892,
+      "step": 7030
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27473559874173337,
+      "learning_rate": 0.00015841574729626935,
+      "loss": 0.9814,
+      "step": 7031
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30313556481972587,
+      "learning_rate": 0.00015840290682853428,
+      "loss": 1.016,
+      "step": 7032
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28512884288704177,
+      "learning_rate": 0.0001583900648992298,
+      "loss": 0.9909,
+      "step": 7033
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2551073361295876,
+      "learning_rate": 0.00015837722150867722,
+      "loss": 0.9779,
+      "step": 7034
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2939908816315301,
+      "learning_rate": 0.00015836437665719798,
+      "loss": 1.1285,
+      "step": 7035
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2725649426346149,
+      "learning_rate": 0.00015835153034511357,
+      "loss": 1.1472,
+      "step": 7036
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2718133362148875,
+      "learning_rate": 0.0001583386825727454,
+      "loss": 1.0443,
+      "step": 7037
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27419111607826285,
+      "learning_rate": 0.00015832583334041505,
+      "loss": 1.0012,
+      "step": 7038
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2849089224699415,
+      "learning_rate": 0.00015831298264844406,
+      "loss": 1.0507,
+      "step": 7039
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3068910572836713,
+      "learning_rate": 0.00015830013049715404,
+      "loss": 1.0838,
+      "step": 7040
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2779576724478555,
+      "learning_rate": 0.0001582872768868666,
+      "loss": 1.0766,
+      "step": 7041
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3302291612230447,
+      "learning_rate": 0.00015827442181790344,
+      "loss": 0.9843,
+      "step": 7042
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.25384666865589006,
+      "learning_rate": 0.00015826156529058624,
+      "loss": 1.0086,
+      "step": 7043
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27280371588140856,
+      "learning_rate": 0.00015824870730523675,
+      "loss": 1.0686,
+      "step": 7044
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3030164504363727,
+      "learning_rate": 0.00015823584786217677,
+      "loss": 1.0226,
+      "step": 7045
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26909739332574156,
+      "learning_rate": 0.00015822298696172805,
+      "loss": 0.99,
+      "step": 7046
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2877874749400455,
+      "learning_rate": 0.00015821012460421255,
+      "loss": 0.9923,
+      "step": 7047
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2877538039040546,
+      "learning_rate": 0.00015819726078995208,
+      "loss": 1.0115,
+      "step": 7048
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30848201813052933,
+      "learning_rate": 0.00015818439551926856,
+      "loss": 1.0591,
+      "step": 7049
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.29140546183062116,
+      "learning_rate": 0.000158171528792484,
+      "loss": 1.0429,
+      "step": 7050
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.26994340789160964,
+      "learning_rate": 0.00015815866060992035,
+      "loss": 1.011,
+      "step": 7051
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2792969580723623,
+      "learning_rate": 0.00015814579097189966,
+      "loss": 1.0891,
+      "step": 7052
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.28250949172509243,
+      "learning_rate": 0.00015813291987874407,
+      "loss": 1.1065,
+      "step": 7053
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2978217599721609,
+      "learning_rate": 0.00015812004733077554,
+      "loss": 0.9388,
+      "step": 7054
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2908978622536936,
+      "learning_rate": 0.00015810717332831635,
+      "loss": 1.0988,
+      "step": 7055
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2799056103902689,
+      "learning_rate": 0.0001580942978716886,
+      "loss": 1.1076,
+      "step": 7056
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30269500535419785,
+      "learning_rate": 0.00015808142096121456,
+      "loss": 1.1737,
+      "step": 7057
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28238103258320846,
+      "learning_rate": 0.00015806854259721646,
+      "loss": 1.0724,
+      "step": 7058
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2985043169531798,
+      "learning_rate": 0.00015805566278001657,
+      "loss": 0.9103,
+      "step": 7059
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.293719688850402,
+      "learning_rate": 0.00015804278150993722,
+      "loss": 1.0877,
+      "step": 7060
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27699549920215416,
+      "learning_rate": 0.00015802989878730084,
+      "loss": 1.1226,
+      "step": 7061
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26487298015420774,
+      "learning_rate": 0.00015801701461242974,
+      "loss": 1.1144,
+      "step": 7062
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26324552701590875,
+      "learning_rate": 0.00015800412898564636,
+      "loss": 1.173,
+      "step": 7063
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27384596081191975,
+      "learning_rate": 0.00015799124190727322,
+      "loss": 0.9868,
+      "step": 7064
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25963681695290847,
+      "learning_rate": 0.00015797835337763282,
+      "loss": 0.9288,
+      "step": 7065
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29013003407961624,
+      "learning_rate": 0.00015796546339704766,
+      "loss": 1.0005,
+      "step": 7066
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3020156469079804,
+      "learning_rate": 0.00015795257196584038,
+      "loss": 0.9843,
+      "step": 7067
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2895277107493972,
+      "learning_rate": 0.00015793967908433353,
+      "loss": 1.105,
+      "step": 7068
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2755240824809856,
+      "learning_rate": 0.0001579267847528498,
+      "loss": 1.082,
+      "step": 7069
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30036604934270644,
+      "learning_rate": 0.00015791388897171192,
+      "loss": 1.2018,
+      "step": 7070
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30614155747981436,
+      "learning_rate": 0.00015790099174124253,
+      "loss": 1.0759,
+      "step": 7071
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2941340821458282,
+      "learning_rate": 0.00015788809306176447,
+      "loss": 1.0597,
+      "step": 7072
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30856171230251395,
+      "learning_rate": 0.00015787519293360044,
+      "loss": 1.1318,
+      "step": 7073
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29580057657767084,
+      "learning_rate": 0.00015786229135707338,
+      "loss": 1.0522,
+      "step": 7074
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.274616244018645,
+      "learning_rate": 0.0001578493883325061,
+      "loss": 1.0299,
+      "step": 7075
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3114772800268081,
+      "learning_rate": 0.00015783648386022151,
+      "loss": 1.161,
+      "step": 7076
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2802238017688198,
+      "learning_rate": 0.0001578235779405426,
+      "loss": 1.0197,
+      "step": 7077
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2944810135727739,
+      "learning_rate": 0.00015781067057379228,
+      "loss": 1.1013,
+      "step": 7078
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24998357909349614,
+      "learning_rate": 0.00015779776176029356,
+      "loss": 1.0692,
+      "step": 7079
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2945811953423457,
+      "learning_rate": 0.0001577848515003696,
+      "loss": 1.068,
+      "step": 7080
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25806411996635104,
+      "learning_rate": 0.00015777193979434338,
+      "loss": 1.0476,
+      "step": 7081
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2598704969021484,
+      "learning_rate": 0.00015775902664253808,
+      "loss": 1.0567,
+      "step": 7082
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28645413042100387,
+      "learning_rate": 0.00015774611204527681,
+      "loss": 1.0646,
+      "step": 7083
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25410792181961595,
+      "learning_rate": 0.0001577331960028828,
+      "loss": 0.9907,
+      "step": 7084
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26092421697879675,
+      "learning_rate": 0.00015772027851567927,
+      "loss": 0.9608,
+      "step": 7085
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28943346794498837,
+      "learning_rate": 0.00015770735958398952,
+      "loss": 1.0073,
+      "step": 7086
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26566974144868033,
+      "learning_rate": 0.0001576944392081368,
+      "loss": 1.109,
+      "step": 7087
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28616557026410305,
+      "learning_rate": 0.0001576815173884445,
+      "loss": 1.1788,
+      "step": 7088
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3140389757089799,
+      "learning_rate": 0.00015766859412523596,
+      "loss": 1.1028,
+      "step": 7089
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2903534504256985,
+      "learning_rate": 0.0001576556694188346,
+      "loss": 1.0848,
+      "step": 7090
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2739660674757886,
+      "learning_rate": 0.00015764274326956392,
+      "loss": 1.088,
+      "step": 7091
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26183889146766914,
+      "learning_rate": 0.00015762981567774733,
+      "loss": 1.045,
+      "step": 7092
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2773165562228698,
+      "learning_rate": 0.0001576168866437084,
+      "loss": 1.0625,
+      "step": 7093
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28250207452060827,
+      "learning_rate": 0.00015760395616777064,
+      "loss": 1.0738,
+      "step": 7094
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2664810800693953,
+      "learning_rate": 0.0001575910242502577,
+      "loss": 1.1199,
+      "step": 7095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29450171685022813,
+      "learning_rate": 0.00015757809089149319,
+      "loss": 1.0634,
+      "step": 7096
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28549635512987454,
+      "learning_rate": 0.00015756515609180073,
+      "loss": 1.0337,
+      "step": 7097
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28429692726911365,
+      "learning_rate": 0.00015755221985150412,
+      "loss": 1.1475,
+      "step": 7098
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30253389968364547,
+      "learning_rate": 0.000157539282170927,
+      "loss": 1.047,
+      "step": 7099
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3098958322121182,
+      "learning_rate": 0.00015752634305039317,
+      "loss": 0.965,
+      "step": 7100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25143769789194087,
+      "learning_rate": 0.0001575134024902265,
+      "loss": 0.9581,
+      "step": 7101
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3016367636629631,
+      "learning_rate": 0.00015750046049075076,
+      "loss": 1.0633,
+      "step": 7102
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3271227303031079,
+      "learning_rate": 0.00015748751705228984,
+      "loss": 0.9946,
+      "step": 7103
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25230564082743323,
+      "learning_rate": 0.00015747457217516768,
+      "loss": 0.998,
+      "step": 7104
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29985887178796994,
+      "learning_rate": 0.00015746162585970826,
+      "loss": 1.0747,
+      "step": 7105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3268684025188319,
+      "learning_rate": 0.00015744867810623553,
+      "loss": 1.2323,
+      "step": 7106
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24907462724647375,
+      "learning_rate": 0.0001574357289150735,
+      "loss": 1.0212,
+      "step": 7107
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2670820715697481,
+      "learning_rate": 0.0001574227782865463,
+      "loss": 0.9549,
+      "step": 7108
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2872956326475909,
+      "learning_rate": 0.00015740982622097793,
+      "loss": 1.0685,
+      "step": 7109
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3131384318834513,
+      "learning_rate": 0.00015739687271869258,
+      "loss": 1.0596,
+      "step": 7110
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2658074059936171,
+      "learning_rate": 0.00015738391778001446,
+      "loss": 1.0484,
+      "step": 7111
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2628890032193891,
+      "learning_rate": 0.00015737096140526773,
+      "loss": 0.9861,
+      "step": 7112
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3111781620355991,
+      "learning_rate": 0.0001573580035947766,
+      "loss": 1.0708,
+      "step": 7113
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28054369956222347,
+      "learning_rate": 0.0001573450443488654,
+      "loss": 1.0379,
+      "step": 7114
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2972730553838039,
+      "learning_rate": 0.00015733208366785847,
+      "loss": 1.1547,
+      "step": 7115
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2731099842689578,
+      "learning_rate": 0.00015731912155208004,
+      "loss": 1.0216,
+      "step": 7116
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25449973903583906,
+      "learning_rate": 0.0001573061580018546,
+      "loss": 1.0567,
+      "step": 7117
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3130389985554061,
+      "learning_rate": 0.00015729319301750655,
+      "loss": 1.0886,
+      "step": 7118
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2764607660175635,
+      "learning_rate": 0.00015728022659936033,
+      "loss": 0.9724,
+      "step": 7119
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24734793912425435,
+      "learning_rate": 0.0001572672587477404,
+      "loss": 1.0651,
+      "step": 7120
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28747093554729464,
+      "learning_rate": 0.00015725428946297137,
+      "loss": 1.1001,
+      "step": 7121
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.270729761427449,
+      "learning_rate": 0.0001572413187453778,
+      "loss": 1.0384,
+      "step": 7122
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2928842971220595,
+      "learning_rate": 0.00015722834659528422,
+      "loss": 1.1656,
+      "step": 7123
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2793916142108935,
+      "learning_rate": 0.00015721537301301527,
+      "loss": 0.9688,
+      "step": 7124
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29909818823600665,
+      "learning_rate": 0.0001572023979988957,
+      "loss": 0.9318,
+      "step": 7125
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2696167713839793,
+      "learning_rate": 0.00015718942155325011,
+      "loss": 1.1889,
+      "step": 7126
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29698260893246986,
+      "learning_rate": 0.00015717644367640334,
+      "loss": 1.1648,
+      "step": 7127
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27565722448673946,
+      "learning_rate": 0.00015716346436868016,
+      "loss": 1.0083,
+      "step": 7128
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27200588461690844,
+      "learning_rate": 0.00015715048363040533,
+      "loss": 1.0808,
+      "step": 7129
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.274483688022852,
+      "learning_rate": 0.00015713750146190372,
+      "loss": 1.0244,
+      "step": 7130
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2603147750459217,
+      "learning_rate": 0.00015712451786350023,
+      "loss": 1.0541,
+      "step": 7131
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2611050046236202,
+      "learning_rate": 0.0001571115328355198,
+      "loss": 0.9513,
+      "step": 7132
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2611354213480838,
+      "learning_rate": 0.00015709854637828733,
+      "loss": 1.0073,
+      "step": 7133
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29780693358025156,
+      "learning_rate": 0.0001570855584921279,
+      "loss": 1.0521,
+      "step": 7134
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.298538656430755,
+      "learning_rate": 0.00015707256917736647,
+      "loss": 1.2377,
+      "step": 7135
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2748841653644941,
+      "learning_rate": 0.0001570595784343281,
+      "loss": 0.9403,
+      "step": 7136
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2746937829833261,
+      "learning_rate": 0.00015704658626333794,
+      "loss": 1.1224,
+      "step": 7137
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2842293587504884,
+      "learning_rate": 0.00015703359266472112,
+      "loss": 1.1599,
+      "step": 7138
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2670841998105975,
+      "learning_rate": 0.0001570205976388028,
+      "loss": 1.138,
+      "step": 7139
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27756874671553816,
+      "learning_rate": 0.00015700760118590815,
+      "loss": 0.9721,
+      "step": 7140
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28756494584263737,
+      "learning_rate": 0.00015699460330636248,
+      "loss": 1.1181,
+      "step": 7141
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.32898360339581917,
+      "learning_rate": 0.00015698160400049105,
+      "loss": 1.219,
+      "step": 7142
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2803241831359587,
+      "learning_rate": 0.00015696860326861917,
+      "loss": 1.1124,
+      "step": 7143
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.27535423833378253,
+      "learning_rate": 0.00015695560111107218,
+      "loss": 1.1046,
+      "step": 7144
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2606122473280347,
+      "learning_rate": 0.0001569425975281755,
+      "loss": 1.158,
+      "step": 7145
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26610041445452237,
+      "learning_rate": 0.00015692959252025447,
+      "loss": 1.1423,
+      "step": 7146
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2953405796549938,
+      "learning_rate": 0.00015691658608763467,
+      "loss": 1.1065,
+      "step": 7147
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2809208524967904,
+      "learning_rate": 0.00015690357823064147,
+      "loss": 1.0091,
+      "step": 7148
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.298379105314993,
+      "learning_rate": 0.00015689056894960054,
+      "loss": 0.9916,
+      "step": 7149
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29683307271349985,
+      "learning_rate": 0.00015687755824483733,
+      "loss": 1.1125,
+      "step": 7150
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.29899393435333316,
+      "learning_rate": 0.00015686454611667745,
+      "loss": 1.0796,
+      "step": 7151
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31478792374668824,
+      "learning_rate": 0.00015685153256544658,
+      "loss": 1.0593,
+      "step": 7152
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2930746130040074,
+      "learning_rate": 0.0001568385175914704,
+      "loss": 1.0662,
+      "step": 7153
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2882046106628799,
+      "learning_rate": 0.00015682550119507457,
+      "loss": 1.0464,
+      "step": 7154
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.24774454066142032,
+      "learning_rate": 0.0001568124833765849,
+      "loss": 1.0727,
+      "step": 7155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28400220165986323,
+      "learning_rate": 0.0001567994641363271,
+      "loss": 1.2001,
+      "step": 7156
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2847089193485156,
+      "learning_rate": 0.000156786443474627,
+      "loss": 1.0203,
+      "step": 7157
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.25521289971018285,
+      "learning_rate": 0.0001567734213918105,
+      "loss": 1.0875,
+      "step": 7158
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.26777210499225285,
+      "learning_rate": 0.0001567603978882034,
+      "loss": 1.0413,
+      "step": 7159
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29958229941874526,
+      "learning_rate": 0.00015674737296413171,
+      "loss": 1.1779,
+      "step": 7160
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2973923709791303,
+      "learning_rate": 0.00015673434661992133,
+      "loss": 1.0605,
+      "step": 7161
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29890629646161765,
+      "learning_rate": 0.00015672131885589827,
+      "loss": 1.1378,
+      "step": 7162
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27465315968911586,
+      "learning_rate": 0.00015670828967238857,
+      "loss": 1.0176,
+      "step": 7163
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27330309587431983,
+      "learning_rate": 0.00015669525906971825,
+      "loss": 1.099,
+      "step": 7164
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2830955541275837,
+      "learning_rate": 0.00015668222704821346,
+      "loss": 1.0174,
+      "step": 7165
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2918518106052513,
+      "learning_rate": 0.00015666919360820034,
+      "loss": 1.0828,
+      "step": 7166
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2938118596159364,
+      "learning_rate": 0.000156656158750005,
+      "loss": 1.1368,
+      "step": 7167
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33564117245367897,
+      "learning_rate": 0.0001566431224739537,
+      "loss": 1.1594,
+      "step": 7168
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2791063027357139,
+      "learning_rate": 0.00015663008478037263,
+      "loss": 1.0643,
+      "step": 7169
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2836589857893266,
+      "learning_rate": 0.00015661704566958816,
+      "loss": 1.0865,
+      "step": 7170
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3268578411481459,
+      "learning_rate": 0.00015660400514192648,
+      "loss": 1.1146,
+      "step": 7171
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26432781847496883,
+      "learning_rate": 0.00015659096319771401,
+      "loss": 1.2087,
+      "step": 7172
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2947543226709198,
+      "learning_rate": 0.00015657791983727715,
+      "loss": 1.0841,
+      "step": 7173
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3064445919955032,
+      "learning_rate": 0.00015656487506094226,
+      "loss": 1.1338,
+      "step": 7174
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2768287567420236,
+      "learning_rate": 0.00015655182886903582,
+      "loss": 1.1217,
+      "step": 7175
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3115480635288813,
+      "learning_rate": 0.00015653878126188433,
+      "loss": 1.1119,
+      "step": 7176
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26658700110300837,
+      "learning_rate": 0.00015652573223981432,
+      "loss": 1.1532,
+      "step": 7177
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30729315346172287,
+      "learning_rate": 0.0001565126818031523,
+      "loss": 1.0824,
+      "step": 7178
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27545018818322364,
+      "learning_rate": 0.00015649962995222493,
+      "loss": 1.0008,
+      "step": 7179
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2840294196959938,
+      "learning_rate": 0.0001564865766873588,
+      "loss": 0.9888,
+      "step": 7180
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2704386331778757,
+      "learning_rate": 0.00015647352200888056,
+      "loss": 1.0845,
+      "step": 7181
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.309378454139959,
+      "learning_rate": 0.00015646046591711698,
+      "loss": 1.1094,
+      "step": 7182
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33236253311844427,
+      "learning_rate": 0.00015644740841239477,
+      "loss": 1.0607,
+      "step": 7183
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2654938165964374,
+      "learning_rate": 0.00015643434949504066,
+      "loss": 1.0977,
+      "step": 7184
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.25891560294027866,
+      "learning_rate": 0.00015642128916538148,
+      "loss": 1.0862,
+      "step": 7185
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30191735715591245,
+      "learning_rate": 0.00015640822742374411,
+      "loss": 1.0986,
+      "step": 7186
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3016522569398602,
+      "learning_rate": 0.00015639516427045538,
+      "loss": 1.1071,
+      "step": 7187
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3747262023726705,
+      "learning_rate": 0.00015638209970584218,
+      "loss": 0.9347,
+      "step": 7188
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26969894694899865,
+      "learning_rate": 0.0001563690337302316,
+      "loss": 0.9472,
+      "step": 7189
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27707566766666775,
+      "learning_rate": 0.00015635596634395045,
+      "loss": 1.0062,
+      "step": 7190
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.299605953934922,
+      "learning_rate": 0.00015634289754732584,
+      "loss": 1.0789,
+      "step": 7191
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2756403847856143,
+      "learning_rate": 0.00015632982734068479,
+      "loss": 1.0478,
+      "step": 7192
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28164477492401824,
+      "learning_rate": 0.00015631675572435442,
+      "loss": 1.0234,
+      "step": 7193
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3139943484530815,
+      "learning_rate": 0.00015630368269866187,
+      "loss": 1.0499,
+      "step": 7194
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30405542664327845,
+      "learning_rate": 0.0001562906082639342,
+      "loss": 1.023,
+      "step": 7195
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2874490807251814,
+      "learning_rate": 0.00015627753242049877,
+      "loss": 0.9521,
+      "step": 7196
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2643259838229851,
+      "learning_rate": 0.0001562644551686827,
+      "loss": 1.0435,
+      "step": 7197
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2865014648758085,
+      "learning_rate": 0.00015625137650881324,
+      "loss": 1.0175,
+      "step": 7198
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30454373875693874,
+      "learning_rate": 0.00015623829644121777,
+      "loss": 1.134,
+      "step": 7199
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31859855773654133,
+      "learning_rate": 0.00015622521496622355,
+      "loss": 1.0659,
+      "step": 7200
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2858177196834928,
+      "learning_rate": 0.00015621213208415804,
+      "loss": 0.9598,
+      "step": 7201
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26598227866462754,
+      "learning_rate": 0.00015619904779534856,
+      "loss": 0.9907,
+      "step": 7202
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2951400583438705,
+      "learning_rate": 0.00015618596210012256,
+      "loss": 1.0765,
+      "step": 7203
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3114702561846883,
+      "learning_rate": 0.00015617287499880762,
+      "loss": 1.1041,
+      "step": 7204
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.278499185049389,
+      "learning_rate": 0.00015615978649173112,
+      "loss": 0.9533,
+      "step": 7205
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28555299537963824,
+      "learning_rate": 0.0001561466965792207,
+      "loss": 1.0967,
+      "step": 7206
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2636715267212951,
+      "learning_rate": 0.00015613360526160392,
+      "loss": 1.086,
+      "step": 7207
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29854739591999657,
+      "learning_rate": 0.00015612051253920836,
+      "loss": 1.1259,
+      "step": 7208
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29584977312779975,
+      "learning_rate": 0.00015610741841236173,
+      "loss": 1.1717,
+      "step": 7209
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2853540873126679,
+      "learning_rate": 0.00015609432288139167,
+      "loss": 1.0673,
+      "step": 7210
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26104277026079287,
+      "learning_rate": 0.00015608122594662596,
+      "loss": 1.0772,
+      "step": 7211
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3189693042004204,
+      "learning_rate": 0.00015606812760839226,
+      "loss": 1.0342,
+      "step": 7212
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27816239495102424,
+      "learning_rate": 0.00015605502786701848,
+      "loss": 1.0825,
+      "step": 7213
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2635774551759029,
+      "learning_rate": 0.0001560419267228324,
+      "loss": 1.0783,
+      "step": 7214
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29195022738920845,
+      "learning_rate": 0.00015602882417616184,
+      "loss": 1.1758,
+      "step": 7215
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3144273867404004,
+      "learning_rate": 0.0001560157202273348,
+      "loss": 1.0973,
+      "step": 7216
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3065507511047062,
+      "learning_rate": 0.00015600261487667912,
+      "loss": 1.1111,
+      "step": 7217
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3131853451412242,
+      "learning_rate": 0.0001559895081245228,
+      "loss": 1.2187,
+      "step": 7218
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26515742013497673,
+      "learning_rate": 0.00015597639997119389,
+      "loss": 1.0605,
+      "step": 7219
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3129582152206669,
+      "learning_rate": 0.00015596329041702036,
+      "loss": 1.0686,
+      "step": 7220
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34295960089795735,
+      "learning_rate": 0.00015595017946233033,
+      "loss": 1.0771,
+      "step": 7221
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2721307338351633,
+      "learning_rate": 0.00015593706710745187,
+      "loss": 1.0284,
+      "step": 7222
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.304506700589439,
+      "learning_rate": 0.00015592395335271316,
+      "loss": 1.055,
+      "step": 7223
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.31053881654022203,
+      "learning_rate": 0.0001559108381984424,
+      "loss": 1.1429,
+      "step": 7224
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30691636576138964,
+      "learning_rate": 0.00015589772164496774,
+      "loss": 1.1403,
+      "step": 7225
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2833916505818309,
+      "learning_rate": 0.00015588460369261748,
+      "loss": 0.9685,
+      "step": 7226
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28649595640599806,
+      "learning_rate": 0.0001558714843417199,
+      "loss": 0.9536,
+      "step": 7227
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27284801442293516,
+      "learning_rate": 0.0001558583635926033,
+      "loss": 1.1062,
+      "step": 7228
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27007347326614567,
+      "learning_rate": 0.00015584524144559604,
+      "loss": 1.1615,
+      "step": 7229
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3039328987168462,
+      "learning_rate": 0.00015583211790102652,
+      "loss": 1.1376,
+      "step": 7230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26912448480042633,
+      "learning_rate": 0.00015581899295922318,
+      "loss": 1.0096,
+      "step": 7231
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2653339147614556,
+      "learning_rate": 0.00015580586662051444,
+      "loss": 1.0514,
+      "step": 7232
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26697292715319065,
+      "learning_rate": 0.0001557927388852288,
+      "loss": 0.9873,
+      "step": 7233
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2937309164974668,
+      "learning_rate": 0.00015577960975369484,
+      "loss": 1.1235,
+      "step": 7234
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2781751496250474,
+      "learning_rate": 0.00015576647922624105,
+      "loss": 1.0094,
+      "step": 7235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28016398794516684,
+      "learning_rate": 0.00015575334730319611,
+      "loss": 1.1144,
+      "step": 7236
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2756048290515205,
+      "learning_rate": 0.00015574021398488862,
+      "loss": 1.0359,
+      "step": 7237
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.24131864496426217,
+      "learning_rate": 0.0001557270792716472,
+      "loss": 1.0668,
+      "step": 7238
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2840317470056723,
+      "learning_rate": 0.00015571394316380062,
+      "loss": 1.1596,
+      "step": 7239
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27376264865951977,
+      "learning_rate": 0.0001557008056616776,
+      "loss": 0.9976,
+      "step": 7240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3019067450221086,
+      "learning_rate": 0.0001556876667656069,
+      "loss": 1.0813,
+      "step": 7241
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3162835708665788,
+      "learning_rate": 0.00015567452647591732,
+      "loss": 1.1084,
+      "step": 7242
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27896797749608704,
+      "learning_rate": 0.00015566138479293775,
+      "loss": 1.0595,
+      "step": 7243
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2992448181482971,
+      "learning_rate": 0.00015564824171699707,
+      "loss": 0.9487,
+      "step": 7244
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27553640224406994,
+      "learning_rate": 0.00015563509724842413,
+      "loss": 1.0404,
+      "step": 7245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27552371209875953,
+      "learning_rate": 0.00015562195138754792,
+      "loss": 1.1528,
+      "step": 7246
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29426928783634326,
+      "learning_rate": 0.00015560880413469742,
+      "loss": 1.0984,
+      "step": 7247
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27853895487179603,
+      "learning_rate": 0.00015559565549020169,
+      "loss": 1.1804,
+      "step": 7248
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3136870635528011,
+      "learning_rate": 0.00015558250545438972,
+      "loss": 1.0698,
+      "step": 7249
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2624215588147093,
+      "learning_rate": 0.0001555693540275906,
+      "loss": 1.1105,
+      "step": 7250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.23782920442038563,
+      "learning_rate": 0.0001555562012101335,
+      "loss": 1.1211,
+      "step": 7251
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2695440813802512,
+      "learning_rate": 0.00015554304700234747,
+      "loss": 1.0031,
+      "step": 7252
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30147711904941943,
+      "learning_rate": 0.00015552989140456185,
+      "loss": 1.1955,
+      "step": 7253
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2430503125575205,
+      "learning_rate": 0.0001555167344171058,
+      "loss": 0.983,
+      "step": 7254
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29554732708566633,
+      "learning_rate": 0.00015550357604030856,
+      "loss": 1.0728,
+      "step": 7255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.33084597515765934,
+      "learning_rate": 0.00015549041627449945,
+      "loss": 1.0754,
+      "step": 7256
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.35533675482762406,
+      "learning_rate": 0.0001554772551200078,
+      "loss": 1.1181,
+      "step": 7257
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27809709182399867,
+      "learning_rate": 0.00015546409257716296,
+      "loss": 1.1706,
+      "step": 7258
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2644018070633185,
+      "learning_rate": 0.00015545092864629437,
+      "loss": 1.1147,
+      "step": 7259
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2749285252651721,
+      "learning_rate": 0.00015543776332773142,
+      "loss": 1.1475,
+      "step": 7260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.26759109696003713,
+      "learning_rate": 0.00015542459662180362,
+      "loss": 1.1167,
+      "step": 7261
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2744664795704848,
+      "learning_rate": 0.00015541142852884044,
+      "loss": 0.9624,
+      "step": 7262
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2656603725314074,
+      "learning_rate": 0.00015539825904917147,
+      "loss": 1.0158,
+      "step": 7263
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.27736864447841364,
+      "learning_rate": 0.0001553850881831262,
+      "loss": 1.0009,
+      "step": 7264
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.23778024632317593,
+      "learning_rate": 0.00015537191593103432,
+      "loss": 1.1653,
+      "step": 7265
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30730004897313506,
+      "learning_rate": 0.00015535874229322545,
+      "loss": 1.0293,
+      "step": 7266
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26576264865967103,
+      "learning_rate": 0.00015534556727002925,
+      "loss": 1.1149,
+      "step": 7267
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2910068427246134,
+      "learning_rate": 0.00015533239086177548,
+      "loss": 1.0587,
+      "step": 7268
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2719761437844997,
+      "learning_rate": 0.0001553192130687938,
+      "loss": 1.0378,
+      "step": 7269
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2578029237737743,
+      "learning_rate": 0.00015530603389141408,
+      "loss": 1.0591,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2799149293630919,
+      "learning_rate": 0.0001552928533299661,
+      "loss": 1.0374,
+      "step": 7271
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.260847576690139,
+      "learning_rate": 0.00015527967138477967,
+      "loss": 0.9975,
+      "step": 7272
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2907795091020474,
+      "learning_rate": 0.00015526648805618478,
+      "loss": 1.1593,
+      "step": 7273
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.22310418172043128,
+      "learning_rate": 0.00015525330334451127,
+      "loss": 0.9877,
+      "step": 7274
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29425332767908924,
+      "learning_rate": 0.00015524011725008912,
+      "loss": 1.0992,
+      "step": 7275
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3105156950201626,
+      "learning_rate": 0.0001552269297732483,
+      "loss": 1.0496,
+      "step": 7276
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2558739957093287,
+      "learning_rate": 0.00015521374091431888,
+      "loss": 1.0964,
+      "step": 7277
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28594690539220524,
+      "learning_rate": 0.00015520055067363089,
+      "loss": 1.0916,
+      "step": 7278
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3061482044460731,
+      "learning_rate": 0.00015518735905151442,
+      "loss": 1.0903,
+      "step": 7279
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3191580668312093,
+      "learning_rate": 0.00015517416604829962,
+      "loss": 0.9281,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26708235106774547,
+      "learning_rate": 0.00015516097166431663,
+      "loss": 1.1284,
+      "step": 7281
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2741503886756114,
+      "learning_rate": 0.00015514777589989564,
+      "loss": 0.9834,
+      "step": 7282
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2721226495347994,
+      "learning_rate": 0.00015513457875536692,
+      "loss": 1.0924,
+      "step": 7283
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2903941706772716,
+      "learning_rate": 0.0001551213802310607,
+      "loss": 1.1032,
+      "step": 7284
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26719034153148796,
+      "learning_rate": 0.0001551081803273073,
+      "loss": 1.0721,
+      "step": 7285
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2410538608083278,
+      "learning_rate": 0.00015509497904443706,
+      "loss": 1.0446,
+      "step": 7286
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.288052793290722,
+      "learning_rate": 0.00015508177638278036,
+      "loss": 1.0968,
+      "step": 7287
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28114598468656504,
+      "learning_rate": 0.00015506857234266755,
+      "loss": 1.2161,
+      "step": 7288
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30363151390349236,
+      "learning_rate": 0.00015505536692442915,
+      "loss": 1.1299,
+      "step": 7289
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3430101339177829,
+      "learning_rate": 0.00015504216012839555,
+      "loss": 1.057,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31124798109609725,
+      "learning_rate": 0.00015502895195489735,
+      "loss": 1.0329,
+      "step": 7291
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31977891525097923,
+      "learning_rate": 0.000155015742404265,
+      "loss": 1.0195,
+      "step": 7292
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2605016731210418,
+      "learning_rate": 0.00015500253147682913,
+      "loss": 1.1187,
+      "step": 7293
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25377938560950036,
+      "learning_rate": 0.00015498931917292037,
+      "loss": 0.9918,
+      "step": 7294
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3046468563219072,
+      "learning_rate": 0.0001549761054928693,
+      "loss": 1.112,
+      "step": 7295
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.34132422940513657,
+      "learning_rate": 0.00015496289043700665,
+      "loss": 1.0857,
+      "step": 7296
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.25593975051636847,
+      "learning_rate": 0.00015494967400566311,
+      "loss": 1.0094,
+      "step": 7297
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30660036228331966,
+      "learning_rate": 0.00015493645619916947,
+      "loss": 1.0384,
+      "step": 7298
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28934490627014536,
+      "learning_rate": 0.0001549232370178565,
+      "loss": 1.1638,
+      "step": 7299
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3082067568322258,
+      "learning_rate": 0.00015491001646205496,
+      "loss": 1.1484,
+      "step": 7300
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.24608283441694623,
+      "learning_rate": 0.00015489679453209578,
+      "loss": 1.0935,
+      "step": 7301
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2878652096204575,
+      "learning_rate": 0.0001548835712283098,
+      "loss": 1.0195,
+      "step": 7302
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2445700870996487,
+      "learning_rate": 0.00015487034655102796,
+      "loss": 1.0676,
+      "step": 7303
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2792976608554407,
+      "learning_rate": 0.00015485712050058125,
+      "loss": 1.0282,
+      "step": 7304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2792852923073424,
+      "learning_rate": 0.00015484389307730056,
+      "loss": 1.0393,
+      "step": 7305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2921531944021332,
+      "learning_rate": 0.00015483066428151703,
+      "loss": 1.1671,
+      "step": 7306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2840087290455877,
+      "learning_rate": 0.00015481743411356163,
+      "loss": 1.0849,
+      "step": 7307
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3040563023087563,
+      "learning_rate": 0.0001548042025737655,
+      "loss": 1.1268,
+      "step": 7308
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29355520065474283,
+      "learning_rate": 0.00015479096966245978,
+      "loss": 1.03,
+      "step": 7309
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2760436996975324,
+      "learning_rate": 0.00015477773537997557,
+      "loss": 1.1871,
+      "step": 7310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27687371076005574,
+      "learning_rate": 0.00015476449972664412,
+      "loss": 1.1466,
+      "step": 7311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2845099189305587,
+      "learning_rate": 0.00015475126270279667,
+      "loss": 1.0636,
+      "step": 7312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2505523226102105,
+      "learning_rate": 0.00015473802430876444,
+      "loss": 1.0668,
+      "step": 7313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32837255120480047,
+      "learning_rate": 0.00015472478454487876,
+      "loss": 1.1224,
+      "step": 7314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26481048965460596,
+      "learning_rate": 0.00015471154341147094,
+      "loss": 1.0428,
+      "step": 7315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2690353020587212,
+      "learning_rate": 0.00015469830090887235,
+      "loss": 1.0432,
+      "step": 7316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27626600703824794,
+      "learning_rate": 0.00015468505703741442,
+      "loss": 1.0638,
+      "step": 7317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3065923845948702,
+      "learning_rate": 0.00015467181179742857,
+      "loss": 0.914,
+      "step": 7318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2829656174177215,
+      "learning_rate": 0.0001546585651892463,
+      "loss": 1.0586,
+      "step": 7319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.277648311025837,
+      "learning_rate": 0.00015464531721319903,
+      "loss": 0.9368,
+      "step": 7320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26414058914433497,
+      "learning_rate": 0.00015463206786961838,
+      "loss": 0.9696,
+      "step": 7321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3161051076868171,
+      "learning_rate": 0.0001546188171588359,
+      "loss": 1.0229,
+      "step": 7322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2799019895774647,
+      "learning_rate": 0.0001546055650811832,
+      "loss": 1.1345,
+      "step": 7323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.305865812690483,
+      "learning_rate": 0.0001545923116369919,
+      "loss": 1.0023,
+      "step": 7324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30382443041697793,
+      "learning_rate": 0.00015457905682659368,
+      "loss": 1.0488,
+      "step": 7325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.31386464628323874,
+      "learning_rate": 0.0001545658006503203,
+      "loss": 1.0361,
+      "step": 7326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27019240058841637,
+      "learning_rate": 0.00015455254310850345,
+      "loss": 1.0565,
+      "step": 7327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.266426547020745,
+      "learning_rate": 0.0001545392842014749,
+      "loss": 0.9752,
+      "step": 7328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2995784354581383,
+      "learning_rate": 0.0001545260239295665,
+      "loss": 1.0005,
+      "step": 7329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.317973794140023,
+      "learning_rate": 0.0001545127622931101,
+      "loss": 1.0538,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2947298100738568,
+      "learning_rate": 0.00015449949929243755,
+      "loss": 1.0116,
+      "step": 7331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3348856644428043,
+      "learning_rate": 0.00015448623492788076,
+      "loss": 0.9802,
+      "step": 7332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2989455629929353,
+      "learning_rate": 0.00015447296919977172,
+      "loss": 1.0376,
+      "step": 7333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3011061423556309,
+      "learning_rate": 0.0001544597021084424,
+      "loss": 1.0908,
+      "step": 7334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3049819395201172,
+      "learning_rate": 0.00015444643365422478,
+      "loss": 1.0768,
+      "step": 7335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26542548997246496,
+      "learning_rate": 0.00015443316383745095,
+      "loss": 1.024,
+      "step": 7336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2797529721086807,
+      "learning_rate": 0.00015441989265845297,
+      "loss": 1.072,
+      "step": 7337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2538780155773843,
+      "learning_rate": 0.000154406620117563,
+      "loss": 1.038,
+      "step": 7338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2831186717367567,
+      "learning_rate": 0.00015439334621511318,
+      "loss": 1.0638,
+      "step": 7339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28787268851064585,
+      "learning_rate": 0.00015438007095143567,
+      "loss": 1.0426,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2948120261226757,
+      "learning_rate": 0.0001543667943268627,
+      "loss": 1.1221,
+      "step": 7341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32242230412435835,
+      "learning_rate": 0.00015435351634172654,
+      "loss": 1.0958,
+      "step": 7342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2654273360662479,
+      "learning_rate": 0.00015434023699635948,
+      "loss": 1.0071,
+      "step": 7343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27327197799382885,
+      "learning_rate": 0.00015432695629109385,
+      "loss": 1.0576,
+      "step": 7344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2754942416375848,
+      "learning_rate": 0.00015431367422626195,
+      "loss": 1.1498,
+      "step": 7345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27801354524717076,
+      "learning_rate": 0.00015430039080219625,
+      "loss": 1.1084,
+      "step": 7346
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2914103943454639,
+      "learning_rate": 0.00015428710601922914,
+      "loss": 1.0507,
+      "step": 7347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2638150945585758,
+      "learning_rate": 0.00015427381987769307,
+      "loss": 1.0324,
+      "step": 7348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2526962446018732,
+      "learning_rate": 0.0001542605323779206,
+      "loss": 0.9352,
+      "step": 7349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2595273949493344,
+      "learning_rate": 0.00015424724352024413,
+      "loss": 1.0355,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28782174706723024,
+      "learning_rate": 0.00015423395330499632,
+      "loss": 1.1083,
+      "step": 7351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3115649702709986,
+      "learning_rate": 0.00015422066173250974,
+      "loss": 1.0422,
+      "step": 7352
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29111738038165,
+      "learning_rate": 0.00015420736880311707,
+      "loss": 1.0315,
+      "step": 7353
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28032086546612484,
+      "learning_rate": 0.00015419407451715088,
+      "loss": 1.1888,
+      "step": 7354
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2716417113526186,
+      "learning_rate": 0.00015418077887494394,
+      "loss": 1.0331,
+      "step": 7355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28116455658959255,
+      "learning_rate": 0.00015416748187682897,
+      "loss": 1.0433,
+      "step": 7356
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3165740582098866,
+      "learning_rate": 0.00015415418352313868,
+      "loss": 1.0488,
+      "step": 7357
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.24167939460825494,
+      "learning_rate": 0.00015414088381420594,
+      "loss": 1.1047,
+      "step": 7358
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2362202925878594,
+      "learning_rate": 0.00015412758275036356,
+      "loss": 1.111,
+      "step": 7359
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.30927705441472164,
+      "learning_rate": 0.0001541142803319444,
+      "loss": 0.9948,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.26694430612905373,
+      "learning_rate": 0.00015410097655928136,
+      "loss": 1.2141,
+      "step": 7361
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28735202961238854,
+      "learning_rate": 0.00015408767143270738,
+      "loss": 1.0482,
+      "step": 7362
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3028944111263906,
+      "learning_rate": 0.00015407436495255543,
+      "loss": 0.9829,
+      "step": 7363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3036916238157443,
+      "learning_rate": 0.0001540610571191585,
+      "loss": 1.1743,
+      "step": 7364
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2897890002077795,
+      "learning_rate": 0.00015404774793284967,
+      "loss": 1.0033,
+      "step": 7365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.29386071816379244,
+      "learning_rate": 0.00015403443739396195,
+      "loss": 1.1666,
+      "step": 7366
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.32062513986248775,
+      "learning_rate": 0.0001540211255028285,
+      "loss": 1.141,
+      "step": 7367
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3180897552174907,
+      "learning_rate": 0.00015400781225978242,
+      "loss": 1.0144,
+      "step": 7368
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.27245479542574785,
+      "learning_rate": 0.00015399449766515688,
+      "loss": 1.0604,
+      "step": 7369
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2915213084795467,
+      "learning_rate": 0.00015398118171928516,
+      "loss": 1.1338,
+      "step": 7370
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23168432157865712,
+      "learning_rate": 0.00015396786442250035,
+      "loss": 1.0496,
+      "step": 7371
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2782432262911835,
+      "learning_rate": 0.0001539545457751359,
+      "loss": 1.0979,
+      "step": 7372
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27977300422557033,
+      "learning_rate": 0.000153941225777525,
+      "loss": 1.0744,
+      "step": 7373
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2545144680006588,
+      "learning_rate": 0.000153927904430001,
+      "loss": 1.0237,
+      "step": 7374
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.23952050585666343,
+      "learning_rate": 0.00015391458173289734,
+      "loss": 1.1329,
+      "step": 7375
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27650814084254066,
+      "learning_rate": 0.00015390125768654738,
+      "loss": 1.128,
+      "step": 7376
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2957362698116198,
+      "learning_rate": 0.00015388793229128455,
+      "loss": 1.1255,
+      "step": 7377
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24485467093779958,
+      "learning_rate": 0.00015387460554744235,
+      "loss": 1.0197,
+      "step": 7378
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25214439017148976,
+      "learning_rate": 0.0001538612774553543,
+      "loss": 1.0601,
+      "step": 7379
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.282880344840166,
+      "learning_rate": 0.00015384794801535394,
+      "loss": 1.1426,
+      "step": 7380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3152140814752052,
+      "learning_rate": 0.00015383461722777482,
+      "loss": 1.0964,
+      "step": 7381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24688971723773373,
+      "learning_rate": 0.00015382128509295058,
+      "loss": 1.0926,
+      "step": 7382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2959651478445274,
+      "learning_rate": 0.00015380795161121485,
+      "loss": 1.0349,
+      "step": 7383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29305053218157,
+      "learning_rate": 0.0001537946167829013,
+      "loss": 1.1347,
+      "step": 7384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2717846083673245,
+      "learning_rate": 0.00015378128060834366,
+      "loss": 1.0096,
+      "step": 7385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2620308073344996,
+      "learning_rate": 0.0001537679430878757,
+      "loss": 1.0911,
+      "step": 7386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2531071813077481,
+      "learning_rate": 0.00015375460422183116,
+      "loss": 0.9925,
+      "step": 7387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2797391994034549,
+      "learning_rate": 0.00015374126401054383,
+      "loss": 1.1076,
+      "step": 7388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2727192530520494,
+      "learning_rate": 0.00015372792245434765,
+      "loss": 1.0334,
+      "step": 7389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29997950673640783,
+      "learning_rate": 0.00015371457955357643,
+      "loss": 0.9707,
+      "step": 7390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2816972910580794,
+      "learning_rate": 0.00015370123530856407,
+      "loss": 1.1254,
+      "step": 7391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2959827743331956,
+      "learning_rate": 0.00015368788971964454,
+      "loss": 1.0022,
+      "step": 7392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3027989628861647,
+      "learning_rate": 0.0001536745427871519,
+      "loss": 1.0555,
+      "step": 7393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27628323794583914,
+      "learning_rate": 0.00015366119451142002,
+      "loss": 0.985,
+      "step": 7394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3098642640517758,
+      "learning_rate": 0.00015364784489278304,
+      "loss": 1.0772,
+      "step": 7395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3045327679601191,
+      "learning_rate": 0.00015363449393157504,
+      "loss": 1.1383,
+      "step": 7396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27647845171066426,
+      "learning_rate": 0.00015362114162813012,
+      "loss": 1.0514,
+      "step": 7397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27708090315989137,
+      "learning_rate": 0.00015360778798278243,
+      "loss": 0.9376,
+      "step": 7398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2903239285477044,
+      "learning_rate": 0.00015359443299586614,
+      "loss": 1.1016,
+      "step": 7399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3010839768456833,
+      "learning_rate": 0.0001535810766677155,
+      "loss": 1.0842,
+      "step": 7400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.276746949232895,
+      "learning_rate": 0.0001535677189986647,
+      "loss": 1.2018,
+      "step": 7401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29483934180185845,
+      "learning_rate": 0.0001535543599890481,
+      "loss": 1.0932,
+      "step": 7402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2944209969280942,
+      "learning_rate": 0.0001535409996392,
+      "loss": 1.1482,
+      "step": 7403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31126699379727446,
+      "learning_rate": 0.0001535276379494547,
+      "loss": 1.1388,
+      "step": 7404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2843244483748962,
+      "learning_rate": 0.00015351427492014662,
+      "loss": 1.122,
+      "step": 7405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2940291993707008,
+      "learning_rate": 0.00015350091055161023,
+      "loss": 0.8989,
+      "step": 7406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3351477452765215,
+      "learning_rate": 0.0001534875448441799,
+      "loss": 1.1108,
+      "step": 7407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2906887381085749,
+      "learning_rate": 0.0001534741777981901,
+      "loss": 0.9912,
+      "step": 7408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3122468085486279,
+      "learning_rate": 0.00015346080941397544,
+      "loss": 1.113,
+      "step": 7409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28278667430577964,
+      "learning_rate": 0.00015344743969187042,
+      "loss": 1.0403,
+      "step": 7410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27569491315509187,
+      "learning_rate": 0.00015343406863220962,
+      "loss": 1.121,
+      "step": 7411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2804750598763221,
+      "learning_rate": 0.0001534206962353277,
+      "loss": 1.0365,
+      "step": 7412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2524239751697441,
+      "learning_rate": 0.00015340732250155927,
+      "loss": 0.9917,
+      "step": 7413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29005569119447944,
+      "learning_rate": 0.000153393947431239,
+      "loss": 1.1511,
+      "step": 7414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2749251352441248,
+      "learning_rate": 0.00015338057102470164,
+      "loss": 1.0658,
+      "step": 7415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2838189917053858,
+      "learning_rate": 0.000153367193282282,
+      "loss": 1.0391,
+      "step": 7416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3244226146833391,
+      "learning_rate": 0.00015335381420431476,
+      "loss": 1.1094,
+      "step": 7417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2574324902880777,
+      "learning_rate": 0.0001533404337911348,
+      "loss": 1.1391,
+      "step": 7418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30913323555851724,
+      "learning_rate": 0.00015332705204307696,
+      "loss": 1.0444,
+      "step": 7419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2902578699002574,
+      "learning_rate": 0.00015331366896047613,
+      "loss": 0.9731,
+      "step": 7420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2736910311823876,
+      "learning_rate": 0.00015330028454366723,
+      "loss": 1.096,
+      "step": 7421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2595210721180384,
+      "learning_rate": 0.0001532868987929852,
+      "loss": 1.0444,
+      "step": 7422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2691043276914828,
+      "learning_rate": 0.00015327351170876504,
+      "loss": 0.9763,
+      "step": 7423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2625617634998602,
+      "learning_rate": 0.00015326012329134177,
+      "loss": 1.0655,
+      "step": 7424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3040586098635329,
+      "learning_rate": 0.00015324673354105044,
+      "loss": 1.0892,
+      "step": 7425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2688279407074845,
+      "learning_rate": 0.00015323334245822613,
+      "loss": 1.0881,
+      "step": 7426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2772663570425649,
+      "learning_rate": 0.00015321995004320398,
+      "loss": 1.0775,
+      "step": 7427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28558771196591126,
+      "learning_rate": 0.00015320655629631915,
+      "loss": 1.1682,
+      "step": 7428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3068876838609169,
+      "learning_rate": 0.00015319316121790676,
+      "loss": 1.1325,
+      "step": 7429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25376782398833597,
+      "learning_rate": 0.00015317976480830214,
+      "loss": 1.051,
+      "step": 7430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26689013628589564,
+      "learning_rate": 0.00015316636706784047,
+      "loss": 1.0535,
+      "step": 7431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27913817719785716,
+      "learning_rate": 0.00015315296799685703,
+      "loss": 1.0799,
+      "step": 7432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27105117018470604,
+      "learning_rate": 0.00015313956759568717,
+      "loss": 0.9799,
+      "step": 7433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2650201668901937,
+      "learning_rate": 0.00015312616586466625,
+      "loss": 1.0213,
+      "step": 7434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2597414213724003,
+      "learning_rate": 0.0001531127628041296,
+      "loss": 1.0696,
+      "step": 7435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2720574465290946,
+      "learning_rate": 0.0001530993584144127,
+      "loss": 1.0064,
+      "step": 7436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2926790787938213,
+      "learning_rate": 0.000153085952695851,
+      "loss": 1.0582,
+      "step": 7437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30451030733173934,
+      "learning_rate": 0.00015307254564877996,
+      "loss": 1.0364,
+      "step": 7438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2969242563944365,
+      "learning_rate": 0.00015305913727353508,
+      "loss": 0.9929,
+      "step": 7439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2843406055001775,
+      "learning_rate": 0.000153045727570452,
+      "loss": 1.0706,
+      "step": 7440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27079473763903633,
+      "learning_rate": 0.0001530323165398662,
+      "loss": 1.13,
+      "step": 7441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31673668159080653,
+      "learning_rate": 0.00015301890418211337,
+      "loss": 1.033,
+      "step": 7442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.25852536375912855,
+      "learning_rate": 0.00015300549049752915,
+      "loss": 0.9558,
+      "step": 7443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2610535794849494,
+      "learning_rate": 0.00015299207548644922,
+      "loss": 0.9075,
+      "step": 7444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3144932740991477,
+      "learning_rate": 0.0001529786591492093,
+      "loss": 1.088,
+      "step": 7445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3002199992017584,
+      "learning_rate": 0.0001529652414861451,
+      "loss": 1.0292,
+      "step": 7446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29137989165257083,
+      "learning_rate": 0.00015295182249759246,
+      "loss": 1.1373,
+      "step": 7447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3000740319743772,
+      "learning_rate": 0.0001529384021838872,
+      "loss": 1.0349,
+      "step": 7448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37952107317617517,
+      "learning_rate": 0.00015292498054536515,
+      "loss": 1.027,
+      "step": 7449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2658313512672852,
+      "learning_rate": 0.00015291155758236219,
+      "loss": 1.0365,
+      "step": 7450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30711739440319363,
+      "learning_rate": 0.00015289813329521427,
+      "loss": 1.0623,
+      "step": 7451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26947346016409285,
+      "learning_rate": 0.0001528847076842573,
+      "loss": 1.0008,
+      "step": 7452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.294244616915276,
+      "learning_rate": 0.00015287128074982728,
+      "loss": 1.1527,
+      "step": 7453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26434224087661645,
+      "learning_rate": 0.00015285785249226025,
+      "loss": 1.0076,
+      "step": 7454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2638878195945205,
+      "learning_rate": 0.00015284442291189224,
+      "loss": 1.0129,
+      "step": 7455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.291822498691914,
+      "learning_rate": 0.0001528309920090593,
+      "loss": 1.0994,
+      "step": 7456
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.24079989611181216,
+      "learning_rate": 0.00015281755978409763,
+      "loss": 0.9589,
+      "step": 7457
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26439658524128423,
+      "learning_rate": 0.00015280412623734331,
+      "loss": 1.0177,
+      "step": 7458
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.28359751564704094,
+      "learning_rate": 0.00015279069136913252,
+      "loss": 1.0652,
+      "step": 7459
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.27983466260304535,
+      "learning_rate": 0.00015277725517980152,
+      "loss": 1.0485,
+      "step": 7460
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29708736109513595,
+      "learning_rate": 0.00015276381766968656,
+      "loss": 1.074,
+      "step": 7461
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29289233835975614,
+      "learning_rate": 0.0001527503788391239,
+      "loss": 1.1547,
+      "step": 7462
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2505395055508232,
+      "learning_rate": 0.00015273693868844983,
+      "loss": 0.9737,
+      "step": 7463
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.32901669963923147,
+      "learning_rate": 0.00015272349721800075,
+      "loss": 1.0536,
+      "step": 7464
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.256290518603839,
+      "learning_rate": 0.000152710054428113,
+      "loss": 1.1606,
+      "step": 7465
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26769516189713743,
+      "learning_rate": 0.000152696610319123,
+      "loss": 0.9284,
+      "step": 7466
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2644771395655536,
+      "learning_rate": 0.00015268316489136722,
+      "loss": 1.1177,
+      "step": 7467
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2806512107610077,
+      "learning_rate": 0.00015266971814518213,
+      "loss": 0.9686,
+      "step": 7468
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2845865397349829,
+      "learning_rate": 0.00015265627008090424,
+      "loss": 1.0744,
+      "step": 7469
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.26504898270571114,
+      "learning_rate": 0.00015264282069887012,
+      "loss": 1.1322,
+      "step": 7470
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2916082135751575,
+      "learning_rate": 0.0001526293699994163,
+      "loss": 1.103,
+      "step": 7471
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31144997066611424,
+      "learning_rate": 0.00015261591798287945,
+      "loss": 1.0014,
+      "step": 7472
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2574195705247452,
+      "learning_rate": 0.00015260246464959614,
+      "loss": 1.1275,
+      "step": 7473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31598031433107676,
+      "learning_rate": 0.00015258900999990313,
+      "loss": 0.9769,
+      "step": 7474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2733626438470682,
+      "learning_rate": 0.00015257555403413707,
+      "loss": 1.1188,
+      "step": 7475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27826283583830214,
+      "learning_rate": 0.00015256209675263473,
+      "loss": 0.9995,
+      "step": 7476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3089275459108109,
+      "learning_rate": 0.0001525486381557329,
+      "loss": 1.0316,
+      "step": 7477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2928951184842569,
+      "learning_rate": 0.00015253517824376838,
+      "loss": 1.0514,
+      "step": 7478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26365816516744844,
+      "learning_rate": 0.00015252171701707798,
+      "loss": 1.0454,
+      "step": 7479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28372734957336343,
+      "learning_rate": 0.00015250825447599863,
+      "loss": 1.0397,
+      "step": 7480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.24476440536696728,
+      "learning_rate": 0.0001524947906208672,
+      "loss": 1.1185,
+      "step": 7481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29553606635449614,
+      "learning_rate": 0.00015248132545202066,
+      "loss": 1.0536,
+      "step": 7482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28192161411982414,
+      "learning_rate": 0.00015246785896979592,
+      "loss": 1.0789,
+      "step": 7483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28836399016991415,
+      "learning_rate": 0.00015245439117453005,
+      "loss": 1.1376,
+      "step": 7484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.285252586401288,
+      "learning_rate": 0.00015244092206656012,
+      "loss": 1.0715,
+      "step": 7485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34180059296617976,
+      "learning_rate": 0.0001524274516462231,
+      "loss": 1.0974,
+      "step": 7486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28411030591653186,
+      "learning_rate": 0.0001524139799138562,
+      "loss": 1.1046,
+      "step": 7487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2688829534104327,
+      "learning_rate": 0.00015240050686979648,
+      "loss": 1.0534,
+      "step": 7488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3171051572047796,
+      "learning_rate": 0.00015238703251438116,
+      "loss": 0.9673,
+      "step": 7489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2715311978491321,
+      "learning_rate": 0.00015237355684794742,
+      "loss": 1.0009,
+      "step": 7490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3107037826927023,
+      "learning_rate": 0.0001523600798708325,
+      "loss": 1.1782,
+      "step": 7491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3004283097955692,
+      "learning_rate": 0.00015234660158337367,
+      "loss": 1.1222,
+      "step": 7492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.34256065220175824,
+      "learning_rate": 0.00015233312198590824,
+      "loss": 1.1572,
+      "step": 7493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30264641686373545,
+      "learning_rate": 0.00015231964107877355,
+      "loss": 0.9944,
+      "step": 7494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29115852098449335,
+      "learning_rate": 0.00015230615886230696,
+      "loss": 1.0331,
+      "step": 7495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2663234013844971,
+      "learning_rate": 0.00015229267533684588,
+      "loss": 0.9686,
+      "step": 7496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2706122042940165,
+      "learning_rate": 0.00015227919050272775,
+      "loss": 1.029,
+      "step": 7497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28192878540642374,
+      "learning_rate": 0.00015226570436028996,
+      "loss": 1.0831,
+      "step": 7498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3136140113978119,
+      "learning_rate": 0.00015225221690987013,
+      "loss": 1.1534,
+      "step": 7499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29044795123500067,
+      "learning_rate": 0.0001522387281518057,
+      "loss": 1.0986,
+      "step": 7500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25709365283504254,
+      "learning_rate": 0.00015222523808643428,
+      "loss": 1.1544,
+      "step": 7501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27642043583466885,
+      "learning_rate": 0.00015221174671409347,
+      "loss": 1.0419,
+      "step": 7502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28235849443833444,
+      "learning_rate": 0.00015219825403512086,
+      "loss": 1.1054,
+      "step": 7503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3347988942349422,
+      "learning_rate": 0.00015218476004985414,
+      "loss": 1.1444,
+      "step": 7504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28011799381802555,
+      "learning_rate": 0.000152171264758631,
+      "loss": 1.0586,
+      "step": 7505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29631516225770976,
+      "learning_rate": 0.00015215776816178918,
+      "loss": 1.066,
+      "step": 7506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.312587038609111,
+      "learning_rate": 0.00015214427025966642,
+      "loss": 1.06,
+      "step": 7507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26267971626273046,
+      "learning_rate": 0.00015213077105260053,
+      "loss": 0.99,
+      "step": 7508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29362317699509005,
+      "learning_rate": 0.00015211727054092932,
+      "loss": 1.0374,
+      "step": 7509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2743031690741454,
+      "learning_rate": 0.00015210376872499068,
+      "loss": 0.9676,
+      "step": 7510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3008169494678444,
+      "learning_rate": 0.0001520902656051225,
+      "loss": 1.0962,
+      "step": 7511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2510179331269226,
+      "learning_rate": 0.00015207676118166266,
+      "loss": 1.0657,
+      "step": 7512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3006570781756079,
+      "learning_rate": 0.00015206325545494913,
+      "loss": 1.1252,
+      "step": 7513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3074556798522242,
+      "learning_rate": 0.00015204974842531995,
+      "loss": 1.1561,
+      "step": 7514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3251122148768213,
+      "learning_rate": 0.00015203624009311307,
+      "loss": 1.1283,
+      "step": 7515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2595576229638888,
+      "learning_rate": 0.0001520227304586666,
+      "loss": 0.9711,
+      "step": 7516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29928452954008067,
+      "learning_rate": 0.00015200921952231858,
+      "loss": 1.0824,
+      "step": 7517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3139867616135075,
+      "learning_rate": 0.0001519957072844072,
+      "loss": 1.1125,
+      "step": 7518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29374530619599887,
+      "learning_rate": 0.00015198219374527053,
+      "loss": 1.1183,
+      "step": 7519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2952280509086445,
+      "learning_rate": 0.00015196867890524676,
+      "loss": 1.0937,
+      "step": 7520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2826852773745996,
+      "learning_rate": 0.00015195516276467422,
+      "loss": 1.1857,
+      "step": 7521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2803858395978367,
+      "learning_rate": 0.000151941645323891,
+      "loss": 0.9876,
+      "step": 7522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27577711685224354,
+      "learning_rate": 0.00015192812658323552,
+      "loss": 1.0477,
+      "step": 7523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29937814744106744,
+      "learning_rate": 0.00015191460654304602,
+      "loss": 1.0781,
+      "step": 7524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.344475738203272,
+      "learning_rate": 0.00015190108520366085,
+      "loss": 1.0677,
+      "step": 7525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.269619028961468,
+      "learning_rate": 0.00015188756256541842,
+      "loss": 1.1207,
+      "step": 7526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2856659171795633,
+      "learning_rate": 0.0001518740386286571,
+      "loss": 0.9971,
+      "step": 7527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2690448461790446,
+      "learning_rate": 0.0001518605133937154,
+      "loss": 1.0434,
+      "step": 7528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3153664787777255,
+      "learning_rate": 0.00015184698686093173,
+      "loss": 1.1609,
+      "step": 7529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31002810501031186,
+      "learning_rate": 0.00015183345903064467,
+      "loss": 1.0097,
+      "step": 7530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2764530793093042,
+      "learning_rate": 0.00015181992990319265,
+      "loss": 1.0459,
+      "step": 7531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2581950187662785,
+      "learning_rate": 0.00015180639947891437,
+      "loss": 1.0612,
+      "step": 7532
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3228690135774238,
+      "learning_rate": 0.0001517928677581484,
+      "loss": 1.1375,
+      "step": 7533
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2830701920026826,
+      "learning_rate": 0.0001517793347412333,
+      "loss": 0.9702,
+      "step": 7534
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30970817365242975,
+      "learning_rate": 0.00015176580042850787,
+      "loss": 1.0011,
+      "step": 7535
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2879799004430592,
+      "learning_rate": 0.00015175226482031073,
+      "loss": 1.0469,
+      "step": 7536
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2721982692551152,
+      "learning_rate": 0.0001517387279169806,
+      "loss": 1.0661,
+      "step": 7537
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30110301529460076,
+      "learning_rate": 0.00015172518971885634,
+      "loss": 1.0512,
+      "step": 7538
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26509470901065213,
+      "learning_rate": 0.00015171165022627667,
+      "loss": 1.0675,
+      "step": 7539
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29294036548567454,
+      "learning_rate": 0.00015169810943958044,
+      "loss": 1.12,
+      "step": 7540
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2756337326490096,
+      "learning_rate": 0.00015168456735910657,
+      "loss": 1.0387,
+      "step": 7541
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2974971175746808,
+      "learning_rate": 0.0001516710239851939,
+      "loss": 1.1836,
+      "step": 7542
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31089357056437184,
+      "learning_rate": 0.0001516574793181814,
+      "loss": 1.1023,
+      "step": 7543
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30077365644294696,
+      "learning_rate": 0.00015164393335840798,
+      "loss": 0.9271,
+      "step": 7544
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2565124797415133,
+      "learning_rate": 0.00015163038610621269,
+      "loss": 1.0294,
+      "step": 7545
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2826401399220724,
+      "learning_rate": 0.00015161683756193456,
+      "loss": 1.0563,
+      "step": 7546
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30362864947281726,
+      "learning_rate": 0.00015160328772591256,
+      "loss": 1.1748,
+      "step": 7547
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29370524398637515,
+      "learning_rate": 0.00015158973659848592,
+      "loss": 1.0426,
+      "step": 7548
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2684686193459749,
+      "learning_rate": 0.00015157618417999366,
+      "loss": 1.0464,
+      "step": 7549
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2761867607331919,
+      "learning_rate": 0.000151562630470775,
+      "loss": 1.0953,
+      "step": 7550
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25846363686227103,
+      "learning_rate": 0.0001515490754711691,
+      "loss": 1.061,
+      "step": 7551
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2642419825902769,
+      "learning_rate": 0.0001515355191815152,
+      "loss": 1.1347,
+      "step": 7552
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2559409653328977,
+      "learning_rate": 0.00015152196160215253,
+      "loss": 1.0638,
+      "step": 7553
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28318731458349305,
+      "learning_rate": 0.00015150840273342038,
+      "loss": 1.0592,
+      "step": 7554
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2850162725399112,
+      "learning_rate": 0.00015149484257565813,
+      "loss": 1.1141,
+      "step": 7555
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.26118830684653604,
+      "learning_rate": 0.000151481281129205,
+      "loss": 1.0173,
+      "step": 7556
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27293247877399707,
+      "learning_rate": 0.0001514677183944005,
+      "loss": 1.0846,
+      "step": 7557
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.304675314149257,
+      "learning_rate": 0.00015145415437158401,
+      "loss": 1.1428,
+      "step": 7558
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2972819357482946,
+      "learning_rate": 0.00015144058906109496,
+      "loss": 1.0655,
+      "step": 7559
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28299905657676133,
+      "learning_rate": 0.00015142702246327287,
+      "loss": 1.0705,
+      "step": 7560
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29542732633039054,
+      "learning_rate": 0.00015141345457845716,
+      "loss": 1.0193,
+      "step": 7561
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2663372553374203,
+      "learning_rate": 0.00015139988540698748,
+      "loss": 0.9831,
+      "step": 7562
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.25393991962427676,
+      "learning_rate": 0.00015138631494920337,
+      "loss": 1.1387,
+      "step": 7563
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27744832277420794,
+      "learning_rate": 0.00015137274320544438,
+      "loss": 1.1088,
+      "step": 7564
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27311375703469265,
+      "learning_rate": 0.00015135917017605025,
+      "loss": 1.092,
+      "step": 7565
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3264586447066704,
+      "learning_rate": 0.0001513455958613606,
+      "loss": 1.1544,
+      "step": 7566
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29926443610547104,
+      "learning_rate": 0.00015133202026171514,
+      "loss": 1.0584,
+      "step": 7567
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2945733521922124,
+      "learning_rate": 0.00015131844337745362,
+      "loss": 0.9397,
+      "step": 7568
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2730547592785977,
+      "learning_rate": 0.00015130486520891582,
+      "loss": 0.9991,
+      "step": 7569
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30119301432734397,
+      "learning_rate": 0.00015129128575644147,
+      "loss": 1.0843,
+      "step": 7570
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.30645684857055866,
+      "learning_rate": 0.00015127770502037052,
+      "loss": 1.0142,
+      "step": 7571
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2821703954502372,
+      "learning_rate": 0.00015126412300104272,
+      "loss": 0.982,
+      "step": 7572
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27920328503608616,
+      "learning_rate": 0.00015125053969879807,
+      "loss": 1.0294,
+      "step": 7573
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2385258107376846,
+      "learning_rate": 0.0001512369551139764,
+      "loss": 1.0406,
+      "step": 7574
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2972641315163676,
+      "learning_rate": 0.0001512233692469178,
+      "loss": 1.1835,
+      "step": 7575
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2602247995915003,
+      "learning_rate": 0.00015120978209796213,
+      "loss": 1.0921,
+      "step": 7576
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.27839769616677684,
+      "learning_rate": 0.00015119619366744952,
+      "loss": 1.099,
+      "step": 7577
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.316376234085856,
+      "learning_rate": 0.00015118260395571994,
+      "loss": 1.0387,
+      "step": 7578
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31660368584142895,
+      "learning_rate": 0.00015116901296311356,
+      "loss": 1.1306,
+      "step": 7579
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.260900438740011,
+      "learning_rate": 0.00015115542068997047,
+      "loss": 0.9071,
+      "step": 7580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2624881211285536,
+      "learning_rate": 0.00015114182713663082,
+      "loss": 1.1411,
+      "step": 7581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.257486571814327,
+      "learning_rate": 0.00015112823230343484,
+      "loss": 1.0307,
+      "step": 7582
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2823321215311315,
+      "learning_rate": 0.00015111463619072265,
+      "loss": 1.1471,
+      "step": 7583
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30117635588768243,
+      "learning_rate": 0.00015110103879883462,
+      "loss": 1.1452,
+      "step": 7584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3022579959192287,
+      "learning_rate": 0.00015108744012811096,
+      "loss": 1.0584,
+      "step": 7585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29740329090227624,
+      "learning_rate": 0.000151073840178892,
+      "loss": 1.0415,
+      "step": 7586
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29372205328515927,
+      "learning_rate": 0.0001510602389515181,
+      "loss": 1.1383,
+      "step": 7587
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2920566881167881,
+      "learning_rate": 0.00015104663644632962,
+      "loss": 1.0683,
+      "step": 7588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22058431176486398,
+      "learning_rate": 0.000151033032663667,
+      "loss": 1.0229,
+      "step": 7589
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25213010351059206,
+      "learning_rate": 0.00015101942760387065,
+      "loss": 1.1396,
+      "step": 7590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25918222437239363,
+      "learning_rate": 0.00015100582126728105,
+      "loss": 1.0704,
+      "step": 7591
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2961632170304152,
+      "learning_rate": 0.00015099221365423872,
+      "loss": 1.1117,
+      "step": 7592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28187090698557027,
+      "learning_rate": 0.0001509786047650842,
+      "loss": 0.9727,
+      "step": 7593
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2892636190812079,
+      "learning_rate": 0.00015096499460015805,
+      "loss": 1.013,
+      "step": 7594
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28236820071622526,
+      "learning_rate": 0.0001509513831598009,
+      "loss": 0.9727,
+      "step": 7595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2722476913875187,
+      "learning_rate": 0.00015093777044435333,
+      "loss": 1.0157,
+      "step": 7596
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2652646040859842,
+      "learning_rate": 0.00015092415645415606,
+      "loss": 1.0091,
+      "step": 7597
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28143692367142764,
+      "learning_rate": 0.00015091054118954978,
+      "loss": 1.0371,
+      "step": 7598
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2767178043247685,
+      "learning_rate": 0.0001508969246508752,
+      "loss": 1.0008,
+      "step": 7599
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25719222763437416,
+      "learning_rate": 0.0001508833068384731,
+      "loss": 1.0056,
+      "step": 7600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30194394764309046,
+      "learning_rate": 0.00015086968775268427,
+      "loss": 1.0588,
+      "step": 7601
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2775361029431778,
+      "learning_rate": 0.00015085606739384953,
+      "loss": 1.1844,
+      "step": 7602
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2529941931845389,
+      "learning_rate": 0.00015084244576230976,
+      "loss": 0.9901,
+      "step": 7603
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3165846646756356,
+      "learning_rate": 0.00015082882285840578,
+      "loss": 0.997,
+      "step": 7604
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2782657643449699,
+      "learning_rate": 0.0001508151986824786,
+      "loss": 1.0908,
+      "step": 7605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2781877735281842,
+      "learning_rate": 0.00015080157323486915,
+      "loss": 1.1314,
+      "step": 7606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.281042713599222,
+      "learning_rate": 0.00015078794651591837,
+      "loss": 1.0458,
+      "step": 7607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27849574824675116,
+      "learning_rate": 0.0001507743185259673,
+      "loss": 1.1229,
+      "step": 7608
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2759377287295315,
+      "learning_rate": 0.00015076068926535706,
+      "loss": 1.044,
+      "step": 7609
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2821493820751014,
+      "learning_rate": 0.00015074705873442863,
+      "loss": 1.0843,
+      "step": 7610
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2607548413248963,
+      "learning_rate": 0.00015073342693352316,
+      "loss": 1.0291,
+      "step": 7611
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3107914480290541,
+      "learning_rate": 0.0001507197938629818,
+      "loss": 1.1632,
+      "step": 7612
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2644427769663581,
+      "learning_rate": 0.0001507061595231457,
+      "loss": 1.112,
+      "step": 7613
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29569519438471115,
+      "learning_rate": 0.00015069252391435614,
+      "loss": 0.8825,
+      "step": 7614
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2725562040988613,
+      "learning_rate": 0.00015067888703695426,
+      "loss": 0.975,
+      "step": 7615
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27141691407540564,
+      "learning_rate": 0.00015066524889128139,
+      "loss": 1.1129,
+      "step": 7616
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.286170558294994,
+      "learning_rate": 0.00015065160947767887,
+      "loss": 1.1549,
+      "step": 7617
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28827191966294335,
+      "learning_rate": 0.00015063796879648793,
+      "loss": 1.0636,
+      "step": 7618
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34763534936988866,
+      "learning_rate": 0.00015062432684805,
+      "loss": 1.0821,
+      "step": 7619
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2928541057535552,
+      "learning_rate": 0.00015061068363270654,
+      "loss": 0.9746,
+      "step": 7620
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.22676940791019204,
+      "learning_rate": 0.00015059703915079888,
+      "loss": 0.9351,
+      "step": 7621
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2932641803272752,
+      "learning_rate": 0.0001505833934026685,
+      "loss": 1.169,
+      "step": 7622
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2545708248714656,
+      "learning_rate": 0.0001505697463886569,
+      "loss": 1.0129,
+      "step": 7623
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31554332098638554,
+      "learning_rate": 0.00015055609810910565,
+      "loss": 1.0664,
+      "step": 7624
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3159256467723614,
+      "learning_rate": 0.00015054244856435624,
+      "loss": 1.1755,
+      "step": 7625
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25618170107349153,
+      "learning_rate": 0.0001505287977547503,
+      "loss": 1.0721,
+      "step": 7626
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3162579358667636,
+      "learning_rate": 0.00015051514568062947,
+      "loss": 0.958,
+      "step": 7627
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25894438378797824,
+      "learning_rate": 0.00015050149234233532,
+      "loss": 1.0019,
+      "step": 7628
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2852760760049263,
+      "learning_rate": 0.00015048783774020962,
+      "loss": 1.1271,
+      "step": 7629
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3912607456759231,
+      "learning_rate": 0.00015047418187459405,
+      "loss": 1.1038,
+      "step": 7630
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.322140496874572,
+      "learning_rate": 0.00015046052474583033,
+      "loss": 1.0758,
+      "step": 7631
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3229270016075276,
+      "learning_rate": 0.0001504468663542603,
+      "loss": 1.0946,
+      "step": 7632
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3081581987259254,
+      "learning_rate": 0.0001504332067002257,
+      "loss": 1.0036,
+      "step": 7633
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29552024618919837,
+      "learning_rate": 0.00015041954578406844,
+      "loss": 1.0949,
+      "step": 7634
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2963238428604876,
+      "learning_rate": 0.00015040588360613034,
+      "loss": 1.0493,
+      "step": 7635
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29301025689881727,
+      "learning_rate": 0.00015039222016675332,
+      "loss": 1.0601,
+      "step": 7636
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31243960952554717,
+      "learning_rate": 0.0001503785554662793,
+      "loss": 1.0141,
+      "step": 7637
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2810933699464373,
+      "learning_rate": 0.00015036488950505032,
+      "loss": 0.9789,
+      "step": 7638
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2800782903010916,
+      "learning_rate": 0.0001503512222834083,
+      "loss": 1.0996,
+      "step": 7639
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2737931904759353,
+      "learning_rate": 0.0001503375538016953,
+      "loss": 1.0249,
+      "step": 7640
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28069227541654734,
+      "learning_rate": 0.0001503238840602534,
+      "loss": 0.9117,
+      "step": 7641
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26949283739174246,
+      "learning_rate": 0.00015031021305942464,
+      "loss": 1.0084,
+      "step": 7642
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31475584607752194,
+      "learning_rate": 0.00015029654079955118,
+      "loss": 1.0414,
+      "step": 7643
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28698683483189275,
+      "learning_rate": 0.0001502828672809752,
+      "loss": 1.0434,
+      "step": 7644
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25256982279011375,
+      "learning_rate": 0.00015026919250403883,
+      "loss": 1.0292,
+      "step": 7645
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29577865157695005,
+      "learning_rate": 0.00015025551646908437,
+      "loss": 1.1659,
+      "step": 7646
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2728622137260916,
+      "learning_rate": 0.000150241839176454,
+      "loss": 1.0527,
+      "step": 7647
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.30338258907392546,
+      "learning_rate": 0.00015022816062649005,
+      "loss": 1.1148,
+      "step": 7648
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26931970248244586,
+      "learning_rate": 0.0001502144808195348,
+      "loss": 1.1357,
+      "step": 7649
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2742596946281973,
+      "learning_rate": 0.0001502007997559306,
+      "loss": 1.0646,
+      "step": 7650
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3011532902374543,
+      "learning_rate": 0.00015018711743601984,
+      "loss": 1.1351,
+      "step": 7651
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2821557355247745,
+      "learning_rate": 0.00015017343386014494,
+      "loss": 1.0087,
+      "step": 7652
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31425399583262464,
+      "learning_rate": 0.0001501597490286483,
+      "loss": 1.0384,
+      "step": 7653
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2931669718113375,
+      "learning_rate": 0.0001501460629418724,
+      "loss": 1.0466,
+      "step": 7654
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2875430271869409,
+      "learning_rate": 0.0001501323756001598,
+      "loss": 1.0878,
+      "step": 7655
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2597750556495404,
+      "learning_rate": 0.00015011868700385298,
+      "loss": 1.0548,
+      "step": 7656
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27705706713516576,
+      "learning_rate": 0.00015010499715329453,
+      "loss": 0.984,
+      "step": 7657
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.26538202477264977,
+      "learning_rate": 0.00015009130604882702,
+      "loss": 1.0417,
+      "step": 7658
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.28174584601024333,
+      "learning_rate": 0.0001500776136907931,
+      "loss": 1.0257,
+      "step": 7659
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.25036859344556583,
+      "learning_rate": 0.00015006392007953543,
+      "loss": 1.1059,
+      "step": 7660
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27900752760462394,
+      "learning_rate": 0.00015005022521539672,
+      "loss": 1.1089,
+      "step": 7661
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31107819751826155,
+      "learning_rate": 0.0001500365290987196,
+      "loss": 1.0433,
+      "step": 7662
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2843202029978909,
+      "learning_rate": 0.00015002283172984695,
+      "loss": 1.0845,
+      "step": 7663
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2589541502663174,
+      "learning_rate": 0.00015000913310912148,
+      "loss": 0.9485,
+      "step": 7664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.300671018081089,
+      "learning_rate": 0.00014999543323688603,
+      "loss": 1.0407,
+      "step": 7665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.27922119836694304,
+      "learning_rate": 0.00014998173211348343,
+      "loss": 0.9969,
+      "step": 7666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29425343441337976,
+      "learning_rate": 0.0001499680297392566,
+      "loss": 0.9274,
+      "step": 7667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2941232322325698,
+      "learning_rate": 0.0001499543261145484,
+      "loss": 1.0088,
+      "step": 7668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2914706729504272,
+      "learning_rate": 0.0001499406212397018,
+      "loss": 1.0058,
+      "step": 7669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2953576562328182,
+      "learning_rate": 0.00014992691511505975,
+      "loss": 1.0051,
+      "step": 7670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.34306316772779066,
+      "learning_rate": 0.0001499132077409653,
+      "loss": 1.153,
+      "step": 7671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2679983439498134,
+      "learning_rate": 0.0001498994991177614,
+      "loss": 1.0104,
+      "step": 7672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2594539874622118,
+      "learning_rate": 0.00014988578924579122,
+      "loss": 1.1645,
+      "step": 7673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3048213376270825,
+      "learning_rate": 0.0001498720781253978,
+      "loss": 1.098,
+      "step": 7674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3077629403566935,
+      "learning_rate": 0.0001498583657569243,
+      "loss": 1.1484,
+      "step": 7675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31902620346830557,
+      "learning_rate": 0.00014984465214071385,
+      "loss": 1.009,
+      "step": 7676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.29403015567312174,
+      "learning_rate": 0.00014983093727710965,
+      "loss": 0.9773,
+      "step": 7677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2548359871414429,
+      "learning_rate": 0.00014981722116645495,
+      "loss": 1.0751,
+      "step": 7678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2787506686083023,
+      "learning_rate": 0.00014980350380909294,
+      "loss": 1.1147,
+      "step": 7679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2992261122500776,
+      "learning_rate": 0.00014978978520536698,
+      "loss": 1.0858,
+      "step": 7680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3011621462376174,
+      "learning_rate": 0.00014977606535562034,
+      "loss": 1.046,
+      "step": 7681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2986573455838004,
+      "learning_rate": 0.0001497623442601964,
+      "loss": 1.0208,
+      "step": 7682
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28656498019974624,
+      "learning_rate": 0.00014974862191943848,
+      "loss": 1.1759,
+      "step": 7683
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2762485785593129,
+      "learning_rate": 0.00014973489833369004,
+      "loss": 1.1052,
+      "step": 7684
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3083477893434321,
+      "learning_rate": 0.00014972117350329456,
+      "loss": 0.9617,
+      "step": 7685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3003280251753417,
+      "learning_rate": 0.0001497074474285954,
+      "loss": 1.029,
+      "step": 7686
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2572025859628947,
+      "learning_rate": 0.00014969372010993618,
+      "loss": 1.2292,
+      "step": 7687
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24791565205884394,
+      "learning_rate": 0.00014967999154766036,
+      "loss": 1.1202,
+      "step": 7688
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2834724391112066,
+      "learning_rate": 0.00014966626174211153,
+      "loss": 1.1773,
+      "step": 7689
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2793645427994029,
+      "learning_rate": 0.0001496525306936333,
+      "loss": 1.0002,
+      "step": 7690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2742995002805119,
+      "learning_rate": 0.00014963879840256927,
+      "loss": 1.1219,
+      "step": 7691
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31255435123133335,
+      "learning_rate": 0.0001496250648692631,
+      "loss": 1.0062,
+      "step": 7692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3095875184013366,
+      "learning_rate": 0.00014961133009405852,
+      "loss": 1.0721,
+      "step": 7693
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2804828390216687,
+      "learning_rate": 0.00014959759407729922,
+      "loss": 0.9511,
+      "step": 7694
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28459275473311924,
+      "learning_rate": 0.00014958385681932893,
+      "loss": 0.8945,
+      "step": 7695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30693979635734253,
+      "learning_rate": 0.00014957011832049147,
+      "loss": 1.0574,
+      "step": 7696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2661866180961575,
+      "learning_rate": 0.00014955637858113065,
+      "loss": 1.1286,
+      "step": 7697
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2748632910585887,
+      "learning_rate": 0.00014954263760159033,
+      "loss": 1.0264,
+      "step": 7698
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2639660707256238,
+      "learning_rate": 0.00014952889538221434,
+      "loss": 0.9609,
+      "step": 7699
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28950568503410234,
+      "learning_rate": 0.00014951515192334665,
+      "loss": 1.0243,
+      "step": 7700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2741181044596097,
+      "learning_rate": 0.00014950140722533114,
+      "loss": 1.1703,
+      "step": 7701
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2939600698983923,
+      "learning_rate": 0.0001494876612885118,
+      "loss": 1.0576,
+      "step": 7702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27769959394242444,
+      "learning_rate": 0.00014947391411323263,
+      "loss": 1.032,
+      "step": 7703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27126944244428325,
+      "learning_rate": 0.0001494601656998377,
+      "loss": 0.9935,
+      "step": 7704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2842820549012656,
+      "learning_rate": 0.000149446416048671,
+      "loss": 1.1503,
+      "step": 7705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2669012700932571,
+      "learning_rate": 0.0001494326651600767,
+      "loss": 1.0303,
+      "step": 7706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30368700424850303,
+      "learning_rate": 0.00014941891303439886,
+      "loss": 1.0606,
+      "step": 7707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30976284634068807,
+      "learning_rate": 0.0001494051596719817,
+      "loss": 1.0383,
+      "step": 7708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3052315022073979,
+      "learning_rate": 0.00014939140507316934,
+      "loss": 1.0763,
+      "step": 7709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28582065805745044,
+      "learning_rate": 0.00014937764923830598,
+      "loss": 1.2036,
+      "step": 7710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2783946932750405,
+      "learning_rate": 0.000149363892167736,
+      "loss": 1.0478,
+      "step": 7711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2863324153725781,
+      "learning_rate": 0.00014935013386180353,
+      "loss": 1.0988,
+      "step": 7712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2785613479438333,
+      "learning_rate": 0.00014933637432085295,
+      "loss": 1.0189,
+      "step": 7713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28499263197797736,
+      "learning_rate": 0.00014932261354522864,
+      "loss": 1.1063,
+      "step": 7714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30255835930843594,
+      "learning_rate": 0.00014930885153527492,
+      "loss": 1.0056,
+      "step": 7715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.23887877046224554,
+      "learning_rate": 0.00014929508829133616,
+      "loss": 1.0982,
+      "step": 7716
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25941492372004377,
+      "learning_rate": 0.0001492813238137569,
+      "loss": 1.0775,
+      "step": 7717
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2840455310981083,
+      "learning_rate": 0.00014926755810288147,
+      "loss": 1.0016,
+      "step": 7718
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3159294197735588,
+      "learning_rate": 0.00014925379115905444,
+      "loss": 1.1289,
+      "step": 7719
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3001807062258611,
+      "learning_rate": 0.00014924002298262034,
+      "loss": 0.9867,
+      "step": 7720
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2783216610430766,
+      "learning_rate": 0.00014922625357392376,
+      "loss": 1.0793,
+      "step": 7721
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28020041552297037,
+      "learning_rate": 0.00014921248293330922,
+      "loss": 1.0532,
+      "step": 7722
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3424716991146364,
+      "learning_rate": 0.00014919871106112135,
+      "loss": 1.0768,
+      "step": 7723
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2779795066958149,
+      "learning_rate": 0.00014918493795770482,
+      "loss": 0.9554,
+      "step": 7724
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3181269730793016,
+      "learning_rate": 0.00014917116362340435,
+      "loss": 1.0211,
+      "step": 7725
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28616822041259976,
+      "learning_rate": 0.00014915738805856458,
+      "loss": 1.1136,
+      "step": 7726
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25638910226945644,
+      "learning_rate": 0.00014914361126353026,
+      "loss": 0.9941,
+      "step": 7727
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25181470798369615,
+      "learning_rate": 0.0001491298332386462,
+      "loss": 1.0635,
+      "step": 7728
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.279163837257328,
+      "learning_rate": 0.0001491160539842572,
+      "loss": 1.0245,
+      "step": 7729
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28555043885157716,
+      "learning_rate": 0.00014910227350070805,
+      "loss": 1.1297,
+      "step": 7730
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28507597908988613,
+      "learning_rate": 0.00014908849178834366,
+      "loss": 1.1125,
+      "step": 7731
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28046175174288296,
+      "learning_rate": 0.00014907470884750892,
+      "loss": 1.1712,
+      "step": 7732
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2841570962434119,
+      "learning_rate": 0.00014906092467854875,
+      "loss": 1.061,
+      "step": 7733
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25589378852673406,
+      "learning_rate": 0.00014904713928180806,
+      "loss": 1.0623,
+      "step": 7734
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29195274168710955,
+      "learning_rate": 0.00014903335265763193,
+      "loss": 1.008,
+      "step": 7735
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2742605176860971,
+      "learning_rate": 0.00014901956480636535,
+      "loss": 1.153,
+      "step": 7736
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28128186302274166,
+      "learning_rate": 0.0001490057757283533,
+      "loss": 1.0106,
+      "step": 7737
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2829056481616711,
+      "learning_rate": 0.00014899198542394094,
+      "loss": 1.1351,
+      "step": 7738
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2715236434676326,
+      "learning_rate": 0.00014897819389347335,
+      "loss": 1.0404,
+      "step": 7739
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2911015197712587,
+      "learning_rate": 0.00014896440113729568,
+      "loss": 1.2098,
+      "step": 7740
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2752066845807117,
+      "learning_rate": 0.0001489506071557531,
+      "loss": 1.0378,
+      "step": 7741
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33207199501826795,
+      "learning_rate": 0.00014893681194919084,
+      "loss": 0.967,
+      "step": 7742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2838659301448611,
+      "learning_rate": 0.00014892301551795408,
+      "loss": 1.0683,
+      "step": 7743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29348859811094435,
+      "learning_rate": 0.0001489092178623881,
+      "loss": 1.1144,
+      "step": 7744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2682746013538995,
+      "learning_rate": 0.00014889541898283821,
+      "loss": 1.1194,
+      "step": 7745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30171726380766734,
+      "learning_rate": 0.00014888161887964974,
+      "loss": 1.023,
+      "step": 7746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25296517645030825,
+      "learning_rate": 0.000148867817553168,
+      "loss": 1.1727,
+      "step": 7747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31768749485918707,
+      "learning_rate": 0.00014885401500373845,
+      "loss": 1.0391,
+      "step": 7748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2438411570540729,
+      "learning_rate": 0.0001488402112317065,
+      "loss": 1.0145,
+      "step": 7749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3034976664176374,
+      "learning_rate": 0.0001488264062374175,
+      "loss": 1.1215,
+      "step": 7750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.265851976673612,
+      "learning_rate": 0.00014881260002121705,
+      "loss": 0.9724,
+      "step": 7751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29764121681730704,
+      "learning_rate": 0.00014879879258345057,
+      "loss": 1.0749,
+      "step": 7752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2797392427717897,
+      "learning_rate": 0.00014878498392446366,
+      "loss": 0.8735,
+      "step": 7753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.24366220050786846,
+      "learning_rate": 0.00014877117404460185,
+      "loss": 1.0055,
+      "step": 7754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28588196020457907,
+      "learning_rate": 0.00014875736294421078,
+      "loss": 1.16,
+      "step": 7755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.30525823027606785,
+      "learning_rate": 0.00014874355062363605,
+      "loss": 1.1435,
+      "step": 7756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.248479759938403,
+      "learning_rate": 0.00014872973708322332,
+      "loss": 1.0148,
+      "step": 7757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2705904252639747,
+      "learning_rate": 0.00014871592232331833,
+      "loss": 1.1533,
+      "step": 7758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2624337205195208,
+      "learning_rate": 0.0001487021063442667,
+      "loss": 0.9806,
+      "step": 7759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2526960220148543,
+      "learning_rate": 0.00014868828914641431,
+      "loss": 0.9377,
+      "step": 7760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2974913829902583,
+      "learning_rate": 0.00014867447073010686,
+      "loss": 1.0434,
+      "step": 7761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.22591683924408296,
+      "learning_rate": 0.0001486606510956902,
+      "loss": 1.145,
+      "step": 7762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26106572494311425,
+      "learning_rate": 0.00014864683024351017,
+      "loss": 1.0687,
+      "step": 7763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2540501633396988,
+      "learning_rate": 0.00014863300817391262,
+      "loss": 1.0217,
+      "step": 7764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2801665117949245,
+      "learning_rate": 0.0001486191848872435,
+      "loss": 1.1218,
+      "step": 7765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3187353776069642,
+      "learning_rate": 0.0001486053603838487,
+      "loss": 1.0898,
+      "step": 7766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25763576552152534,
+      "learning_rate": 0.0001485915346640742,
+      "loss": 1.0585,
+      "step": 7767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2876319448148584,
+      "learning_rate": 0.00014857770772826602,
+      "loss": 1.0169,
+      "step": 7768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25997267525577705,
+      "learning_rate": 0.0001485638795767702,
+      "loss": 1.0577,
+      "step": 7769
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2870532110656952,
+      "learning_rate": 0.00014855005020993276,
+      "loss": 1.1589,
+      "step": 7770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28744735564229734,
+      "learning_rate": 0.00014853621962809975,
+      "loss": 1.0786,
+      "step": 7771
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.26983452322850415,
+      "learning_rate": 0.0001485223878316174,
+      "loss": 1.2123,
+      "step": 7772
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.33765627577778073,
+      "learning_rate": 0.00014850855482083177,
+      "loss": 1.1044,
+      "step": 7773
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.272206284878232,
+      "learning_rate": 0.00014849472059608906,
+      "loss": 1.0212,
+      "step": 7774
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.29277937755205086,
+      "learning_rate": 0.00014848088515773553,
+      "loss": 1.1695,
+      "step": 7775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25356023491550367,
+      "learning_rate": 0.00014846704850611736,
+      "loss": 1.1737,
+      "step": 7776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3022516410233581,
+      "learning_rate": 0.00014845321064158083,
+      "loss": 1.0756,
+      "step": 7777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2530511912100064,
+      "learning_rate": 0.00014843937156447227,
+      "loss": 1.0036,
+      "step": 7778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3022412340923616,
+      "learning_rate": 0.000148425531275138,
+      "loss": 0.9716,
+      "step": 7779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2678799188134652,
+      "learning_rate": 0.00014841168977392432,
+      "loss": 1.1076,
+      "step": 7780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2963039241793785,
+      "learning_rate": 0.00014839784706117775,
+      "loss": 1.0797,
+      "step": 7781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.27673978607030375,
+      "learning_rate": 0.00014838400313724458,
+      "loss": 1.1354,
+      "step": 7782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2599969777043114,
+      "learning_rate": 0.00014837015800247137,
+      "loss": 1.0195,
+      "step": 7783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.28872142479078317,
+      "learning_rate": 0.0001483563116572045,
+      "loss": 0.9883,
+      "step": 7784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2728653682651342,
+      "learning_rate": 0.0001483424641017906,
+      "loss": 1.1095,
+      "step": 7785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3069290092163076,
+      "learning_rate": 0.00014832861533657613,
+      "loss": 1.093,
+      "step": 7786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.25616829490396015,
+      "learning_rate": 0.0001483147653619077,
+      "loss": 1.1398,
+      "step": 7787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.33842983259832143,
+      "learning_rate": 0.00014830091417813188,
+      "loss": 1.0526,
+      "step": 7788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29499869905887005,
+      "learning_rate": 0.00014828706178559534,
+      "loss": 1.1995,
+      "step": 7789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2867836759486996,
+      "learning_rate": 0.00014827320818464474,
+      "loss": 1.1038,
+      "step": 7790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27597630873723183,
+      "learning_rate": 0.00014825935337562673,
+      "loss": 1.1934,
+      "step": 7791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2566617689778122,
+      "learning_rate": 0.0001482454973588881,
+      "loss": 1.0383,
+      "step": 7792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28408572266180726,
+      "learning_rate": 0.0001482316401347756,
+      "loss": 0.9326,
+      "step": 7793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2745042716972083,
+      "learning_rate": 0.00014821778170363595,
+      "loss": 1.0336,
+      "step": 7794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3032701457073397,
+      "learning_rate": 0.00014820392206581602,
+      "loss": 1.0234,
+      "step": 7795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2751280009313665,
+      "learning_rate": 0.00014819006122166267,
+      "loss": 1.0432,
+      "step": 7796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3125193076347104,
+      "learning_rate": 0.00014817619917152275,
+      "loss": 1.1172,
+      "step": 7797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26116307629794683,
+      "learning_rate": 0.00014816233591574313,
+      "loss": 1.0669,
+      "step": 7798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28349617158355267,
+      "learning_rate": 0.0001481484714546708,
+      "loss": 1.1477,
+      "step": 7799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.34282406819344863,
+      "learning_rate": 0.00014813460578865274,
+      "loss": 1.009,
+      "step": 7800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28578123221740864,
+      "learning_rate": 0.00014812073891803587,
+      "loss": 1.156,
+      "step": 7801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2980354742362817,
+      "learning_rate": 0.0001481068708431673,
+      "loss": 1.0768,
+      "step": 7802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3220759599695297,
+      "learning_rate": 0.00014809300156439406,
+      "loss": 1.0764,
+      "step": 7803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2640835765034414,
+      "learning_rate": 0.00014807913108206322,
+      "loss": 1.0768,
+      "step": 7804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27220178093073455,
+      "learning_rate": 0.00014806525939652188,
+      "loss": 1.1449,
+      "step": 7805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27363838878553143,
+      "learning_rate": 0.00014805138650811724,
+      "loss": 1.0526,
+      "step": 7806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2883264760303332,
+      "learning_rate": 0.0001480375124171965,
+      "loss": 1.0072,
+      "step": 7807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28230583793684294,
+      "learning_rate": 0.00014802363712410673,
+      "loss": 1.043,
+      "step": 7808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2615628239719908,
+      "learning_rate": 0.00014800976062919532,
+      "loss": 1.0759,
+      "step": 7809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29229275325827664,
+      "learning_rate": 0.00014799588293280946,
+      "loss": 1.0116,
+      "step": 7810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2738661601965768,
+      "learning_rate": 0.00014798200403529646,
+      "loss": 1.1214,
+      "step": 7811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2743653194914421,
+      "learning_rate": 0.00014796812393700368,
+      "loss": 1.0277,
+      "step": 7812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27052008019631224,
+      "learning_rate": 0.00014795424263827842,
+      "loss": 1.0564,
+      "step": 7813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30231813442957706,
+      "learning_rate": 0.00014794036013946813,
+      "loss": 1.1308,
+      "step": 7814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29045069706562976,
+      "learning_rate": 0.00014792647644092016,
+      "loss": 1.0502,
+      "step": 7815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2663099880383512,
+      "learning_rate": 0.000147912591542982,
+      "loss": 1.1342,
+      "step": 7816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2671935344264389,
+      "learning_rate": 0.00014789870544600116,
+      "loss": 1.0337,
+      "step": 7817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2376028787105682,
+      "learning_rate": 0.00014788481815032509,
+      "loss": 1.0951,
+      "step": 7818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2873919541494794,
+      "learning_rate": 0.00014787092965630135,
+      "loss": 1.0885,
+      "step": 7819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28326748757574954,
+      "learning_rate": 0.00014785703996427754,
+      "loss": 0.9765,
+      "step": 7820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26333898401252714,
+      "learning_rate": 0.0001478431490746012,
+      "loss": 0.9915,
+      "step": 7821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2638238903309792,
+      "learning_rate": 0.00014782925698761997,
+      "loss": 1.0908,
+      "step": 7822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30274071491597526,
+      "learning_rate": 0.00014781536370368157,
+      "loss": 1.0742,
+      "step": 7823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2610374416233597,
+      "learning_rate": 0.0001478014692231336,
+      "loss": 1.1589,
+      "step": 7824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.312177588979833,
+      "learning_rate": 0.00014778757354632382,
+      "loss": 1.023,
+      "step": 7825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2817902834771327,
+      "learning_rate": 0.00014777367667360002,
+      "loss": 1.0925,
+      "step": 7826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24540218148082688,
+      "learning_rate": 0.00014775977860530988,
+      "loss": 0.9525,
+      "step": 7827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.292375111038645,
+      "learning_rate": 0.0001477458793418013,
+      "loss": 1.0295,
+      "step": 7828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25821078004181164,
+      "learning_rate": 0.0001477319788834221,
+      "loss": 1.0538,
+      "step": 7829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26371249547752684,
+      "learning_rate": 0.00014771807723052013,
+      "loss": 1.0396,
+      "step": 7830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24548178512781826,
+      "learning_rate": 0.00014770417438344325,
+      "loss": 1.1311,
+      "step": 7831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28303419109838207,
+      "learning_rate": 0.00014769027034253944,
+      "loss": 1.0892,
+      "step": 7832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.27687609304875627,
+      "learning_rate": 0.00014767636510815667,
+      "loss": 1.0869,
+      "step": 7833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28746479888055076,
+      "learning_rate": 0.00014766245868064285,
+      "loss": 1.1869,
+      "step": 7834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2879143036391422,
+      "learning_rate": 0.00014764855106034607,
+      "loss": 1.0875,
+      "step": 7835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2653922351709626,
+      "learning_rate": 0.00014763464224761436,
+      "loss": 1.1238,
+      "step": 7836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2748579199210656,
+      "learning_rate": 0.00014762073224279578,
+      "loss": 1.0366,
+      "step": 7837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29359776941173216,
+      "learning_rate": 0.00014760682104623845,
+      "loss": 0.9695,
+      "step": 7838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2671398097358774,
+      "learning_rate": 0.00014759290865829053,
+      "loss": 1.0089,
+      "step": 7839
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.1271681785583496,
+      "eval_runtime": 4227.9142,
+      "eval_samples_per_second": 19.778,
+      "eval_steps_per_second": 2.472,
+      "step": 7839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28108668366507505,
+      "learning_rate": 0.00014757899507930012,
+      "loss": 1.1518,
+      "step": 7840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2889156834394568,
+      "learning_rate": 0.00014756508030961543,
+      "loss": 1.057,
+      "step": 7841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.23620442034659647,
+      "learning_rate": 0.00014755116434958477,
+      "loss": 1.0139,
+      "step": 7842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26963408436529646,
+      "learning_rate": 0.00014753724719955634,
+      "loss": 1.2173,
+      "step": 7843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2930307296639112,
+      "learning_rate": 0.0001475233288598784,
+      "loss": 1.1191,
+      "step": 7844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26447900586343337,
+      "learning_rate": 0.00014750940933089927,
+      "loss": 1.0972,
+      "step": 7845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28229996008223884,
+      "learning_rate": 0.00014749548861296734,
+      "loss": 1.1229,
+      "step": 7846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2670260276637567,
+      "learning_rate": 0.00014748156670643097,
+      "loss": 1.2654,
+      "step": 7847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29233732430571957,
+      "learning_rate": 0.00014746764361163854,
+      "loss": 1.1009,
+      "step": 7848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2872352622058822,
+      "learning_rate": 0.00014745371932893848,
+      "loss": 1.0123,
+      "step": 7849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26320292513113624,
+      "learning_rate": 0.00014743979385867928,
+      "loss": 1.056,
+      "step": 7850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29353850314037555,
+      "learning_rate": 0.00014742586720120943,
+      "loss": 1.0711,
+      "step": 7851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.24643873017384735,
+      "learning_rate": 0.00014741193935687743,
+      "loss": 0.942,
+      "step": 7852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30839731486986416,
+      "learning_rate": 0.00014739801032603186,
+      "loss": 0.932,
+      "step": 7853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.25565523985069327,
+      "learning_rate": 0.0001473840801090213,
+      "loss": 1.0941,
+      "step": 7854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26831609426984876,
+      "learning_rate": 0.00014737014870619438,
+      "loss": 1.079,
+      "step": 7855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2782662024037316,
+      "learning_rate": 0.00014735621611789963,
+      "loss": 1.085,
+      "step": 7856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29723517036293495,
+      "learning_rate": 0.0001473422823444859,
+      "loss": 1.0157,
+      "step": 7857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26674243695427563,
+      "learning_rate": 0.00014732834738630178,
+      "loss": 1.0748,
+      "step": 7858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3178208105289837,
+      "learning_rate": 0.00014731441124369598,
+      "loss": 1.047,
+      "step": 7859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3797982501806393,
+      "learning_rate": 0.00014730047391701737,
+      "loss": 0.9933,
+      "step": 7860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2901368038846893,
+      "learning_rate": 0.00014728653540661463,
+      "loss": 1.0821,
+      "step": 7861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2955915921047405,
+      "learning_rate": 0.0001472725957128366,
+      "loss": 0.9773,
+      "step": 7862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29647306511427657,
+      "learning_rate": 0.00014725865483603218,
+      "loss": 1.1752,
+      "step": 7863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2918367246052348,
+      "learning_rate": 0.00014724471277655023,
+      "loss": 0.9763,
+      "step": 7864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2815720894283948,
+      "learning_rate": 0.00014723076953473965,
+      "loss": 1.098,
+      "step": 7865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2930097766676262,
+      "learning_rate": 0.00014721682511094935,
+      "loss": 1.0655,
+      "step": 7866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29679633835988606,
+      "learning_rate": 0.00014720287950552836,
+      "loss": 1.103,
+      "step": 7867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29540666426295414,
+      "learning_rate": 0.00014718893271882562,
+      "loss": 1.1377,
+      "step": 7868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28866732431657105,
+      "learning_rate": 0.0001471749847511902,
+      "loss": 1.1552,
+      "step": 7869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2897250322336478,
+      "learning_rate": 0.00014716103560297116,
+      "loss": 1.1178,
+      "step": 7870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2871031890874088,
+      "learning_rate": 0.00014714708527451752,
+      "loss": 1.0608,
+      "step": 7871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2660012537881913,
+      "learning_rate": 0.00014713313376617845,
+      "loss": 0.9785,
+      "step": 7872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30288071506941316,
+      "learning_rate": 0.00014711918107830312,
+      "loss": 1.0374,
+      "step": 7873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2623082689214895,
+      "learning_rate": 0.00014710522721124062,
+      "loss": 1.1882,
+      "step": 7874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29218912698161004,
+      "learning_rate": 0.00014709127216534025,
+      "loss": 1.1121,
+      "step": 7875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2926792107209431,
+      "learning_rate": 0.00014707731594095118,
+      "loss": 1.1558,
+      "step": 7876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3036786630685544,
+      "learning_rate": 0.00014706335853842268,
+      "loss": 1.1285,
+      "step": 7877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3070257247499537,
+      "learning_rate": 0.00014704939995810408,
+      "loss": 1.1566,
+      "step": 7878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2753834785951813,
+      "learning_rate": 0.00014703544020034463,
+      "loss": 1.1336,
+      "step": 7879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30835133105536705,
+      "learning_rate": 0.00014702147926549377,
+      "loss": 1.0733,
+      "step": 7880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.30129377981823663,
+      "learning_rate": 0.00014700751715390084,
+      "loss": 1.1398,
+      "step": 7881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.26568110306618115,
+      "learning_rate": 0.00014699355386591523,
+      "loss": 1.0773,
+      "step": 7882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28052397758769976,
+      "learning_rate": 0.00014697958940188642,
+      "loss": 1.0731,
+      "step": 7883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2696382620489903,
+      "learning_rate": 0.00014696562376216381,
+      "loss": 0.9844,
+      "step": 7884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2532216857709228,
+      "learning_rate": 0.000146951656947097,
+      "loss": 1.0863,
+      "step": 7885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2622266542497967,
+      "learning_rate": 0.00014693768895703544,
+      "loss": 1.0305,
+      "step": 7886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2800195946028661,
+      "learning_rate": 0.00014692371979232872,
+      "loss": 1.1075,
+      "step": 7887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2953080493463318,
+      "learning_rate": 0.00014690974945332644,
+      "loss": 0.9948,
+      "step": 7888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2886266231002391,
+      "learning_rate": 0.00014689577794037818,
+      "loss": 0.9764,
+      "step": 7889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29706661719229843,
+      "learning_rate": 0.0001468818052538336,
+      "loss": 1.055,
+      "step": 7890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.28179682965161795,
+      "learning_rate": 0.0001468678313940424,
+      "loss": 1.0903,
+      "step": 7891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3030242128434006,
+      "learning_rate": 0.00014685385636135426,
+      "loss": 1.1486,
+      "step": 7892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2647061302202202,
+      "learning_rate": 0.00014683988015611892,
+      "loss": 1.0548,
+      "step": 7893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.300733623893973,
+      "learning_rate": 0.00014682590277868612,
+      "loss": 1.1367,
+      "step": 7894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2995793418222508,
+      "learning_rate": 0.0001468119242294057,
+      "loss": 1.0878,
+      "step": 7895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28708772163438406,
+      "learning_rate": 0.00014679794450862745,
+      "loss": 0.9396,
+      "step": 7896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.311063842263335,
+      "learning_rate": 0.0001467839636167012,
+      "loss": 1.1202,
+      "step": 7897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2986192805747378,
+      "learning_rate": 0.0001467699815539769,
+      "loss": 1.0607,
+      "step": 7898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.258096740209968,
+      "learning_rate": 0.0001467559983208044,
+      "loss": 0.9521,
+      "step": 7899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27283666156818637,
+      "learning_rate": 0.00014674201391753363,
+      "loss": 1.1342,
+      "step": 7900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28773671837545195,
+      "learning_rate": 0.0001467280283445146,
+      "loss": 1.1128,
+      "step": 7901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.307556564712858,
+      "learning_rate": 0.00014671404160209733,
+      "loss": 0.9963,
+      "step": 7902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2684719138920676,
+      "learning_rate": 0.0001467000536906318,
+      "loss": 1.0201,
+      "step": 7903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26240667501651854,
+      "learning_rate": 0.00014668606461046806,
+      "loss": 1.0589,
+      "step": 7904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27922442657332547,
+      "learning_rate": 0.0001466720743619562,
+      "loss": 1.0683,
+      "step": 7905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2540084140712392,
+      "learning_rate": 0.00014665808294544633,
+      "loss": 0.9643,
+      "step": 7906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31142085724795016,
+      "learning_rate": 0.00014664409036128866,
+      "loss": 1.045,
+      "step": 7907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3206848269424768,
+      "learning_rate": 0.00014663009660983328,
+      "loss": 1.1147,
+      "step": 7908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.33340917807727916,
+      "learning_rate": 0.00014661610169143044,
+      "loss": 1.0744,
+      "step": 7909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32071991522507715,
+      "learning_rate": 0.00014660210560643036,
+      "loss": 1.1297,
+      "step": 7910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2761333180916678,
+      "learning_rate": 0.00014658810835518332,
+      "loss": 1.0477,
+      "step": 7911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2859125633527157,
+      "learning_rate": 0.00014657410993803956,
+      "loss": 1.0781,
+      "step": 7912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3274797315976737,
+      "learning_rate": 0.00014656011035534943,
+      "loss": 1.0144,
+      "step": 7913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28598470593343445,
+      "learning_rate": 0.00014654610960746327,
+      "loss": 1.0887,
+      "step": 7914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28926204149947815,
+      "learning_rate": 0.00014653210769473147,
+      "loss": 1.0627,
+      "step": 7915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26324720206303615,
+      "learning_rate": 0.00014651810461750446,
+      "loss": 1.1053,
+      "step": 7916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30675528347879105,
+      "learning_rate": 0.0001465041003761326,
+      "loss": 1.1054,
+      "step": 7917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26066528416246476,
+      "learning_rate": 0.0001464900949709664,
+      "loss": 1.0405,
+      "step": 7918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29329946245835375,
+      "learning_rate": 0.0001464760884023564,
+      "loss": 1.0615,
+      "step": 7919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2679473490528041,
+      "learning_rate": 0.00014646208067065305,
+      "loss": 1.0988,
+      "step": 7920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29866219772813785,
+      "learning_rate": 0.00014644807177620694,
+      "loss": 1.1193,
+      "step": 7921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3004083603649936,
+      "learning_rate": 0.00014643406171936863,
+      "loss": 1.0458,
+      "step": 7922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3255922513505334,
+      "learning_rate": 0.00014642005050048877,
+      "loss": 1.1641,
+      "step": 7923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27331269973566386,
+      "learning_rate": 0.00014640603811991794,
+      "loss": 1.0904,
+      "step": 7924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3021895089941293,
+      "learning_rate": 0.00014639202457800688,
+      "loss": 1.0426,
+      "step": 7925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31651111587188924,
+      "learning_rate": 0.0001463780098751062,
+      "loss": 0.9561,
+      "step": 7926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27245193343641233,
+      "learning_rate": 0.00014636399401156668,
+      "loss": 0.9084,
+      "step": 7927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28171140443893944,
+      "learning_rate": 0.0001463499769877391,
+      "loss": 1.0122,
+      "step": 7928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2722172920609029,
+      "learning_rate": 0.00014633595880397422,
+      "loss": 0.9773,
+      "step": 7929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.314701712266059,
+      "learning_rate": 0.00014632193946062283,
+      "loss": 1.0967,
+      "step": 7930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2766120012080896,
+      "learning_rate": 0.0001463079189580358,
+      "loss": 1.1545,
+      "step": 7931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26393567149311037,
+      "learning_rate": 0.00014629389729656399,
+      "loss": 0.8912,
+      "step": 7932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30280952487557516,
+      "learning_rate": 0.0001462798744765583,
+      "loss": 1.1161,
+      "step": 7933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31616202197131665,
+      "learning_rate": 0.0001462658504983697,
+      "loss": 1.179,
+      "step": 7934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26565495442579756,
+      "learning_rate": 0.0001462518253623491,
+      "loss": 1.0568,
+      "step": 7935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.318280638548035,
+      "learning_rate": 0.00014623779906884748,
+      "loss": 1.1108,
+      "step": 7936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2586299065330331,
+      "learning_rate": 0.00014622377161821587,
+      "loss": 1.148,
+      "step": 7937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2715217059364236,
+      "learning_rate": 0.00014620974301080537,
+      "loss": 1.0559,
+      "step": 7938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24985474931051407,
+      "learning_rate": 0.00014619571324696697,
+      "loss": 1.1047,
+      "step": 7939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2961624550424931,
+      "learning_rate": 0.00014618168232705182,
+      "loss": 0.9659,
+      "step": 7940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2966955141929237,
+      "learning_rate": 0.00014616765025141106,
+      "loss": 1.0266,
+      "step": 7941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2986341505874412,
+      "learning_rate": 0.00014615361702039582,
+      "loss": 1.0165,
+      "step": 7942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.291306108234052,
+      "learning_rate": 0.00014613958263435734,
+      "loss": 1.0463,
+      "step": 7943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24770026210630766,
+      "learning_rate": 0.00014612554709364677,
+      "loss": 1.0362,
+      "step": 7944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30727250672800005,
+      "learning_rate": 0.00014611151039861542,
+      "loss": 1.0903,
+      "step": 7945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30656468148580795,
+      "learning_rate": 0.00014609747254961452,
+      "loss": 1.0681,
+      "step": 7946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28351015720243533,
+      "learning_rate": 0.0001460834335469954,
+      "loss": 1.166,
+      "step": 7947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28579851887764157,
+      "learning_rate": 0.0001460693933911094,
+      "loss": 1.0082,
+      "step": 7948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26715615685634764,
+      "learning_rate": 0.00014605535208230789,
+      "loss": 1.1272,
+      "step": 7949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2908957018166822,
+      "learning_rate": 0.0001460413096209422,
+      "loss": 1.0386,
+      "step": 7950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30421751423231963,
+      "learning_rate": 0.00014602726600736388,
+      "loss": 1.1368,
+      "step": 7951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24654599232461855,
+      "learning_rate": 0.00014601322124192426,
+      "loss": 1.0704,
+      "step": 7952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2858412396022753,
+      "learning_rate": 0.00014599917532497487,
+      "loss": 0.9629,
+      "step": 7953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31986567150353395,
+      "learning_rate": 0.00014598512825686718,
+      "loss": 1.1515,
+      "step": 7954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26767724316265096,
+      "learning_rate": 0.0001459710800379528,
+      "loss": 1.0711,
+      "step": 7955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2683093058976096,
+      "learning_rate": 0.0001459570306685832,
+      "loss": 0.9834,
+      "step": 7956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26418718399717656,
+      "learning_rate": 0.00014594298014911005,
+      "loss": 1.1891,
+      "step": 7957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2997490177491894,
+      "learning_rate": 0.00014592892847988494,
+      "loss": 1.1303,
+      "step": 7958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3092712454899324,
+      "learning_rate": 0.00014591487566125957,
+      "loss": 1.0605,
+      "step": 7959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.270510283984863,
+      "learning_rate": 0.00014590082169358554,
+      "loss": 1.0886,
+      "step": 7960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.24910690948882463,
+      "learning_rate": 0.0001458867665772146,
+      "loss": 1.0897,
+      "step": 7961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29126461074740784,
+      "learning_rate": 0.0001458727103124985,
+      "loss": 1.1572,
+      "step": 7962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32863806873359025,
+      "learning_rate": 0.000145858652899789,
+      "loss": 1.0036,
+      "step": 7963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27115363531745906,
+      "learning_rate": 0.00014584459433943786,
+      "loss": 1.0609,
+      "step": 7964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26632334307016087,
+      "learning_rate": 0.00014583053463179695,
+      "loss": 0.929,
+      "step": 7965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30896648642067515,
+      "learning_rate": 0.00014581647377721812,
+      "loss": 1.1302,
+      "step": 7966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31271610841373415,
+      "learning_rate": 0.00014580241177605322,
+      "loss": 1.0555,
+      "step": 7967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2862084572430668,
+      "learning_rate": 0.0001457883486286542,
+      "loss": 1.1462,
+      "step": 7968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2722638712262504,
+      "learning_rate": 0.00014577428433537297,
+      "loss": 1.0671,
+      "step": 7969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29776545230446366,
+      "learning_rate": 0.0001457602188965615,
+      "loss": 1.0892,
+      "step": 7970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28339621303738605,
+      "learning_rate": 0.00014574615231257177,
+      "loss": 1.0042,
+      "step": 7971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2872649900886519,
+      "learning_rate": 0.00014573208458375586,
+      "loss": 1.0962,
+      "step": 7972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29073436106492967,
+      "learning_rate": 0.0001457180157104658,
+      "loss": 1.096,
+      "step": 7973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2782075864070279,
+      "learning_rate": 0.00014570394569305366,
+      "loss": 1.0278,
+      "step": 7974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31350190862970956,
+      "learning_rate": 0.00014568987453187154,
+      "loss": 1.091,
+      "step": 7975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2652268268602932,
+      "learning_rate": 0.0001456758022272716,
+      "loss": 1.128,
+      "step": 7976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2810839334912306,
+      "learning_rate": 0.00014566172877960603,
+      "loss": 1.0408,
+      "step": 7977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.25013935049819563,
+      "learning_rate": 0.00014564765418922696,
+      "loss": 1.1124,
+      "step": 7978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.23668034177376465,
+      "learning_rate": 0.00014563357845648667,
+      "loss": 1.028,
+      "step": 7979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2736819717949704,
+      "learning_rate": 0.0001456195015817374,
+      "loss": 0.9969,
+      "step": 7980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2771456339595817,
+      "learning_rate": 0.00014560542356533142,
+      "loss": 1.1049,
+      "step": 7981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.294782138767455,
+      "learning_rate": 0.00014559134440762108,
+      "loss": 1.13,
+      "step": 7982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26345886128180046,
+      "learning_rate": 0.0001455772641089587,
+      "loss": 1.1155,
+      "step": 7983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2829069470535079,
+      "learning_rate": 0.00014556318266969656,
+      "loss": 1.0918,
+      "step": 7984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28542191746861517,
+      "learning_rate": 0.00014554910009018722,
+      "loss": 1.0391,
+      "step": 7985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2538797891395937,
+      "learning_rate": 0.000145535016370783,
+      "loss": 0.9585,
+      "step": 7986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.26884615169189024,
+      "learning_rate": 0.0001455209315118364,
+      "loss": 1.0022,
+      "step": 7987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2716483059032336,
+      "learning_rate": 0.00014550684551369985,
+      "loss": 1.0702,
+      "step": 7988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.27843349775471815,
+      "learning_rate": 0.00014549275837672586,
+      "loss": 1.1401,
+      "step": 7989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2518481347111056,
+      "learning_rate": 0.00014547867010126706,
+      "loss": 1.0111,
+      "step": 7990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3216392004906807,
+      "learning_rate": 0.00014546458068767594,
+      "loss": 1.0472,
+      "step": 7991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28909463374066535,
+      "learning_rate": 0.00014545049013630512,
+      "loss": 1.0201,
+      "step": 7992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.31788442875623607,
+      "learning_rate": 0.0001454363984475072,
+      "loss": 1.0955,
+      "step": 7993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2758373823123496,
+      "learning_rate": 0.00014542230562163488,
+      "loss": 1.0873,
+      "step": 7994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2818308726562138,
+      "learning_rate": 0.0001454082116590408,
+      "loss": 1.2114,
+      "step": 7995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2641986472531651,
+      "learning_rate": 0.0001453941165600777,
+      "loss": 1.1287,
+      "step": 7996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27941463601912914,
+      "learning_rate": 0.0001453800203250983,
+      "loss": 1.0574,
+      "step": 7997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2756789148686793,
+      "learning_rate": 0.00014536592295445532,
+      "loss": 0.9556,
+      "step": 7998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27533489457097576,
+      "learning_rate": 0.00014535182444850165,
+      "loss": 1.1021,
+      "step": 7999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2964512342749194,
+      "learning_rate": 0.00014533772480759008,
+      "loss": 1.0421,
+      "step": 8000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26212904131408093,
+      "learning_rate": 0.00014532362403207346,
+      "loss": 1.0015,
+      "step": 8001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2949190595576002,
+      "learning_rate": 0.00014530952212230463,
+      "loss": 1.0511,
+      "step": 8002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3066540690434103,
+      "learning_rate": 0.00014529541907863655,
+      "loss": 1.0643,
+      "step": 8003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27584654577393586,
+      "learning_rate": 0.00014528131490142217,
+      "loss": 0.9773,
+      "step": 8004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27774031420621953,
+      "learning_rate": 0.00014526720959101436,
+      "loss": 0.9925,
+      "step": 8005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2512278740693556,
+      "learning_rate": 0.00014525310314776623,
+      "loss": 1.0008,
+      "step": 8006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2630439424936692,
+      "learning_rate": 0.00014523899557203075,
+      "loss": 1.1098,
+      "step": 8007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27426353058067837,
+      "learning_rate": 0.00014522488686416097,
+      "loss": 1.0728,
+      "step": 8008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2763756735231871,
+      "learning_rate": 0.00014521077702450995,
+      "loss": 1.0639,
+      "step": 8009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3325734526066199,
+      "learning_rate": 0.00014519666605343083,
+      "loss": 1.0027,
+      "step": 8010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26469084339952254,
+      "learning_rate": 0.00014518255395127677,
+      "loss": 1.0773,
+      "step": 8011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2755841427968526,
+      "learning_rate": 0.00014516844071840086,
+      "loss": 1.118,
+      "step": 8012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2970488079219033,
+      "learning_rate": 0.00014515432635515635,
+      "loss": 1.0362,
+      "step": 8013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2954323242845108,
+      "learning_rate": 0.00014514021086189645,
+      "loss": 1.0746,
+      "step": 8014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2617764506782068,
+      "learning_rate": 0.00014512609423897438,
+      "loss": 1.0385,
+      "step": 8015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30127357515303343,
+      "learning_rate": 0.00014511197648674348,
+      "loss": 1.0966,
+      "step": 8016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3004406925080267,
+      "learning_rate": 0.00014509785760555697,
+      "loss": 1.0441,
+      "step": 8017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25228201013427193,
+      "learning_rate": 0.00014508373759576824,
+      "loss": 1.0692,
+      "step": 8018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3052520210235085,
+      "learning_rate": 0.00014506961645773068,
+      "loss": 1.1513,
+      "step": 8019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2709522809700183,
+      "learning_rate": 0.00014505549419179765,
+      "loss": 0.991,
+      "step": 8020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26525464351023736,
+      "learning_rate": 0.00014504137079832252,
+      "loss": 0.8964,
+      "step": 8021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2877907446668608,
+      "learning_rate": 0.00014502724627765877,
+      "loss": 1.0461,
+      "step": 8022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29764029690496574,
+      "learning_rate": 0.00014501312063015993,
+      "loss": 1.0284,
+      "step": 8023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2883370401645907,
+      "learning_rate": 0.00014499899385617943,
+      "loss": 1.0268,
+      "step": 8024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2563877310771201,
+      "learning_rate": 0.0001449848659560708,
+      "loss": 1.1029,
+      "step": 8025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29710595080293883,
+      "learning_rate": 0.00014497073693018768,
+      "loss": 1.1315,
+      "step": 8026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2559529397106755,
+      "learning_rate": 0.00014495660677888358,
+      "loss": 1.009,
+      "step": 8027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29014013639604846,
+      "learning_rate": 0.00014494247550251213,
+      "loss": 1.0681,
+      "step": 8028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2875485426750669,
+      "learning_rate": 0.00014492834310142702,
+      "loss": 1.1195,
+      "step": 8029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.286722769874282,
+      "learning_rate": 0.00014491420957598184,
+      "loss": 1.0529,
+      "step": 8030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2674676429533566,
+      "learning_rate": 0.0001449000749265304,
+      "loss": 1.0475,
+      "step": 8031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29840922095847167,
+      "learning_rate": 0.00014488593915342628,
+      "loss": 1.123,
+      "step": 8032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29901175240287103,
+      "learning_rate": 0.0001448718022570234,
+      "loss": 1.148,
+      "step": 8033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2915382990881953,
+      "learning_rate": 0.00014485766423767544,
+      "loss": 1.0401,
+      "step": 8034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26281760733012527,
+      "learning_rate": 0.00014484352509573626,
+      "loss": 1.1292,
+      "step": 8035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2756692483739733,
+      "learning_rate": 0.00014482938483155965,
+      "loss": 1.0544,
+      "step": 8036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28042375304828776,
+      "learning_rate": 0.00014481524344549953,
+      "loss": 1.1094,
+      "step": 8037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2666048364114282,
+      "learning_rate": 0.00014480110093790976,
+      "loss": 1.0748,
+      "step": 8038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5049686356926107,
+      "learning_rate": 0.0001447869573091443,
+      "loss": 1.0392,
+      "step": 8039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2898604153674101,
+      "learning_rate": 0.0001447728125595571,
+      "loss": 1.0527,
+      "step": 8040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2938423642503213,
+      "learning_rate": 0.0001447586666895021,
+      "loss": 1.0593,
+      "step": 8041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3199560544400549,
+      "learning_rate": 0.00014474451969933333,
+      "loss": 1.1253,
+      "step": 8042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.279058551647216,
+      "learning_rate": 0.00014473037158940484,
+      "loss": 1.0904,
+      "step": 8043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3317633487011477,
+      "learning_rate": 0.0001447162223600707,
+      "loss": 1.0621,
+      "step": 8044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31113174621821804,
+      "learning_rate": 0.00014470207201168497,
+      "loss": 1.1192,
+      "step": 8045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2732527847810805,
+      "learning_rate": 0.00014468792054460184,
+      "loss": 1.0248,
+      "step": 8046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29153439327487507,
+      "learning_rate": 0.00014467376795917537,
+      "loss": 1.124,
+      "step": 8047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2860523862518081,
+      "learning_rate": 0.0001446596142557598,
+      "loss": 1.0621,
+      "step": 8048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2862949081963216,
+      "learning_rate": 0.00014464545943470932,
+      "loss": 1.0017,
+      "step": 8049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2968576640237807,
+      "learning_rate": 0.00014463130349637814,
+      "loss": 1.0521,
+      "step": 8050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3203158707213885,
+      "learning_rate": 0.00014461714644112053,
+      "loss": 1.0283,
+      "step": 8051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.23962634495433535,
+      "learning_rate": 0.0001446029882692908,
+      "loss": 1.0052,
+      "step": 8052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29338455713698386,
+      "learning_rate": 0.0001445888289812433,
+      "loss": 1.1304,
+      "step": 8053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2650099607677807,
+      "learning_rate": 0.0001445746685773323,
+      "loss": 0.9916,
+      "step": 8054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.26198071838847975,
+      "learning_rate": 0.00014456050705791216,
+      "loss": 1.0875,
+      "step": 8055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.292914913997817,
+      "learning_rate": 0.00014454634442333738,
+      "loss": 1.0666,
+      "step": 8056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3015285693686079,
+      "learning_rate": 0.00014453218067396231,
+      "loss": 1.1363,
+      "step": 8057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30434995117078845,
+      "learning_rate": 0.00014451801581014147,
+      "loss": 1.0074,
+      "step": 8058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30572536430411557,
+      "learning_rate": 0.00014450384983222926,
+      "loss": 1.0941,
+      "step": 8059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2665325169928879,
+      "learning_rate": 0.00014448968274058025,
+      "loss": 1.0584,
+      "step": 8060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.291238529425458,
+      "learning_rate": 0.000144475514535549,
+      "loss": 0.9758,
+      "step": 8061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.320923065004721,
+      "learning_rate": 0.00014446134521749,
+      "loss": 0.9887,
+      "step": 8062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2695062501972563,
+      "learning_rate": 0.00014444717478675792,
+      "loss": 1.063,
+      "step": 8063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29036294842452276,
+      "learning_rate": 0.00014443300324370738,
+      "loss": 1.0825,
+      "step": 8064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2667481516709462,
+      "learning_rate": 0.00014441883058869298,
+      "loss": 1.0139,
+      "step": 8065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31330268446835163,
+      "learning_rate": 0.00014440465682206944,
+      "loss": 1.1519,
+      "step": 8066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29426667401104506,
+      "learning_rate": 0.00014439048194419141,
+      "loss": 1.001,
+      "step": 8067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30892190035705414,
+      "learning_rate": 0.00014437630595541374,
+      "loss": 1.0714,
+      "step": 8068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28625371937125405,
+      "learning_rate": 0.00014436212885609106,
+      "loss": 1.1328,
+      "step": 8069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2904730209899161,
+      "learning_rate": 0.00014434795064657827,
+      "loss": 0.9807,
+      "step": 8070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29946573442663216,
+      "learning_rate": 0.0001443337713272301,
+      "loss": 1.108,
+      "step": 8071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.25138650851267935,
+      "learning_rate": 0.0001443195908984015,
+      "loss": 0.9354,
+      "step": 8072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3046129695319591,
+      "learning_rate": 0.00014430540936044724,
+      "loss": 1.0464,
+      "step": 8073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2815772869608019,
+      "learning_rate": 0.0001442912267137223,
+      "loss": 1.1689,
+      "step": 8074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28788534180895897,
+      "learning_rate": 0.00014427704295858154,
+      "loss": 1.1042,
+      "step": 8075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28620999022686416,
+      "learning_rate": 0.00014426285809537997,
+      "loss": 1.0644,
+      "step": 8076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2718948107560663,
+      "learning_rate": 0.00014424867212447254,
+      "loss": 1.0555,
+      "step": 8077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2745313908908246,
+      "learning_rate": 0.0001442344850462143,
+      "loss": 0.953,
+      "step": 8078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2782122491457099,
+      "learning_rate": 0.0001442202968609603,
+      "loss": 1.098,
+      "step": 8079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27832697424885927,
+      "learning_rate": 0.00014420610756906552,
+      "loss": 0.9896,
+      "step": 8080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2850630781424915,
+      "learning_rate": 0.00014419191717088517,
+      "loss": 1.0579,
+      "step": 8081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30698153794733524,
+      "learning_rate": 0.00014417772566677428,
+      "loss": 0.9522,
+      "step": 8082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2412252387912062,
+      "learning_rate": 0.00014416353305708802,
+      "loss": 0.963,
+      "step": 8083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2760933371029509,
+      "learning_rate": 0.00014414933934218165,
+      "loss": 1.039,
+      "step": 8084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27478977307311525,
+      "learning_rate": 0.0001441351445224103,
+      "loss": 1.0191,
+      "step": 8085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27460908719231536,
+      "learning_rate": 0.0001441209485981292,
+      "loss": 0.942,
+      "step": 8086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.300097647062556,
+      "learning_rate": 0.00014410675156969362,
+      "loss": 1.1083,
+      "step": 8087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.301434940986813,
+      "learning_rate": 0.0001440925534374589,
+      "loss": 1.1201,
+      "step": 8088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34304082819084863,
+      "learning_rate": 0.00014407835420178028,
+      "loss": 1.2234,
+      "step": 8089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.30266233132954434,
+      "learning_rate": 0.00014406415386301319,
+      "loss": 1.0731,
+      "step": 8090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29052074003899026,
+      "learning_rate": 0.00014404995242151293,
+      "loss": 1.0351,
+      "step": 8091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31236829856827036,
+      "learning_rate": 0.00014403574987763493,
+      "loss": 1.1238,
+      "step": 8092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27979550559379457,
+      "learning_rate": 0.0001440215462317346,
+      "loss": 1.142,
+      "step": 8093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.303167287374647,
+      "learning_rate": 0.00014400734148416742,
+      "loss": 1.1546,
+      "step": 8094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2722206268311316,
+      "learning_rate": 0.00014399313563528886,
+      "loss": 1.0722,
+      "step": 8095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29950103549697416,
+      "learning_rate": 0.00014397892868545442,
+      "loss": 1.0869,
+      "step": 8096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.29083600055293385,
+      "learning_rate": 0.00014396472063501968,
+      "loss": 1.0576,
+      "step": 8097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2883189452608079,
+      "learning_rate": 0.00014395051148434015,
+      "loss": 0.9811,
+      "step": 8098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2614201885833928,
+      "learning_rate": 0.0001439363012337715,
+      "loss": 1.0721,
+      "step": 8099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.27854811920048617,
+      "learning_rate": 0.00014392208988366921,
+      "loss": 1.0133,
+      "step": 8100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.33867360910508465,
+      "learning_rate": 0.00014390787743438907,
+      "loss": 1.0265,
+      "step": 8101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27479933367603354,
+      "learning_rate": 0.0001438936638862867,
+      "loss": 0.9831,
+      "step": 8102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3270439622691649,
+      "learning_rate": 0.00014387944923971782,
+      "loss": 1.1035,
+      "step": 8103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29562739459578125,
+      "learning_rate": 0.0001438652334950381,
+      "loss": 1.0352,
+      "step": 8104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2771019668036566,
+      "learning_rate": 0.00014385101665260338,
+      "loss": 1.0598,
+      "step": 8105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26504488308797247,
+      "learning_rate": 0.0001438367987127694,
+      "loss": 1.1604,
+      "step": 8106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3218630057212102,
+      "learning_rate": 0.000143822579675892,
+      "loss": 1.0142,
+      "step": 8107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28024189474691924,
+      "learning_rate": 0.00014380835954232697,
+      "loss": 1.0721,
+      "step": 8108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2852976599263691,
+      "learning_rate": 0.00014379413831243026,
+      "loss": 0.9966,
+      "step": 8109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2904612714774512,
+      "learning_rate": 0.00014377991598655765,
+      "loss": 1.0675,
+      "step": 8110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26588060857502965,
+      "learning_rate": 0.00014376569256506516,
+      "loss": 1.0215,
+      "step": 8111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28963536427741476,
+      "learning_rate": 0.0001437514680483087,
+      "loss": 0.9806,
+      "step": 8112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2669044448329335,
+      "learning_rate": 0.00014373724243664423,
+      "loss": 1.1047,
+      "step": 8113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2904302678513009,
+      "learning_rate": 0.00014372301573042782,
+      "loss": 1.0147,
+      "step": 8114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25491810409780064,
+      "learning_rate": 0.00014370878793001546,
+      "loss": 1.0453,
+      "step": 8115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2592085806502541,
+      "learning_rate": 0.0001436945590357632,
+      "loss": 1.0276,
+      "step": 8116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2821215453803441,
+      "learning_rate": 0.00014368032904802714,
+      "loss": 1.0488,
+      "step": 8117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2615363952942239,
+      "learning_rate": 0.00014366609796716338,
+      "loss": 1.1121,
+      "step": 8118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.31736889976384736,
+      "learning_rate": 0.0001436518657935281,
+      "loss": 1.1233,
+      "step": 8119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2598336195194818,
+      "learning_rate": 0.00014363763252747745,
+      "loss": 1.0805,
+      "step": 8120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3284573374828336,
+      "learning_rate": 0.0001436233981693676,
+      "loss": 1.0575,
+      "step": 8121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2908417184290244,
+      "learning_rate": 0.00014360916271955482,
+      "loss": 1.0383,
+      "step": 8122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28781407930130587,
+      "learning_rate": 0.0001435949261783953,
+      "loss": 0.9912,
+      "step": 8123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2929647743747517,
+      "learning_rate": 0.0001435806885462454,
+      "loss": 1.0553,
+      "step": 8124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2753194722230229,
+      "learning_rate": 0.00014356644982346133,
+      "loss": 0.9769,
+      "step": 8125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3013577274487869,
+      "learning_rate": 0.0001435522100103995,
+      "loss": 1.0244,
+      "step": 8126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2901508132919651,
+      "learning_rate": 0.00014353796910741623,
+      "loss": 1.011,
+      "step": 8127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29254447956246066,
+      "learning_rate": 0.0001435237271148679,
+      "loss": 1.0236,
+      "step": 8128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3042957195012631,
+      "learning_rate": 0.000143509484033111,
+      "loss": 1.0665,
+      "step": 8129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29428251771956543,
+      "learning_rate": 0.0001434952398625019,
+      "loss": 1.065,
+      "step": 8130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2734033579364972,
+      "learning_rate": 0.00014348099460339707,
+      "loss": 1.001,
+      "step": 8131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3303462730760466,
+      "learning_rate": 0.00014346674825615303,
+      "loss": 1.0916,
+      "step": 8132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27668942643869243,
+      "learning_rate": 0.0001434525008211263,
+      "loss": 1.0658,
+      "step": 8133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2953192209869138,
+      "learning_rate": 0.00014343825229867343,
+      "loss": 1.0055,
+      "step": 8134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30590123544431624,
+      "learning_rate": 0.00014342400268915097,
+      "loss": 1.1254,
+      "step": 8135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.289254633390817,
+      "learning_rate": 0.00014340975199291558,
+      "loss": 0.8961,
+      "step": 8136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2554256044992401,
+      "learning_rate": 0.00014339550021032384,
+      "loss": 1.0181,
+      "step": 8137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3141083587137128,
+      "learning_rate": 0.00014338124734173245,
+      "loss": 1.044,
+      "step": 8138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2569367079798127,
+      "learning_rate": 0.0001433669933874981,
+      "loss": 1.1267,
+      "step": 8139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2678458091756298,
+      "learning_rate": 0.00014335273834797745,
+      "loss": 1.1156,
+      "step": 8140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29169647370051666,
+      "learning_rate": 0.0001433384822235273,
+      "loss": 1.0323,
+      "step": 8141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2742299999761089,
+      "learning_rate": 0.0001433242250145044,
+      "loss": 1.0005,
+      "step": 8142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2685403344557798,
+      "learning_rate": 0.00014330996672126553,
+      "loss": 0.9437,
+      "step": 8143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29515305992400265,
+      "learning_rate": 0.0001432957073441675,
+      "loss": 1.1712,
+      "step": 8144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2932051822985775,
+      "learning_rate": 0.00014328144688356722,
+      "loss": 1.083,
+      "step": 8145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28221973458585115,
+      "learning_rate": 0.00014326718533982154,
+      "loss": 1.0657,
+      "step": 8146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2528985840125888,
+      "learning_rate": 0.00014325292271328733,
+      "loss": 1.1343,
+      "step": 8147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3099604595171793,
+      "learning_rate": 0.00014323865900432153,
+      "loss": 1.1354,
+      "step": 8148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28724423689538947,
+      "learning_rate": 0.00014322439421328114,
+      "loss": 1.0258,
+      "step": 8149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29908396514945557,
+      "learning_rate": 0.0001432101283405231,
+      "loss": 1.0992,
+      "step": 8150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25176928414336436,
+      "learning_rate": 0.00014319586138640447,
+      "loss": 1.127,
+      "step": 8151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27010420300606885,
+      "learning_rate": 0.00014318159335128226,
+      "loss": 1.112,
+      "step": 8152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27285928105597573,
+      "learning_rate": 0.0001431673242355135,
+      "loss": 1.0926,
+      "step": 8153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2606896768983877,
+      "learning_rate": 0.00014315305403945534,
+      "loss": 1.0197,
+      "step": 8154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2828962978680743,
+      "learning_rate": 0.0001431387827634649,
+      "loss": 1.1644,
+      "step": 8155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30392229600605375,
+      "learning_rate": 0.00014312451040789928,
+      "loss": 1.0046,
+      "step": 8156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.22992598628886599,
+      "learning_rate": 0.0001431102369731157,
+      "loss": 1.1334,
+      "step": 8157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27602772065447234,
+      "learning_rate": 0.00014309596245947134,
+      "loss": 1.1262,
+      "step": 8158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25737443548054917,
+      "learning_rate": 0.00014308168686732344,
+      "loss": 0.9536,
+      "step": 8159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2674977275846256,
+      "learning_rate": 0.00014306741019702926,
+      "loss": 1.1088,
+      "step": 8160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26488508791506327,
+      "learning_rate": 0.00014305313244894604,
+      "loss": 1.0541,
+      "step": 8161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2689193524609303,
+      "learning_rate": 0.00014303885362343115,
+      "loss": 1.1577,
+      "step": 8162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2879200289332181,
+      "learning_rate": 0.00014302457372084192,
+      "loss": 1.1757,
+      "step": 8163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2867128935586349,
+      "learning_rate": 0.00014301029274153563,
+      "loss": 1.0072,
+      "step": 8164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24341160678709342,
+      "learning_rate": 0.00014299601068586978,
+      "loss": 1.0225,
+      "step": 8165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2762861135172704,
+      "learning_rate": 0.00014298172755420173,
+      "loss": 1.1304,
+      "step": 8166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2769763284115676,
+      "learning_rate": 0.00014296744334688893,
+      "loss": 1.0157,
+      "step": 8167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26734305962867105,
+      "learning_rate": 0.0001429531580642889,
+      "loss": 1.1552,
+      "step": 8168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26480907501831297,
+      "learning_rate": 0.00014293887170675903,
+      "loss": 0.9893,
+      "step": 8169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29281043302440196,
+      "learning_rate": 0.00014292458427465695,
+      "loss": 0.9863,
+      "step": 8170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2888301828367874,
+      "learning_rate": 0.00014291029576834013,
+      "loss": 1.1172,
+      "step": 8171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3039150196032684,
+      "learning_rate": 0.00014289600618816627,
+      "loss": 1.0797,
+      "step": 8172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2841118966636821,
+      "learning_rate": 0.0001428817155344928,
+      "loss": 1.1855,
+      "step": 8173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28079473707368974,
+      "learning_rate": 0.0001428674238076775,
+      "loss": 1.0272,
+      "step": 8174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2862456351557848,
+      "learning_rate": 0.00014285313100807797,
+      "loss": 1.0838,
+      "step": 8175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2602397115203145,
+      "learning_rate": 0.00014283883713605192,
+      "loss": 1.083,
+      "step": 8176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.30104111034073594,
+      "learning_rate": 0.00014282454219195702,
+      "loss": 1.0324,
+      "step": 8177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.24143983810247652,
+      "learning_rate": 0.00014281024617615105,
+      "loss": 1.0873,
+      "step": 8178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2932603381628741,
+      "learning_rate": 0.00014279594908899175,
+      "loss": 1.017,
+      "step": 8179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3006594349278346,
+      "learning_rate": 0.00014278165093083696,
+      "loss": 1.0753,
+      "step": 8180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28022927773599904,
+      "learning_rate": 0.00014276735170204444,
+      "loss": 1.0561,
+      "step": 8181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27844733340976957,
+      "learning_rate": 0.0001427530514029721,
+      "loss": 1.1647,
+      "step": 8182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2571022498864135,
+      "learning_rate": 0.00014273875003397774,
+      "loss": 0.92,
+      "step": 8183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2692587322659136,
+      "learning_rate": 0.0001427244475954193,
+      "loss": 1.0409,
+      "step": 8184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2868029667790984,
+      "learning_rate": 0.00014271014408765472,
+      "loss": 0.9207,
+      "step": 8185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3122721251692393,
+      "learning_rate": 0.00014269583951104196,
+      "loss": 1.0165,
+      "step": 8186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.240316991720147,
+      "learning_rate": 0.00014268153386593898,
+      "loss": 1.1213,
+      "step": 8187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.283970934006317,
+      "learning_rate": 0.00014266722715270376,
+      "loss": 1.1421,
+      "step": 8188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2848531576221525,
+      "learning_rate": 0.0001426529193716944,
+      "loss": 1.0823,
+      "step": 8189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2683429644781905,
+      "learning_rate": 0.0001426386105232689,
+      "loss": 1.0785,
+      "step": 8190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2862688618200794,
+      "learning_rate": 0.00014262430060778538,
+      "loss": 1.0695,
+      "step": 8191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26946047893455116,
+      "learning_rate": 0.00014260998962560195,
+      "loss": 0.9804,
+      "step": 8192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3044969174847609,
+      "learning_rate": 0.00014259567757707675,
+      "loss": 1.0052,
+      "step": 8193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.27698543650468876,
+      "learning_rate": 0.00014258136446256795,
+      "loss": 0.9854,
+      "step": 8194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2971561209977439,
+      "learning_rate": 0.00014256705028243375,
+      "loss": 1.0358,
+      "step": 8195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2770540731397757,
+      "learning_rate": 0.00014255273503703238,
+      "loss": 1.0042,
+      "step": 8196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28561457756539305,
+      "learning_rate": 0.00014253841872672202,
+      "loss": 1.1068,
+      "step": 8197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2813535778333693,
+      "learning_rate": 0.00014252410135186103,
+      "loss": 1.1568,
+      "step": 8198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3051058643154024,
+      "learning_rate": 0.00014250978291280766,
+      "loss": 1.0966,
+      "step": 8199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2995687199626439,
+      "learning_rate": 0.00014249546340992027,
+      "loss": 0.9566,
+      "step": 8200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.25038359737911414,
+      "learning_rate": 0.0001424811428435572,
+      "loss": 0.9642,
+      "step": 8201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2755441297303153,
+      "learning_rate": 0.00014246682121407686,
+      "loss": 0.9964,
+      "step": 8202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28771744037000346,
+      "learning_rate": 0.0001424524985218376,
+      "loss": 1.1178,
+      "step": 8203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.26096812303528166,
+      "learning_rate": 0.00014243817476719789,
+      "loss": 1.0602,
+      "step": 8204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2910822777089926,
+      "learning_rate": 0.00014242384995051617,
+      "loss": 1.1028,
+      "step": 8205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2946214010693757,
+      "learning_rate": 0.000142409524072151,
+      "loss": 1.0582,
+      "step": 8206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2782362063554336,
+      "learning_rate": 0.00014239519713246077,
+      "loss": 1.0472,
+      "step": 8207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3096824457906856,
+      "learning_rate": 0.00014238086913180407,
+      "loss": 1.1764,
+      "step": 8208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2705085127084864,
+      "learning_rate": 0.00014236654007053956,
+      "loss": 1.0761,
+      "step": 8209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2746196439259208,
+      "learning_rate": 0.00014235220994902572,
+      "loss": 1.0842,
+      "step": 8210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2818643359206753,
+      "learning_rate": 0.0001423378787676212,
+      "loss": 1.0669,
+      "step": 8211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3016521365181387,
+      "learning_rate": 0.0001423235465266847,
+      "loss": 1.1674,
+      "step": 8212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3097236944936014,
+      "learning_rate": 0.0001423092132265748,
+      "loss": 1.019,
+      "step": 8213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25229115162540394,
+      "learning_rate": 0.00014229487886765026,
+      "loss": 1.1445,
+      "step": 8214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27446316421143896,
+      "learning_rate": 0.0001422805434502698,
+      "loss": 1.1201,
+      "step": 8215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.265891783959781,
+      "learning_rate": 0.00014226620697479217,
+      "loss": 0.9653,
+      "step": 8216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2858901663372903,
+      "learning_rate": 0.00014225186944157614,
+      "loss": 1.1012,
+      "step": 8217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24368459446009416,
+      "learning_rate": 0.00014223753085098052,
+      "loss": 1.0438,
+      "step": 8218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2658297205018217,
+      "learning_rate": 0.00014222319120336415,
+      "loss": 1.0923,
+      "step": 8219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27643416390290804,
+      "learning_rate": 0.0001422088504990859,
+      "loss": 1.0201,
+      "step": 8220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27516078932172305,
+      "learning_rate": 0.00014219450873850464,
+      "loss": 0.9544,
+      "step": 8221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27601315150275746,
+      "learning_rate": 0.00014218016592197925,
+      "loss": 1.0343,
+      "step": 8222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24192057612707965,
+      "learning_rate": 0.00014216582204986872,
+      "loss": 1.0806,
+      "step": 8223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28077409693408506,
+      "learning_rate": 0.000142151477122532,
+      "loss": 0.8763,
+      "step": 8224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27896630435292175,
+      "learning_rate": 0.00014213713114032803,
+      "loss": 1.0493,
+      "step": 8225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2737335624375359,
+      "learning_rate": 0.0001421227841036159,
+      "loss": 1.0675,
+      "step": 8226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3407111785009769,
+      "learning_rate": 0.00014210843601275466,
+      "loss": 1.0628,
+      "step": 8227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2617914284389135,
+      "learning_rate": 0.00014209408686810329,
+      "loss": 1.0987,
+      "step": 8228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3013100303523883,
+      "learning_rate": 0.00014207973667002097,
+      "loss": 0.9312,
+      "step": 8229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2846920379517571,
+      "learning_rate": 0.00014206538541886677,
+      "loss": 1.0237,
+      "step": 8230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30283490816190023,
+      "learning_rate": 0.0001420510331149999,
+      "loss": 1.0535,
+      "step": 8231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24872498659564674,
+      "learning_rate": 0.00014203667975877946,
+      "loss": 1.057,
+      "step": 8232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27630382799540015,
+      "learning_rate": 0.00014202232535056472,
+      "loss": 1.0121,
+      "step": 8233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28477175943154037,
+      "learning_rate": 0.00014200796989071487,
+      "loss": 1.045,
+      "step": 8234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2719512123825706,
+      "learning_rate": 0.00014199361337958915,
+      "loss": 1.075,
+      "step": 8235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27304851623487275,
+      "learning_rate": 0.0001419792558175469,
+      "loss": 0.9973,
+      "step": 8236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2749777621769799,
+      "learning_rate": 0.0001419648972049474,
+      "loss": 1.0225,
+      "step": 8237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26176184932542573,
+      "learning_rate": 0.0001419505375421499,
+      "loss": 1.061,
+      "step": 8238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2825657955231182,
+      "learning_rate": 0.0001419361768295139,
+      "loss": 1.039,
+      "step": 8239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.26238648793597996,
+      "learning_rate": 0.00014192181506739868,
+      "loss": 0.9806,
+      "step": 8240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2596777187775208,
+      "learning_rate": 0.0001419074522561637,
+      "loss": 1.0234,
+      "step": 8241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2883448002693972,
+      "learning_rate": 0.0001418930883961684,
+      "loss": 1.113,
+      "step": 8242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31700421775940696,
+      "learning_rate": 0.00014187872348777223,
+      "loss": 0.9246,
+      "step": 8243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2502687124845023,
+      "learning_rate": 0.00014186435753133468,
+      "loss": 0.9632,
+      "step": 8244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2721584066704403,
+      "learning_rate": 0.00014184999052721528,
+      "loss": 1.0489,
+      "step": 8245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2672093711111685,
+      "learning_rate": 0.00014183562247577358,
+      "loss": 1.029,
+      "step": 8246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31595242015080555,
+      "learning_rate": 0.00014182125337736912,
+      "loss": 1.0484,
+      "step": 8247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.217568494553999,
+      "learning_rate": 0.0001418068832323615,
+      "loss": 0.8719,
+      "step": 8248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27618501018822034,
+      "learning_rate": 0.00014179251204111037,
+      "loss": 0.9553,
+      "step": 8249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28614850226175115,
+      "learning_rate": 0.00014177813980397535,
+      "loss": 1.0244,
+      "step": 8250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2804392960583254,
+      "learning_rate": 0.00014176376652131614,
+      "loss": 1.117,
+      "step": 8251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.296161246928998,
+      "learning_rate": 0.0001417493921934924,
+      "loss": 0.9899,
+      "step": 8252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27500753507752407,
+      "learning_rate": 0.00014173501682086389,
+      "loss": 1.0332,
+      "step": 8253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27240297524975027,
+      "learning_rate": 0.00014172064040379037,
+      "loss": 1.1169,
+      "step": 8254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2862065599891984,
+      "learning_rate": 0.00014170626294263158,
+      "loss": 0.9608,
+      "step": 8255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29049647009040475,
+      "learning_rate": 0.00014169188443774737,
+      "loss": 1.0737,
+      "step": 8256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30650214779789686,
+      "learning_rate": 0.00014167750488949753,
+      "loss": 1.1447,
+      "step": 8257
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25682892320136824,
+      "learning_rate": 0.00014166312429824196,
+      "loss": 1.1008,
+      "step": 8258
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.289522043516876,
+      "learning_rate": 0.0001416487426643405,
+      "loss": 0.9996,
+      "step": 8259
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2612733947909529,
+      "learning_rate": 0.00014163435998815308,
+      "loss": 1.0454,
+      "step": 8260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3017746734519574,
+      "learning_rate": 0.00014161997627003964,
+      "loss": 1.1057,
+      "step": 8261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3026134541065002,
+      "learning_rate": 0.0001416055915103601,
+      "loss": 1.0563,
+      "step": 8262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2921549638360822,
+      "learning_rate": 0.00014159120570947454,
+      "loss": 1.2412,
+      "step": 8263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.27462991805196874,
+      "learning_rate": 0.00014157681886774293,
+      "loss": 1.1579,
+      "step": 8264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30353840216135913,
+      "learning_rate": 0.0001415624309855253,
+      "loss": 1.0529,
+      "step": 8265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2628785239134127,
+      "learning_rate": 0.00014154804206318165,
+      "loss": 1.0783,
+      "step": 8266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.279824355560124,
+      "learning_rate": 0.00014153365210107217,
+      "loss": 0.992,
+      "step": 8267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25845741171089465,
+      "learning_rate": 0.00014151926109955696,
+      "loss": 0.9018,
+      "step": 8268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2949235009130459,
+      "learning_rate": 0.0001415048690589961,
+      "loss": 1.1514,
+      "step": 8269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3091426382260449,
+      "learning_rate": 0.00014149047597974984,
+      "loss": 1.128,
+      "step": 8270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.261254800590016,
+      "learning_rate": 0.00014147608186217836,
+      "loss": 1.0248,
+      "step": 8271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28155062611196985,
+      "learning_rate": 0.0001414616867066418,
+      "loss": 1.1088,
+      "step": 8272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2802398085480191,
+      "learning_rate": 0.00014144729051350055,
+      "loss": 1.0411,
+      "step": 8273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28950599861246645,
+      "learning_rate": 0.00014143289328311478,
+      "loss": 1.1909,
+      "step": 8274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3156838267805892,
+      "learning_rate": 0.0001414184950158448,
+      "loss": 1.0001,
+      "step": 8275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2761437699110209,
+      "learning_rate": 0.00014140409571205095,
+      "loss": 1.0747,
+      "step": 8276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2556738607507545,
+      "learning_rate": 0.00014138969537209358,
+      "loss": 1.1727,
+      "step": 8277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3116455068455913,
+      "learning_rate": 0.0001413752939963331,
+      "loss": 1.0714,
+      "step": 8278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3023777310208979,
+      "learning_rate": 0.00014136089158512985,
+      "loss": 1.1541,
+      "step": 8279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29881211105271765,
+      "learning_rate": 0.00014134648813884433,
+      "loss": 1.1441,
+      "step": 8280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2564362293165476,
+      "learning_rate": 0.00014133208365783693,
+      "loss": 1.2231,
+      "step": 8281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25459170299471473,
+      "learning_rate": 0.00014131767814246817,
+      "loss": 1.0638,
+      "step": 8282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3164293509155231,
+      "learning_rate": 0.00014130327159309853,
+      "loss": 1.0238,
+      "step": 8283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3161096823663635,
+      "learning_rate": 0.0001412888640100886,
+      "loss": 1.0997,
+      "step": 8284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2926302601160675,
+      "learning_rate": 0.00014127445539379886,
+      "loss": 1.0269,
+      "step": 8285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3026119265810598,
+      "learning_rate": 0.00014126004574458996,
+      "loss": 1.0535,
+      "step": 8286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.31239765394033814,
+      "learning_rate": 0.00014124563506282247,
+      "loss": 1.1042,
+      "step": 8287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2730026191662197,
+      "learning_rate": 0.00014123122334885706,
+      "loss": 1.0406,
+      "step": 8288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2802587764341756,
+      "learning_rate": 0.00014121681060305435,
+      "loss": 1.0393,
+      "step": 8289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2965134782106928,
+      "learning_rate": 0.00014120239682577506,
+      "loss": 1.1487,
+      "step": 8290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3604661013234594,
+      "learning_rate": 0.0001411879820173799,
+      "loss": 1.0004,
+      "step": 8291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2754245502665909,
+      "learning_rate": 0.0001411735661782296,
+      "loss": 0.9554,
+      "step": 8292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.24638138967620207,
+      "learning_rate": 0.00014115914930868493,
+      "loss": 1.1353,
+      "step": 8293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30600504343829077,
+      "learning_rate": 0.00014114473140910668,
+      "loss": 1.1201,
+      "step": 8294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3035845944676315,
+      "learning_rate": 0.00014113031247985566,
+      "loss": 1.0861,
+      "step": 8295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3384934195188291,
+      "learning_rate": 0.00014111589252129272,
+      "loss": 1.1501,
+      "step": 8296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2716845796991304,
+      "learning_rate": 0.00014110147153377874,
+      "loss": 1.0534,
+      "step": 8297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3062067246424864,
+      "learning_rate": 0.0001410870495176746,
+      "loss": 1.1802,
+      "step": 8298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2673331008196946,
+      "learning_rate": 0.0001410726264733412,
+      "loss": 1.0404,
+      "step": 8299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2860719333532579,
+      "learning_rate": 0.00014105820240113955,
+      "loss": 1.1296,
+      "step": 8300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2924049271387296,
+      "learning_rate": 0.0001410437773014306,
+      "loss": 1.0039,
+      "step": 8301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.32493379544572254,
+      "learning_rate": 0.00014102935117457524,
+      "loss": 0.9896,
+      "step": 8302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.308689353469358,
+      "learning_rate": 0.00014101492402093463,
+      "loss": 0.9233,
+      "step": 8303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2713635705097109,
+      "learning_rate": 0.00014100049584086979,
+      "loss": 1.0322,
+      "step": 8304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2853711362692117,
+      "learning_rate": 0.00014098606663474176,
+      "loss": 1.0401,
+      "step": 8305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.25822823305650644,
+      "learning_rate": 0.00014097163640291164,
+      "loss": 1.0283,
+      "step": 8306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3343555252521715,
+      "learning_rate": 0.00014095720514574058,
+      "loss": 1.0707,
+      "step": 8307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28266257518096666,
+      "learning_rate": 0.00014094277286358972,
+      "loss": 1.1636,
+      "step": 8308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2649055365462347,
+      "learning_rate": 0.00014092833955682026,
+      "loss": 1.0662,
+      "step": 8309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31075468177448395,
+      "learning_rate": 0.00014091390522579333,
+      "loss": 1.0101,
+      "step": 8310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3011177221612655,
+      "learning_rate": 0.00014089946987087023,
+      "loss": 1.136,
+      "step": 8311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29851607856732926,
+      "learning_rate": 0.00014088503349241223,
+      "loss": 1.1139,
+      "step": 8312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29855074002538673,
+      "learning_rate": 0.00014087059609078052,
+      "loss": 1.0843,
+      "step": 8313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2768067900609332,
+      "learning_rate": 0.0001408561576663365,
+      "loss": 1.0972,
+      "step": 8314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2819211049236916,
+      "learning_rate": 0.00014084171821944144,
+      "loss": 1.1288,
+      "step": 8315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28684815717935336,
+      "learning_rate": 0.00014082727775045667,
+      "loss": 1.1531,
+      "step": 8316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29080240408238495,
+      "learning_rate": 0.00014081283625974367,
+      "loss": 1.1591,
+      "step": 8317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29429553590146723,
+      "learning_rate": 0.0001407983937476638,
+      "loss": 1.1297,
+      "step": 8318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30787470171045134,
+      "learning_rate": 0.00014078395021457845,
+      "loss": 1.2024,
+      "step": 8319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27975097914183483,
+      "learning_rate": 0.0001407695056608491,
+      "loss": 0.9402,
+      "step": 8320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2832395257132092,
+      "learning_rate": 0.0001407550600868373,
+      "loss": 1.0379,
+      "step": 8321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29437871608032573,
+      "learning_rate": 0.00014074061349290447,
+      "loss": 1.0362,
+      "step": 8322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.269143724602035,
+      "learning_rate": 0.00014072616587941218,
+      "loss": 1.0537,
+      "step": 8323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29710110944603785,
+      "learning_rate": 0.00014071171724672202,
+      "loss": 1.0264,
+      "step": 8324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30415387019725554,
+      "learning_rate": 0.00014069726759519553,
+      "loss": 1.231,
+      "step": 8325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29030887320948506,
+      "learning_rate": 0.00014068281692519434,
+      "loss": 1.1106,
+      "step": 8326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.32054128469538806,
+      "learning_rate": 0.0001406683652370801,
+      "loss": 1.1057,
+      "step": 8327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30650933574016137,
+      "learning_rate": 0.00014065391253121446,
+      "loss": 1.1166,
+      "step": 8328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2694980594720075,
+      "learning_rate": 0.0001406394588079591,
+      "loss": 0.9278,
+      "step": 8329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2696507606928678,
+      "learning_rate": 0.00014062500406767574,
+      "loss": 1.079,
+      "step": 8330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2759447283420335,
+      "learning_rate": 0.00014061054831072614,
+      "loss": 0.9918,
+      "step": 8331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2738734236935922,
+      "learning_rate": 0.00014059609153747204,
+      "loss": 1.1079,
+      "step": 8332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28515443322196976,
+      "learning_rate": 0.00014058163374827521,
+      "loss": 1.0569,
+      "step": 8333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3080422030491291,
+      "learning_rate": 0.0001405671749434975,
+      "loss": 1.0172,
+      "step": 8334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6582786045483043,
+      "learning_rate": 0.00014055271512350079,
+      "loss": 1.458,
+      "step": 8335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29640316228647345,
+      "learning_rate": 0.00014053825428864686,
+      "loss": 0.9889,
+      "step": 8336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3224808939799605,
+      "learning_rate": 0.00014052379243929762,
+      "loss": 1.143,
+      "step": 8337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26780388495165136,
+      "learning_rate": 0.00014050932957581505,
+      "loss": 1.1065,
+      "step": 8338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3007238046192213,
+      "learning_rate": 0.000140494865698561,
+      "loss": 1.2191,
+      "step": 8339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3150280632023674,
+      "learning_rate": 0.00014048040080789752,
+      "loss": 1.1275,
+      "step": 8340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25501610087568716,
+      "learning_rate": 0.00014046593490418656,
+      "loss": 1.0839,
+      "step": 8341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3241245476281749,
+      "learning_rate": 0.00014045146798779014,
+      "loss": 1.0121,
+      "step": 8342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28717103288948476,
+      "learning_rate": 0.00014043700005907033,
+      "loss": 1.0748,
+      "step": 8343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2949240771446552,
+      "learning_rate": 0.00014042253111838917,
+      "loss": 1.1718,
+      "step": 8344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3080849599666515,
+      "learning_rate": 0.00014040806116610873,
+      "loss": 1.1352,
+      "step": 8345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2757839282426602,
+      "learning_rate": 0.0001403935902025912,
+      "loss": 1.0233,
+      "step": 8346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2993473446158299,
+      "learning_rate": 0.00014037911822819868,
+      "loss": 1.1544,
+      "step": 8347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.292778124826498,
+      "learning_rate": 0.00014036464524329337,
+      "loss": 1.1381,
+      "step": 8348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2958157472586977,
+      "learning_rate": 0.00014035017124823743,
+      "loss": 1.0634,
+      "step": 8349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2973614768918341,
+      "learning_rate": 0.00014033569624339308,
+      "loss": 1.0993,
+      "step": 8350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2877291411238009,
+      "learning_rate": 0.0001403212202291226,
+      "loss": 1.1405,
+      "step": 8351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3161483762081105,
+      "learning_rate": 0.00014030674320578823,
+      "loss": 1.0157,
+      "step": 8352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3097883664079197,
+      "learning_rate": 0.0001402922651737523,
+      "loss": 1.1105,
+      "step": 8353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3052718311315742,
+      "learning_rate": 0.00014027778613337708,
+      "loss": 1.1271,
+      "step": 8354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.24254225363550241,
+      "learning_rate": 0.00014026330608502496,
+      "loss": 0.9638,
+      "step": 8355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28453519098553043,
+      "learning_rate": 0.00014024882502905833,
+      "loss": 1.095,
+      "step": 8356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3148338032717982,
+      "learning_rate": 0.0001402343429658395,
+      "loss": 1.1137,
+      "step": 8357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3117328371653278,
+      "learning_rate": 0.000140219859895731,
+      "loss": 1.1287,
+      "step": 8358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26695229050684094,
+      "learning_rate": 0.00014020537581909524,
+      "loss": 1.0255,
+      "step": 8359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28843280197149884,
+      "learning_rate": 0.00014019089073629464,
+      "loss": 1.0416,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3067274794496299,
+      "learning_rate": 0.00014017640464769176,
+      "loss": 1.0825,
+      "step": 8361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2739855692621317,
+      "learning_rate": 0.00014016191755364908,
+      "loss": 1.0509,
+      "step": 8362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2395995459480104,
+      "learning_rate": 0.0001401474294545292,
+      "loss": 1.0843,
+      "step": 8363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2759339406448203,
+      "learning_rate": 0.00014013294035069467,
+      "loss": 1.1089,
+      "step": 8364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3012501447545405,
+      "learning_rate": 0.00014011845024250805,
+      "loss": 1.0218,
+      "step": 8365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2743545412783308,
+      "learning_rate": 0.00014010395913033202,
+      "loss": 0.9881,
+      "step": 8366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.22999177193777853,
+      "learning_rate": 0.00014008946701452921,
+      "loss": 1.0122,
+      "step": 8367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3145753551754261,
+      "learning_rate": 0.00014007497389546228,
+      "loss": 1.1592,
+      "step": 8368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29504190170329436,
+      "learning_rate": 0.000140060479773494,
+      "loss": 1.0783,
+      "step": 8369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27650998107466573,
+      "learning_rate": 0.00014004598464898698,
+      "loss": 1.1227,
+      "step": 8370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2594677164964031,
+      "learning_rate": 0.00014003148852230403,
+      "loss": 1.0184,
+      "step": 8371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26498120130511177,
+      "learning_rate": 0.00014001699139380792,
+      "loss": 1.016,
+      "step": 8372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33720870260470887,
+      "learning_rate": 0.00014000249326386147,
+      "loss": 0.9954,
+      "step": 8373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27003696605507693,
+      "learning_rate": 0.0001399879941328275,
+      "loss": 1.0899,
+      "step": 8374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25731301945171275,
+      "learning_rate": 0.0001399734940010688,
+      "loss": 1.1751,
+      "step": 8375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.288234443384105,
+      "learning_rate": 0.0001399589928689483,
+      "loss": 1.1105,
+      "step": 8376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3086646943980646,
+      "learning_rate": 0.0001399444907368289,
+      "loss": 1.0816,
+      "step": 8377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2857761154797601,
+      "learning_rate": 0.00013992998760507352,
+      "loss": 1.1379,
+      "step": 8378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2848709741740318,
+      "learning_rate": 0.00013991548347404512,
+      "loss": 1.1337,
+      "step": 8379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26144857764795576,
+      "learning_rate": 0.00013990097834410664,
+      "loss": 0.9858,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28140647243062356,
+      "learning_rate": 0.0001398864722156211,
+      "loss": 1.1319,
+      "step": 8381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2757644481768791,
+      "learning_rate": 0.00013987196508895153,
+      "loss": 0.9251,
+      "step": 8382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29461811148669015,
+      "learning_rate": 0.00013985745696446097,
+      "loss": 1.1373,
+      "step": 8383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3141106450117521,
+      "learning_rate": 0.0001398429478425125,
+      "loss": 1.0046,
+      "step": 8384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2932601643453963,
+      "learning_rate": 0.00013982843772346922,
+      "loss": 1.0836,
+      "step": 8385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3080641745939993,
+      "learning_rate": 0.00013981392660769424,
+      "loss": 1.1587,
+      "step": 8386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3161172120816325,
+      "learning_rate": 0.00013979941449555075,
+      "loss": 1.0562,
+      "step": 8387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2919139460605944,
+      "learning_rate": 0.00013978490138740187,
+      "loss": 1.1742,
+      "step": 8388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27923451724228243,
+      "learning_rate": 0.00013977038728361086,
+      "loss": 1.1083,
+      "step": 8389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31621277440252016,
+      "learning_rate": 0.0001397558721845409,
+      "loss": 1.0143,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3053969306091787,
+      "learning_rate": 0.00013974135609055527,
+      "loss": 1.0729,
+      "step": 8391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2825597454845708,
+      "learning_rate": 0.00013972683900201723,
+      "loss": 1.0619,
+      "step": 8392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29068615722749824,
+      "learning_rate": 0.00013971232091929006,
+      "loss": 1.065,
+      "step": 8393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2853464043964099,
+      "learning_rate": 0.00013969780184273705,
+      "loss": 0.9789,
+      "step": 8394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2849572721200672,
+      "learning_rate": 0.0001396832817727217,
+      "loss": 1.0263,
+      "step": 8395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28211914596567356,
+      "learning_rate": 0.00013966876070960722,
+      "loss": 1.1037,
+      "step": 8396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2995697358994056,
+      "learning_rate": 0.00013965423865375712,
+      "loss": 1.1653,
+      "step": 8397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.29293561141447033,
+      "learning_rate": 0.0001396397156055347,
+      "loss": 1.094,
+      "step": 8398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25956145112228035,
+      "learning_rate": 0.00013962519156530354,
+      "loss": 1.0856,
+      "step": 8399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27947255917814445,
+      "learning_rate": 0.00013961066653342706,
+      "loss": 1.0245,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2745380656704247,
+      "learning_rate": 0.00013959614051026873,
+      "loss": 1.0483,
+      "step": 8401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2980202904500265,
+      "learning_rate": 0.0001395816134961921,
+      "loss": 1.0607,
+      "step": 8402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.25864322411654844,
+      "learning_rate": 0.00013956708549156072,
+      "loss": 0.9817,
+      "step": 8403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.27059445234233453,
+      "learning_rate": 0.00013955255649673816,
+      "loss": 1.0721,
+      "step": 8404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30462552256042574,
+      "learning_rate": 0.00013953802651208802,
+      "loss": 1.0382,
+      "step": 8405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2863563756673883,
+      "learning_rate": 0.0001395234955379739,
+      "loss": 1.1493,
+      "step": 8406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2862298189990451,
+      "learning_rate": 0.0001395089635747595,
+      "loss": 1.1067,
+      "step": 8407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2993560987033485,
+      "learning_rate": 0.00013949443062280842,
+      "loss": 1.081,
+      "step": 8408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2717526863604508,
+      "learning_rate": 0.00013947989668248442,
+      "loss": 1.0246,
+      "step": 8409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2843109420913706,
+      "learning_rate": 0.00013946536175415118,
+      "loss": 1.1101,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.31237587843637793,
+      "learning_rate": 0.00013945082583817245,
+      "loss": 1.0994,
+      "step": 8411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2655591048449623,
+      "learning_rate": 0.00013943628893491202,
+      "loss": 1.1926,
+      "step": 8412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2923686496830888,
+      "learning_rate": 0.0001394217510447337,
+      "loss": 1.1175,
+      "step": 8413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.26432275264898286,
+      "learning_rate": 0.00013940721216800127,
+      "loss": 0.9868,
+      "step": 8414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2767290301083631,
+      "learning_rate": 0.00013939267230507856,
+      "loss": 0.9424,
+      "step": 8415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2865371880691285,
+      "learning_rate": 0.0001393781314563295,
+      "loss": 1.1268,
+      "step": 8416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27359931565493517,
+      "learning_rate": 0.00013936358962211794,
+      "loss": 0.894,
+      "step": 8417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2682490335376522,
+      "learning_rate": 0.00013934904680280781,
+      "loss": 1.0634,
+      "step": 8418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.323666262645315,
+      "learning_rate": 0.00013933450299876305,
+      "loss": 1.0572,
+      "step": 8419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.313542678895405,
+      "learning_rate": 0.00013931995821034766,
+      "loss": 1.0228,
+      "step": 8420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2741537767451146,
+      "learning_rate": 0.00013930541243792555,
+      "loss": 1.1194,
+      "step": 8421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2727440924938454,
+      "learning_rate": 0.00013929086568186083,
+      "loss": 1.1567,
+      "step": 8422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27596495736712096,
+      "learning_rate": 0.0001392763179425175,
+      "loss": 1.0803,
+      "step": 8423
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2789242022899579,
+      "learning_rate": 0.00013926176922025963,
+      "loss": 0.912,
+      "step": 8424
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28533265323613294,
+      "learning_rate": 0.00013924721951545128,
+      "loss": 1.1343,
+      "step": 8425
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29008984332342763,
+      "learning_rate": 0.00013923266882845666,
+      "loss": 1.0476,
+      "step": 8426
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24250004605487083,
+      "learning_rate": 0.00013921811715963977,
+      "loss": 1.062,
+      "step": 8427
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29532956323803095,
+      "learning_rate": 0.0001392035645093649,
+      "loss": 1.0561,
+      "step": 8428
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3001458878270363,
+      "learning_rate": 0.00013918901087799616,
+      "loss": 1.081,
+      "step": 8429
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3037811128687649,
+      "learning_rate": 0.0001391744562658978,
+      "loss": 1.098,
+      "step": 8430
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3304936166647709,
+      "learning_rate": 0.00013915990067343408,
+      "loss": 0.9485,
+      "step": 8431
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2831506029753717,
+      "learning_rate": 0.0001391453441009692,
+      "loss": 1.0727,
+      "step": 8432
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2528874867012647,
+      "learning_rate": 0.0001391307865488675,
+      "loss": 1.046,
+      "step": 8433
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2926620284475519,
+      "learning_rate": 0.00013911622801749326,
+      "loss": 1.066,
+      "step": 8434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28658384259986724,
+      "learning_rate": 0.00013910166850721086,
+      "loss": 1.1164,
+      "step": 8435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2611948977130402,
+      "learning_rate": 0.0001390871080183846,
+      "loss": 1.1024,
+      "step": 8436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29623740606511145,
+      "learning_rate": 0.0001390725465513789,
+      "loss": 0.9396,
+      "step": 8437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2759634164236242,
+      "learning_rate": 0.00013905798410655817,
+      "loss": 0.991,
+      "step": 8438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2559555603571331,
+      "learning_rate": 0.00013904342068428688,
+      "loss": 0.9583,
+      "step": 8439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2834397140569843,
+      "learning_rate": 0.00013902885628492938,
+      "loss": 1.1322,
+      "step": 8440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2959073999001046,
+      "learning_rate": 0.00013901429090885028,
+      "loss": 1.0687,
+      "step": 8441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25399362804027736,
+      "learning_rate": 0.000138999724556414,
+      "loss": 1.0374,
+      "step": 8442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29065321813906453,
+      "learning_rate": 0.00013898515722798513,
+      "loss": 1.0508,
+      "step": 8443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37210670978861016,
+      "learning_rate": 0.00013897058892392818,
+      "loss": 0.9855,
+      "step": 8444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2800545837279327,
+      "learning_rate": 0.00013895601964460775,
+      "loss": 1.1169,
+      "step": 8445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2926130011031746,
+      "learning_rate": 0.00013894144939038844,
+      "loss": 1.0492,
+      "step": 8446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31446837274856276,
+      "learning_rate": 0.00013892687816163487,
+      "loss": 1.077,
+      "step": 8447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2894693179667743,
+      "learning_rate": 0.00013891230595871175,
+      "loss": 1.1005,
+      "step": 8448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2658040957365846,
+      "learning_rate": 0.0001388977327819837,
+      "loss": 1.129,
+      "step": 8449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30570591750106335,
+      "learning_rate": 0.0001388831586318154,
+      "loss": 1.1457,
+      "step": 8450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30282216470657447,
+      "learning_rate": 0.00013886858350857167,
+      "loss": 1.0694,
+      "step": 8451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3179740476883664,
+      "learning_rate": 0.00013885400741261717,
+      "loss": 1.1539,
+      "step": 8452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2714947725647423,
+      "learning_rate": 0.00013883943034431677,
+      "loss": 1.0887,
+      "step": 8453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2617135621115103,
+      "learning_rate": 0.0001388248523040352,
+      "loss": 1.1484,
+      "step": 8454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2589470456450907,
+      "learning_rate": 0.00013881027329213727,
+      "loss": 0.976,
+      "step": 8455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2971405117388152,
+      "learning_rate": 0.0001387956933089879,
+      "loss": 1.1232,
+      "step": 8456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2638200620165654,
+      "learning_rate": 0.0001387811123549519,
+      "loss": 1.1529,
+      "step": 8457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2792026970736888,
+      "learning_rate": 0.00013876653043039418,
+      "loss": 1.1957,
+      "step": 8458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27823742191201595,
+      "learning_rate": 0.0001387519475356797,
+      "loss": 1.0414,
+      "step": 8459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25238962999634595,
+      "learning_rate": 0.00013873736367117336,
+      "loss": 1.0228,
+      "step": 8460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28008622103734476,
+      "learning_rate": 0.00013872277883724015,
+      "loss": 1.023,
+      "step": 8461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2777949786844138,
+      "learning_rate": 0.00013870819303424506,
+      "loss": 1.1302,
+      "step": 8462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.24701542829888543,
+      "learning_rate": 0.0001386936062625531,
+      "loss": 1.1174,
+      "step": 8463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29033108955425646,
+      "learning_rate": 0.00013867901852252935,
+      "loss": 1.0872,
+      "step": 8464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2747150319565986,
+      "learning_rate": 0.00013866442981453887,
+      "loss": 0.95,
+      "step": 8465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28133751977616195,
+      "learning_rate": 0.00013864984013894669,
+      "loss": 1.04,
+      "step": 8466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2603016003322199,
+      "learning_rate": 0.00013863524949611798,
+      "loss": 1.0926,
+      "step": 8467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28873463989684345,
+      "learning_rate": 0.00013862065788641787,
+      "loss": 1.0727,
+      "step": 8468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3043522634022851,
+      "learning_rate": 0.00013860606531021155,
+      "loss": 1.017,
+      "step": 8469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.258266209093827,
+      "learning_rate": 0.00013859147176786417,
+      "loss": 1.0572,
+      "step": 8470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2960901666392521,
+      "learning_rate": 0.00013857687725974093,
+      "loss": 0.9861,
+      "step": 8471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29910616519981614,
+      "learning_rate": 0.00013856228178620709,
+      "loss": 1.1596,
+      "step": 8472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2990576464485771,
+      "learning_rate": 0.00013854768534762795,
+      "loss": 1.0265,
+      "step": 8473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31441877420729514,
+      "learning_rate": 0.00013853308794436876,
+      "loss": 1.0613,
+      "step": 8474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2634200444546087,
+      "learning_rate": 0.0001385184895767948,
+      "loss": 1.004,
+      "step": 8475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2556240322403397,
+      "learning_rate": 0.0001385038902452714,
+      "loss": 1.1238,
+      "step": 8476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27888482542250587,
+      "learning_rate": 0.00013848928995016403,
+      "loss": 1.1326,
+      "step": 8477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27815317220422986,
+      "learning_rate": 0.00013847468869183796,
+      "loss": 1.123,
+      "step": 8478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2831745331072525,
+      "learning_rate": 0.00013846008647065857,
+      "loss": 1.0664,
+      "step": 8479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3041401975450099,
+      "learning_rate": 0.0001384454832869914,
+      "loss": 1.0162,
+      "step": 8480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27417387963872447,
+      "learning_rate": 0.00013843087914120185,
+      "loss": 1.1522,
+      "step": 8481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2551238865778478,
+      "learning_rate": 0.00013841627403365537,
+      "loss": 1.1159,
+      "step": 8482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2772835560718097,
+      "learning_rate": 0.0001384016679647175,
+      "loss": 1.0014,
+      "step": 8483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.266588619864142,
+      "learning_rate": 0.00013838706093475379,
+      "loss": 1.0242,
+      "step": 8484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2950429695678513,
+      "learning_rate": 0.0001383724529441297,
+      "loss": 1.0789,
+      "step": 8485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26853491046599237,
+      "learning_rate": 0.00013835784399321088,
+      "loss": 1.1386,
+      "step": 8486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.27338834723695976,
+      "learning_rate": 0.0001383432340823629,
+      "loss": 0.9977,
+      "step": 8487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2865977979325858,
+      "learning_rate": 0.00013832862321195143,
+      "loss": 1.0583,
+      "step": 8488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28068233052657926,
+      "learning_rate": 0.000138314011382342,
+      "loss": 1.0611,
+      "step": 8489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3053917578678752,
+      "learning_rate": 0.0001382993985939004,
+      "loss": 1.1556,
+      "step": 8490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.26177227751546944,
+      "learning_rate": 0.00013828478484699227,
+      "loss": 1.1853,
+      "step": 8491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28197772368176605,
+      "learning_rate": 0.00013827017014198336,
+      "loss": 1.108,
+      "step": 8492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.25534797811164694,
+      "learning_rate": 0.00013825555447923935,
+      "loss": 1.1122,
+      "step": 8493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2985449459023056,
+      "learning_rate": 0.00013824093785912609,
+      "loss": 1.0912,
+      "step": 8494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3047078671768318,
+      "learning_rate": 0.0001382263202820093,
+      "loss": 0.909,
+      "step": 8495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2566715818782515,
+      "learning_rate": 0.0001382117017482548,
+      "loss": 1.0322,
+      "step": 8496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2612952596331514,
+      "learning_rate": 0.0001381970822582285,
+      "loss": 0.9924,
+      "step": 8497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2933790624492483,
+      "learning_rate": 0.00013818246181229618,
+      "loss": 1.0607,
+      "step": 8498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3051960049265422,
+      "learning_rate": 0.00013816784041082374,
+      "loss": 1.1308,
+      "step": 8499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2790257819761861,
+      "learning_rate": 0.0001381532180541772,
+      "loss": 1.0386,
+      "step": 8500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3118122919886825,
+      "learning_rate": 0.0001381385947427223,
+      "loss": 1.1372,
+      "step": 8501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29570017068411963,
+      "learning_rate": 0.00013812397047682513,
+      "loss": 1.0147,
+      "step": 8502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31683426314572966,
+      "learning_rate": 0.00013810934525685165,
+      "loss": 1.0712,
+      "step": 8503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2909795486157777,
+      "learning_rate": 0.00013809471908316783,
+      "loss": 1.2234,
+      "step": 8504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2955103667345504,
+      "learning_rate": 0.00013808009195613973,
+      "loss": 1.0783,
+      "step": 8505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2867522171152345,
+      "learning_rate": 0.0001380654638761334,
+      "loss": 1.0277,
+      "step": 8506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3681601782152508,
+      "learning_rate": 0.0001380508348435149,
+      "loss": 1.1069,
+      "step": 8507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2932701192842754,
+      "learning_rate": 0.00013803620485865035,
+      "loss": 1.0882,
+      "step": 8508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31028006826949395,
+      "learning_rate": 0.0001380215739219059,
+      "loss": 0.9705,
+      "step": 8509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.311953807926566,
+      "learning_rate": 0.00013800694203364763,
+      "loss": 1.1564,
+      "step": 8510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2959800715326205,
+      "learning_rate": 0.00013799230919424175,
+      "loss": 1.0579,
+      "step": 8511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2605263218657084,
+      "learning_rate": 0.00013797767540405447,
+      "loss": 1.0171,
+      "step": 8512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30387836906175536,
+      "learning_rate": 0.00013796304066345197,
+      "loss": 0.9672,
+      "step": 8513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.299921065982018,
+      "learning_rate": 0.00013794840497280056,
+      "loss": 1.1199,
+      "step": 8514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.32487111245703615,
+      "learning_rate": 0.00013793376833246644,
+      "loss": 1.2367,
+      "step": 8515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2847819157265213,
+      "learning_rate": 0.00013791913074281595,
+      "loss": 1.1113,
+      "step": 8516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2868712720370698,
+      "learning_rate": 0.00013790449220421535,
+      "loss": 1.1692,
+      "step": 8517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2678298209756187,
+      "learning_rate": 0.00013788985271703105,
+      "loss": 1.1474,
+      "step": 8518
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29099670328302274,
+      "learning_rate": 0.00013787521228162934,
+      "loss": 1.0656,
+      "step": 8519
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30474414530355776,
+      "learning_rate": 0.00013786057089837663,
+      "loss": 1.0057,
+      "step": 8520
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28018691357325703,
+      "learning_rate": 0.00013784592856763936,
+      "loss": 1.0637,
+      "step": 8521
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30281968491489586,
+      "learning_rate": 0.00013783128528978395,
+      "loss": 0.9859,
+      "step": 8522
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2833193148012386,
+      "learning_rate": 0.00013781664106517685,
+      "loss": 1.1878,
+      "step": 8523
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27855400568860456,
+      "learning_rate": 0.00013780199589418453,
+      "loss": 1.0736,
+      "step": 8524
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3023424425522343,
+      "learning_rate": 0.00013778734977717348,
+      "loss": 0.9647,
+      "step": 8525
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24163910511953707,
+      "learning_rate": 0.00013777270271451031,
+      "loss": 1.0617,
+      "step": 8526
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2643004336760188,
+      "learning_rate": 0.00013775805470656147,
+      "loss": 1.0083,
+      "step": 8527
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30690050558969756,
+      "learning_rate": 0.00013774340575369357,
+      "loss": 1.1863,
+      "step": 8528
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2803369681980211,
+      "learning_rate": 0.00013772875585627326,
+      "loss": 0.9811,
+      "step": 8529
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27882776591624236,
+      "learning_rate": 0.00013771410501466712,
+      "loss": 1.0438,
+      "step": 8530
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2866694179672929,
+      "learning_rate": 0.00013769945322924179,
+      "loss": 1.0089,
+      "step": 8531
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3193049906912792,
+      "learning_rate": 0.00013768480050036392,
+      "loss": 0.9945,
+      "step": 8532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3134610280410675,
+      "learning_rate": 0.00013767014682840027,
+      "loss": 1.0041,
+      "step": 8533
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.25858959641085166,
+      "learning_rate": 0.0001376554922137175,
+      "loss": 1.0951,
+      "step": 8534
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27495664936656555,
+      "learning_rate": 0.00013764083665668237,
+      "loss": 1.0078,
+      "step": 8535
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26221912952359344,
+      "learning_rate": 0.00013762618015766167,
+      "loss": 1.0953,
+      "step": 8536
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2969797589840666,
+      "learning_rate": 0.00013761152271702214,
+      "loss": 1.104,
+      "step": 8537
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3003478528891654,
+      "learning_rate": 0.00013759686433513062,
+      "loss": 1.0608,
+      "step": 8538
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.35023226379753625,
+      "learning_rate": 0.00013758220501235396,
+      "loss": 1.0195,
+      "step": 8539
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2598816030866285,
+      "learning_rate": 0.000137567544749059,
+      "loss": 0.8741,
+      "step": 8540
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29603444421799885,
+      "learning_rate": 0.0001375528835456126,
+      "loss": 1.124,
+      "step": 8541
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24876455442000533,
+      "learning_rate": 0.0001375382214023817,
+      "loss": 0.9333,
+      "step": 8542
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29348410647341444,
+      "learning_rate": 0.00013752355831973324,
+      "loss": 0.9545,
+      "step": 8543
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31014489060350814,
+      "learning_rate": 0.00013750889429803412,
+      "loss": 1.0946,
+      "step": 8544
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2901149550164192,
+      "learning_rate": 0.00013749422933765135,
+      "loss": 1.1371,
+      "step": 8545
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2796054693890663,
+      "learning_rate": 0.00013747956343895194,
+      "loss": 1.0809,
+      "step": 8546
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2809110179945892,
+      "learning_rate": 0.00013746489660230288,
+      "loss": 1.0668,
+      "step": 8547
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29179871521386425,
+      "learning_rate": 0.00013745022882807127,
+      "loss": 1.1174,
+      "step": 8548
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.310921629295036,
+      "learning_rate": 0.00013743556011662413,
+      "loss": 1.0533,
+      "step": 8549
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2538334627557585,
+      "learning_rate": 0.00013742089046832855,
+      "loss": 0.994,
+      "step": 8550
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2919456298279374,
+      "learning_rate": 0.00013740621988355168,
+      "loss": 1.0132,
+      "step": 8551
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2532952808062169,
+      "learning_rate": 0.00013739154836266064,
+      "loss": 0.9537,
+      "step": 8552
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31354688802504277,
+      "learning_rate": 0.0001373768759060226,
+      "loss": 1.1491,
+      "step": 8553
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2485754538763557,
+      "learning_rate": 0.00013736220251400478,
+      "loss": 1.1273,
+      "step": 8554
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29503224281382395,
+      "learning_rate": 0.00013734752818697434,
+      "loss": 1.0555,
+      "step": 8555
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2671248855695662,
+      "learning_rate": 0.00013733285292529855,
+      "loss": 1.001,
+      "step": 8556
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3018392538404123,
+      "learning_rate": 0.00013731817672934463,
+      "loss": 1.1354,
+      "step": 8557
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.297525745763971,
+      "learning_rate": 0.0001373034995994799,
+      "loss": 1.1386,
+      "step": 8558
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3031989211002846,
+      "learning_rate": 0.00013728882153607165,
+      "loss": 1.0058,
+      "step": 8559
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2925383632398921,
+      "learning_rate": 0.00013727414253948719,
+      "loss": 0.9204,
+      "step": 8560
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28581280375824064,
+      "learning_rate": 0.0001372594626100939,
+      "loss": 1.1115,
+      "step": 8561
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2849917863325391,
+      "learning_rate": 0.00013724478174825916,
+      "loss": 1.0351,
+      "step": 8562
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2867357667558398,
+      "learning_rate": 0.0001372300999543503,
+      "loss": 0.9713,
+      "step": 8563
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2912838079543734,
+      "learning_rate": 0.00013721541722873484,
+      "loss": 1.0435,
+      "step": 8564
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3121331293675006,
+      "learning_rate": 0.00013720073357178017,
+      "loss": 1.1627,
+      "step": 8565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29971607543863277,
+      "learning_rate": 0.00013718604898385375,
+      "loss": 1.1364,
+      "step": 8566
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2740375080807219,
+      "learning_rate": 0.00013717136346532306,
+      "loss": 1.0376,
+      "step": 8567
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26055014984233066,
+      "learning_rate": 0.00013715667701655565,
+      "loss": 0.9892,
+      "step": 8568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28985982137880956,
+      "learning_rate": 0.00013714198963791908,
+      "loss": 1.0258,
+      "step": 8569
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24153544179134467,
+      "learning_rate": 0.00013712730132978083,
+      "loss": 1.0979,
+      "step": 8570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2819940528233919,
+      "learning_rate": 0.00013711261209250857,
+      "loss": 1.0627,
+      "step": 8571
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3014061072273147,
+      "learning_rate": 0.00013709792192646985,
+      "loss": 1.0036,
+      "step": 8572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2943687582185628,
+      "learning_rate": 0.00013708323083203228,
+      "loss": 1.1475,
+      "step": 8573
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29418992771044433,
+      "learning_rate": 0.0001370685388095636,
+      "loss": 1.0402,
+      "step": 8574
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3135438299333758,
+      "learning_rate": 0.00013705384585943145,
+      "loss": 1.1125,
+      "step": 8575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28733629193402527,
+      "learning_rate": 0.00013703915198200347,
+      "loss": 1.0161,
+      "step": 8576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29180260421642634,
+      "learning_rate": 0.00013702445717764746,
+      "loss": 1.0476,
+      "step": 8577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2980601220397487,
+      "learning_rate": 0.00013700976144673116,
+      "loss": 1.0602,
+      "step": 8578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3066488522260643,
+      "learning_rate": 0.00013699506478962231,
+      "loss": 1.0332,
+      "step": 8579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3353162200492158,
+      "learning_rate": 0.00013698036720668873,
+      "loss": 0.9495,
+      "step": 8580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2816337655551374,
+      "learning_rate": 0.00013696566869829816,
+      "loss": 1.1274,
+      "step": 8581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3100139223925439,
+      "learning_rate": 0.00013695096926481855,
+      "loss": 1.0753,
+      "step": 8582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2746129531995423,
+      "learning_rate": 0.0001369362689066177,
+      "loss": 1.0354,
+      "step": 8583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2697359432512603,
+      "learning_rate": 0.00013692156762406347,
+      "loss": 0.9328,
+      "step": 8584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.290636331567875,
+      "learning_rate": 0.00013690686541752384,
+      "loss": 1.0889,
+      "step": 8585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28720885198550766,
+      "learning_rate": 0.0001368921622873667,
+      "loss": 1.1372,
+      "step": 8586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30775748112991347,
+      "learning_rate": 0.00013687745823396007,
+      "loss": 1.1094,
+      "step": 8587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2624465157124863,
+      "learning_rate": 0.0001368627532576718,
+      "loss": 0.9509,
+      "step": 8588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.26983758242587186,
+      "learning_rate": 0.00013684804735887,
+      "loss": 1.0852,
+      "step": 8589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3019798930382428,
+      "learning_rate": 0.00013683334053792262,
+      "loss": 1.0498,
+      "step": 8590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28333579655941465,
+      "learning_rate": 0.00013681863279519776,
+      "loss": 1.0397,
+      "step": 8591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27677050640426154,
+      "learning_rate": 0.0001368039241310635,
+      "loss": 1.0302,
+      "step": 8592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2740272283421442,
+      "learning_rate": 0.00013678921454588787,
+      "loss": 0.9894,
+      "step": 8593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30385602594318123,
+      "learning_rate": 0.00013677450404003905,
+      "loss": 0.9967,
+      "step": 8594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2543360042755725,
+      "learning_rate": 0.0001367597926138851,
+      "loss": 1.1073,
+      "step": 8595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30243436611460256,
+      "learning_rate": 0.0001367450802677943,
+      "loss": 1.0471,
+      "step": 8596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.298278417963026,
+      "learning_rate": 0.00013673036700213476,
+      "loss": 1.108,
+      "step": 8597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.344466314776565,
+      "learning_rate": 0.0001367156528172747,
+      "loss": 1.0931,
+      "step": 8598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2597226647292173,
+      "learning_rate": 0.00013670093771358234,
+      "loss": 1.0871,
+      "step": 8599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.24557911295793564,
+      "learning_rate": 0.00013668622169142597,
+      "loss": 1.0619,
+      "step": 8600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2747878102964318,
+      "learning_rate": 0.00013667150475117382,
+      "loss": 1.1528,
+      "step": 8601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2853209445830605,
+      "learning_rate": 0.00013665678689319424,
+      "loss": 1.0169,
+      "step": 8602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29466027746832646,
+      "learning_rate": 0.00013664206811785554,
+      "loss": 1.0538,
+      "step": 8603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28940310201790015,
+      "learning_rate": 0.000136627348425526,
+      "loss": 0.9515,
+      "step": 8604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28401605740381525,
+      "learning_rate": 0.0001366126278165741,
+      "loss": 1.1516,
+      "step": 8605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2740954715279189,
+      "learning_rate": 0.00013659790629136817,
+      "loss": 1.0099,
+      "step": 8606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2874894186741719,
+      "learning_rate": 0.00013658318385027665,
+      "loss": 0.9993,
+      "step": 8607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2673309894002083,
+      "learning_rate": 0.0001365684604936679,
+      "loss": 1.1178,
+      "step": 8608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.28060184644025465,
+      "learning_rate": 0.0001365537362219105,
+      "loss": 1.1111,
+      "step": 8609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.262926201140649,
+      "learning_rate": 0.00013653901103537287,
+      "loss": 1.0677,
+      "step": 8610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2603806384626445,
+      "learning_rate": 0.0001365242849344235,
+      "loss": 1.0661,
+      "step": 8611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27821013084562063,
+      "learning_rate": 0.00013650955791943097,
+      "loss": 1.1453,
+      "step": 8612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3025962932247498,
+      "learning_rate": 0.0001364948299907638,
+      "loss": 1.2105,
+      "step": 8613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.27742930612541283,
+      "learning_rate": 0.00013648010114879056,
+      "loss": 0.9934,
+      "step": 8614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2436905652104897,
+      "learning_rate": 0.0001364653713938799,
+      "loss": 0.98,
+      "step": 8615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.308431748553756,
+      "learning_rate": 0.00013645064072640036,
+      "loss": 1.1859,
+      "step": 8616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.32785207627829216,
+      "learning_rate": 0.00013643590914672065,
+      "loss": 1.0128,
+      "step": 8617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3089734401683715,
+      "learning_rate": 0.00013642117665520938,
+      "loss": 1.1553,
+      "step": 8618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3077673526631775,
+      "learning_rate": 0.0001364064432522353,
+      "loss": 1.0379,
+      "step": 8619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.29158179240095367,
+      "learning_rate": 0.00013639170893816713,
+      "loss": 1.0954,
+      "step": 8620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2899047445443715,
+      "learning_rate": 0.00013637697371337353,
+      "loss": 1.0493,
+      "step": 8621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2409967915205593,
+      "learning_rate": 0.0001363622375782233,
+      "loss": 1.0478,
+      "step": 8622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.294664521205138,
+      "learning_rate": 0.00013634750053308524,
+      "loss": 1.1147,
+      "step": 8623
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2691622747456586,
+      "learning_rate": 0.00013633276257832814,
+      "loss": 1.0245,
+      "step": 8624
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2662427740875133,
+      "learning_rate": 0.0001363180237143208,
+      "loss": 1.1204,
+      "step": 8625
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.315565873248549,
+      "learning_rate": 0.00013630328394143213,
+      "loss": 1.0977,
+      "step": 8626
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2958160614252357,
+      "learning_rate": 0.00013628854326003093,
+      "loss": 1.0883,
+      "step": 8627
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27828748808814535,
+      "learning_rate": 0.00013627380167048614,
+      "loss": 0.9,
+      "step": 8628
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2605213690147047,
+      "learning_rate": 0.00013625905917316665,
+      "loss": 1.1766,
+      "step": 8629
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28580811147831237,
+      "learning_rate": 0.00013624431576844144,
+      "loss": 1.142,
+      "step": 8630
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2645180239308723,
+      "learning_rate": 0.00013622957145667945,
+      "loss": 1.0222,
+      "step": 8631
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27332832418491226,
+      "learning_rate": 0.00013621482623824965,
+      "loss": 1.1453,
+      "step": 8632
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24290270100131645,
+      "learning_rate": 0.00013620008011352105,
+      "loss": 1.0981,
+      "step": 8633
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2738275452675957,
+      "learning_rate": 0.0001361853330828627,
+      "loss": 1.087,
+      "step": 8634
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2973345148003589,
+      "learning_rate": 0.00013617058514664367,
+      "loss": 1.0528,
+      "step": 8635
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30243238138049405,
+      "learning_rate": 0.000136155836305233,
+      "loss": 1.1979,
+      "step": 8636
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2973554678344532,
+      "learning_rate": 0.0001361410865589998,
+      "loss": 1.1496,
+      "step": 8637
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27907427949892327,
+      "learning_rate": 0.00013612633590831319,
+      "loss": 1.1112,
+      "step": 8638
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2546823283144357,
+      "learning_rate": 0.00013611158435354232,
+      "loss": 0.9379,
+      "step": 8639
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24491672704420625,
+      "learning_rate": 0.0001360968318950564,
+      "loss": 1.1017,
+      "step": 8640
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2397150988377803,
+      "learning_rate": 0.00013608207853322454,
+      "loss": 0.9593,
+      "step": 8641
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27419817711093586,
+      "learning_rate": 0.00013606732426841596,
+      "loss": 0.937,
+      "step": 8642
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3159034039720681,
+      "learning_rate": 0.00013605256910099997,
+      "loss": 1.1914,
+      "step": 8643
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27134589534535863,
+      "learning_rate": 0.00013603781303134576,
+      "loss": 1.1164,
+      "step": 8644
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3025553917420271,
+      "learning_rate": 0.00013602305605982262,
+      "loss": 1.0479,
+      "step": 8645
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27129927828178074,
+      "learning_rate": 0.0001360082981867999,
+      "loss": 1.1052,
+      "step": 8646
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31517001904029224,
+      "learning_rate": 0.00013599353941264684,
+      "loss": 1.0203,
+      "step": 8647
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2851479531322408,
+      "learning_rate": 0.0001359787797377329,
+      "loss": 1.0592,
+      "step": 8648
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2856016662841958,
+      "learning_rate": 0.00013596401916242732,
+      "loss": 1.0729,
+      "step": 8649
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2847359988614896,
+      "learning_rate": 0.00013594925768709959,
+      "loss": 1.0519,
+      "step": 8650
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2656376058979404,
+      "learning_rate": 0.00013593449531211908,
+      "loss": 1.0247,
+      "step": 8651
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2591314203645854,
+      "learning_rate": 0.00013591973203785524,
+      "loss": 1.099,
+      "step": 8652
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28779350087950495,
+      "learning_rate": 0.00013590496786467754,
+      "loss": 1.054,
+      "step": 8653
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2546447105719811,
+      "learning_rate": 0.00013589020279295544,
+      "loss": 1.0369,
+      "step": 8654
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33112794737534357,
+      "learning_rate": 0.00013587543682305847,
+      "loss": 1.0239,
+      "step": 8655
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.261490090551106,
+      "learning_rate": 0.00013586066995535616,
+      "loss": 1.0201,
+      "step": 8656
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2880468769072062,
+      "learning_rate": 0.000135845902190218,
+      "loss": 1.1193,
+      "step": 8657
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.288557868149757,
+      "learning_rate": 0.00013583113352801367,
+      "loss": 1.1048,
+      "step": 8658
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29517378742374495,
+      "learning_rate": 0.00013581636396911266,
+      "loss": 1.0905,
+      "step": 8659
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2617708019113313,
+      "learning_rate": 0.00013580159351388464,
+      "loss": 1.0682,
+      "step": 8660
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31436033052025997,
+      "learning_rate": 0.00013578682216269927,
+      "loss": 0.9827,
+      "step": 8661
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2710558291831824,
+      "learning_rate": 0.00013577204991592617,
+      "loss": 1.0678,
+      "step": 8662
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30589524951652153,
+      "learning_rate": 0.000135757276773935,
+      "loss": 1.0166,
+      "step": 8663
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31535012911315774,
+      "learning_rate": 0.00013574250273709555,
+      "loss": 1.0809,
+      "step": 8664
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.34357963327522195,
+      "learning_rate": 0.0001357277278057775,
+      "loss": 1.0025,
+      "step": 8665
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29336089111450286,
+      "learning_rate": 0.0001357129519803506,
+      "loss": 1.14,
+      "step": 8666
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2988503565133178,
+      "learning_rate": 0.00013569817526118465,
+      "loss": 0.9728,
+      "step": 8667
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2586925578598481,
+      "learning_rate": 0.0001356833976486494,
+      "loss": 1.0163,
+      "step": 8668
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2937963873608329,
+      "learning_rate": 0.0001356686191431147,
+      "loss": 1.0555,
+      "step": 8669
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29571540658867496,
+      "learning_rate": 0.0001356538397449504,
+      "loss": 1.0379,
+      "step": 8670
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27868553879152896,
+      "learning_rate": 0.00013563905945452638,
+      "loss": 1.0352,
+      "step": 8671
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29105897745078474,
+      "learning_rate": 0.00013562427827221244,
+      "loss": 0.993,
+      "step": 8672
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24758183366141243,
+      "learning_rate": 0.0001356094961983786,
+      "loss": 0.8968,
+      "step": 8673
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2605888157367459,
+      "learning_rate": 0.0001355947132333947,
+      "loss": 1.0564,
+      "step": 8674
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2924270653891825,
+      "learning_rate": 0.00013557992937763077,
+      "loss": 1.0911,
+      "step": 8675
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3153612885581894,
+      "learning_rate": 0.00013556514463145672,
+      "loss": 1.0308,
+      "step": 8676
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27974012653020464,
+      "learning_rate": 0.00013555035899524257,
+      "loss": 1.0977,
+      "step": 8677
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2686146240839227,
+      "learning_rate": 0.00013553557246935834,
+      "loss": 0.9855,
+      "step": 8678
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2809894346574622,
+      "learning_rate": 0.00013552078505417412,
+      "loss": 1.1707,
+      "step": 8679
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2869178669204025,
+      "learning_rate": 0.00013550599675005986,
+      "loss": 1.1491,
+      "step": 8680
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2955879869986699,
+      "learning_rate": 0.00013549120755738576,
+      "loss": 1.0608,
+      "step": 8681
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2875101038157737,
+      "learning_rate": 0.00013547641747652187,
+      "loss": 1.0307,
+      "step": 8682
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28175365232286614,
+      "learning_rate": 0.00013546162650783836,
+      "loss": 1.0323,
+      "step": 8683
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27831603189171494,
+      "learning_rate": 0.00013544683465170537,
+      "loss": 1.057,
+      "step": 8684
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3073334793522284,
+      "learning_rate": 0.00013543204190849303,
+      "loss": 1.0045,
+      "step": 8685
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29215172333875483,
+      "learning_rate": 0.00013541724827857157,
+      "loss": 1.1192,
+      "step": 8686
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2771543039799703,
+      "learning_rate": 0.00013540245376231122,
+      "loss": 1.1609,
+      "step": 8687
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29777594479305014,
+      "learning_rate": 0.00013538765836008224,
+      "loss": 0.9681,
+      "step": 8688
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2545360895510685,
+      "learning_rate": 0.00013537286207225484,
+      "loss": 1.0139,
+      "step": 8689
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26647845539942555,
+      "learning_rate": 0.00013535806489919935,
+      "loss": 1.1347,
+      "step": 8690
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.29722399863436927,
+      "learning_rate": 0.00013534326684128605,
+      "loss": 1.1584,
+      "step": 8691
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30891387675873,
+      "learning_rate": 0.00013532846789888532,
+      "loss": 1.0211,
+      "step": 8692
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31589225263631776,
+      "learning_rate": 0.00013531366807236742,
+      "loss": 1.0062,
+      "step": 8693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2972943846537337,
+      "learning_rate": 0.00013529886736210285,
+      "loss": 1.0088,
+      "step": 8694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28937720201774464,
+      "learning_rate": 0.00013528406576846189,
+      "loss": 0.9853,
+      "step": 8695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24765821705508315,
+      "learning_rate": 0.000135269263291815,
+      "loss": 0.9191,
+      "step": 8696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27414169282442363,
+      "learning_rate": 0.00013525445993253267,
+      "loss": 1.0309,
+      "step": 8697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2688265567951683,
+      "learning_rate": 0.0001352396556909853,
+      "loss": 1.0866,
+      "step": 8698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.26180708297498495,
+      "learning_rate": 0.0001352248505675434,
+      "loss": 1.0702,
+      "step": 8699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.296931683455755,
+      "learning_rate": 0.00013521004456257748,
+      "loss": 1.0738,
+      "step": 8700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2518973661220569,
+      "learning_rate": 0.0001351952376764581,
+      "loss": 1.0174,
+      "step": 8701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30852287173282533,
+      "learning_rate": 0.00013518042990955575,
+      "loss": 1.0339,
+      "step": 8702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28709043823004965,
+      "learning_rate": 0.000135165621262241,
+      "loss": 1.127,
+      "step": 8703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2402225336023348,
+      "learning_rate": 0.00013515081173488453,
+      "loss": 1.0512,
+      "step": 8704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3009799139239725,
+      "learning_rate": 0.00013513600132785688,
+      "loss": 0.995,
+      "step": 8705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31841563271518275,
+      "learning_rate": 0.0001351211900415287,
+      "loss": 1.1342,
+      "step": 8706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2726227896301979,
+      "learning_rate": 0.00013510637787627068,
+      "loss": 1.1005,
+      "step": 8707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30718818182896745,
+      "learning_rate": 0.0001350915648324535,
+      "loss": 0.9629,
+      "step": 8708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.24754498668980748,
+      "learning_rate": 0.00013507675091044787,
+      "loss": 1.1344,
+      "step": 8709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3341325246272737,
+      "learning_rate": 0.00013506193611062444,
+      "loss": 1.0433,
+      "step": 8710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2831390747623922,
+      "learning_rate": 0.0001350471204333541,
+      "loss": 0.9654,
+      "step": 8711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27958419761160536,
+      "learning_rate": 0.0001350323038790075,
+      "loss": 1.0065,
+      "step": 8712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3219792061230287,
+      "learning_rate": 0.00013501748644795548,
+      "loss": 1.008,
+      "step": 8713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27010290849276297,
+      "learning_rate": 0.00013500266814056886,
+      "loss": 0.9119,
+      "step": 8714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3035143869003971,
+      "learning_rate": 0.0001349878489572185,
+      "loss": 1.0691,
+      "step": 8715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27992327647557635,
+      "learning_rate": 0.0001349730288982752,
+      "loss": 1.067,
+      "step": 8716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.31109391121037,
+      "learning_rate": 0.00013495820796410987,
+      "loss": 1.0007,
+      "step": 8717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30367401228401525,
+      "learning_rate": 0.00013494338615509344,
+      "loss": 1.1086,
+      "step": 8718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27392132119871376,
+      "learning_rate": 0.00013492856347159678,
+      "loss": 0.9861,
+      "step": 8719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.318345859783819,
+      "learning_rate": 0.00013491373991399088,
+      "loss": 1.042,
+      "step": 8720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3354485387598985,
+      "learning_rate": 0.0001348989154826467,
+      "loss": 1.1203,
+      "step": 8721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3027876660339266,
+      "learning_rate": 0.0001348840901779352,
+      "loss": 1.1517,
+      "step": 8722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3025089370772015,
+      "learning_rate": 0.00013486926400022744,
+      "loss": 1.2937,
+      "step": 8723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.27699234873722095,
+      "learning_rate": 0.00013485443694989443,
+      "loss": 1.0503,
+      "step": 8724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30429620705081767,
+      "learning_rate": 0.00013483960902730725,
+      "loss": 1.0699,
+      "step": 8725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3040291893430194,
+      "learning_rate": 0.00013482478023283694,
+      "loss": 1.0117,
+      "step": 8726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28916049582699416,
+      "learning_rate": 0.00013480995056685462,
+      "loss": 1.0908,
+      "step": 8727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24217865268899172,
+      "learning_rate": 0.00013479512002973143,
+      "loss": 1.056,
+      "step": 8728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27032893449561624,
+      "learning_rate": 0.00013478028862183846,
+      "loss": 1.0668,
+      "step": 8729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27542193970843054,
+      "learning_rate": 0.00013476545634354692,
+      "loss": 0.9492,
+      "step": 8730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23543044415563721,
+      "learning_rate": 0.000134750623195228,
+      "loss": 1.0316,
+      "step": 8731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3223797172938787,
+      "learning_rate": 0.0001347357891772529,
+      "loss": 1.1393,
+      "step": 8732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2998222335002957,
+      "learning_rate": 0.0001347209542899928,
+      "loss": 1.1096,
+      "step": 8733
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2845913693982625,
+      "learning_rate": 0.00013470611853381905,
+      "loss": 1.0304,
+      "step": 8734
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.312734748694851,
+      "learning_rate": 0.00013469128190910285,
+      "loss": 1.0816,
+      "step": 8735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2784778983582345,
+      "learning_rate": 0.00013467644441621552,
+      "loss": 0.9477,
+      "step": 8736
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.294746691747119,
+      "learning_rate": 0.00013466160605552836,
+      "loss": 1.1006,
+      "step": 8737
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2729555754078699,
+      "learning_rate": 0.00013464676682741275,
+      "loss": 1.165,
+      "step": 8738
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27914582732602733,
+      "learning_rate": 0.00013463192673223998,
+      "loss": 1.0035,
+      "step": 8739
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2759299004442835,
+      "learning_rate": 0.00013461708577038154,
+      "loss": 1.1369,
+      "step": 8740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25997588040826375,
+      "learning_rate": 0.00013460224394220871,
+      "loss": 1.0008,
+      "step": 8741
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3006336454437912,
+      "learning_rate": 0.00013458740124809302,
+      "loss": 1.0918,
+      "step": 8742
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2865486476111894,
+      "learning_rate": 0.00013457255768840586,
+      "loss": 1.0318,
+      "step": 8743
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2610230590960148,
+      "learning_rate": 0.00013455771326351874,
+      "loss": 1.0608,
+      "step": 8744
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2747856141015647,
+      "learning_rate": 0.0001345428679738031,
+      "loss": 0.9236,
+      "step": 8745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.250840631667433,
+      "learning_rate": 0.0001345280218196305,
+      "loss": 1.076,
+      "step": 8746
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3427774684297268,
+      "learning_rate": 0.0001345131748013724,
+      "loss": 1.0931,
+      "step": 8747
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30101078325665653,
+      "learning_rate": 0.0001344983269194005,
+      "loss": 0.9994,
+      "step": 8748
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.285861675215507,
+      "learning_rate": 0.00013448347817408623,
+      "loss": 1.0467,
+      "step": 8749
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28454224361070146,
+      "learning_rate": 0.00013446862856580127,
+      "loss": 1.0008,
+      "step": 8750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33159442196308175,
+      "learning_rate": 0.0001344537780949172,
+      "loss": 1.206,
+      "step": 8751
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28094929027693405,
+      "learning_rate": 0.0001344389267618057,
+      "loss": 1.182,
+      "step": 8752
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2904531787454453,
+      "learning_rate": 0.0001344240745668384,
+      "loss": 0.982,
+      "step": 8753
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3124634413609586,
+      "learning_rate": 0.00013440922151038698,
+      "loss": 1.0623,
+      "step": 8754
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28809714187038504,
+      "learning_rate": 0.0001343943675928232,
+      "loss": 1.1198,
+      "step": 8755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2873805808293868,
+      "learning_rate": 0.00013437951281451875,
+      "loss": 1.0509,
+      "step": 8756
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3130611893567796,
+      "learning_rate": 0.00013436465717584533,
+      "loss": 1.0459,
+      "step": 8757
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3002619940368619,
+      "learning_rate": 0.00013434980067717484,
+      "loss": 0.9741,
+      "step": 8758
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29994300371561994,
+      "learning_rate": 0.00013433494331887896,
+      "loss": 1.0306,
+      "step": 8759
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2709309546499614,
+      "learning_rate": 0.00013432008510132955,
+      "loss": 1.0472,
+      "step": 8760
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30275123120474723,
+      "learning_rate": 0.00013430522602489846,
+      "loss": 1.0934,
+      "step": 8761
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27267018063077797,
+      "learning_rate": 0.0001342903660899575,
+      "loss": 1.0372,
+      "step": 8762
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3179594835836974,
+      "learning_rate": 0.0001342755052968786,
+      "loss": 1.0563,
+      "step": 8763
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31426468902967697,
+      "learning_rate": 0.0001342606436460336,
+      "loss": 1.1932,
+      "step": 8764
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2715991486632539,
+      "learning_rate": 0.00013424578113779452,
+      "loss": 1.0489,
+      "step": 8765
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31900892380705087,
+      "learning_rate": 0.00013423091777253323,
+      "loss": 1.109,
+      "step": 8766
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2661697001151372,
+      "learning_rate": 0.0001342160535506217,
+      "loss": 1.0726,
+      "step": 8767
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29001722553751685,
+      "learning_rate": 0.00013420118847243191,
+      "loss": 1.0504,
+      "step": 8768
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26543562475469634,
+      "learning_rate": 0.0001341863225383359,
+      "loss": 1.0057,
+      "step": 8769
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30664308972274495,
+      "learning_rate": 0.0001341714557487057,
+      "loss": 1.0982,
+      "step": 8770
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29541897062886224,
+      "learning_rate": 0.0001341565881039133,
+      "loss": 1.0859,
+      "step": 8771
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.280865811627392,
+      "learning_rate": 0.00013414171960433085,
+      "loss": 1.119,
+      "step": 8772
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24611125528299096,
+      "learning_rate": 0.00013412685025033038,
+      "loss": 1.1245,
+      "step": 8773
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31141735928913117,
+      "learning_rate": 0.00013411198004228405,
+      "loss": 1.0752,
+      "step": 8774
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.23851136770012923,
+      "learning_rate": 0.000134097108980564,
+      "loss": 1.0644,
+      "step": 8775
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2371811186751139,
+      "learning_rate": 0.00013408223706554235,
+      "loss": 1.0813,
+      "step": 8776
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2987397006710603,
+      "learning_rate": 0.0001340673642975913,
+      "loss": 1.0765,
+      "step": 8777
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26831922729943924,
+      "learning_rate": 0.00013405249067708304,
+      "loss": 1.0317,
+      "step": 8778
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2899859345363733,
+      "learning_rate": 0.00013403761620438983,
+      "loss": 1.0402,
+      "step": 8779
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2878652206230178,
+      "learning_rate": 0.00013402274087988384,
+      "loss": 0.9933,
+      "step": 8780
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31702536400845066,
+      "learning_rate": 0.0001340078647039374,
+      "loss": 1.0253,
+      "step": 8781
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30001278045384056,
+      "learning_rate": 0.00013399298767692277,
+      "loss": 1.123,
+      "step": 8782
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2624717683492848,
+      "learning_rate": 0.00013397810979921227,
+      "loss": 1.021,
+      "step": 8783
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.316080617053837,
+      "learning_rate": 0.0001339632310711782,
+      "loss": 1.133,
+      "step": 8784
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27598469027777045,
+      "learning_rate": 0.00013394835149319292,
+      "loss": 0.8538,
+      "step": 8785
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29532824662066726,
+      "learning_rate": 0.00013393347106562884,
+      "loss": 1.0311,
+      "step": 8786
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3293468526276591,
+      "learning_rate": 0.00013391858978885828,
+      "loss": 1.0673,
+      "step": 8787
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2643309414987189,
+      "learning_rate": 0.00013390370766325373,
+      "loss": 1.1561,
+      "step": 8788
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2763364535801901,
+      "learning_rate": 0.00013388882468918758,
+      "loss": 1.0627,
+      "step": 8789
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2862222208420642,
+      "learning_rate": 0.0001338739408670323,
+      "loss": 1.0606,
+      "step": 8790
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2927406980062334,
+      "learning_rate": 0.00013385905619716032,
+      "loss": 1.0987,
+      "step": 8791
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27073771357355814,
+      "learning_rate": 0.00013384417067994423,
+      "loss": 1.0864,
+      "step": 8792
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26681668946322573,
+      "learning_rate": 0.00013382928431575648,
+      "loss": 0.9944,
+      "step": 8793
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2872926853091687,
+      "learning_rate": 0.00013381439710496962,
+      "loss": 1.1929,
+      "step": 8794
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29473601888234835,
+      "learning_rate": 0.00013379950904795625,
+      "loss": 1.0424,
+      "step": 8795
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2767351991760329,
+      "learning_rate": 0.0001337846201450889,
+      "loss": 0.976,
+      "step": 8796
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31278846280951567,
+      "learning_rate": 0.00013376973039674019,
+      "loss": 1.0786,
+      "step": 8797
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29106202302244005,
+      "learning_rate": 0.00013375483980328275,
+      "loss": 1.0206,
+      "step": 8798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3062557626049918,
+      "learning_rate": 0.00013373994836508925,
+      "loss": 1.0934,
+      "step": 8799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2930477172648966,
+      "learning_rate": 0.00013372505608253235,
+      "loss": 1.0262,
+      "step": 8800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.34634712769514453,
+      "learning_rate": 0.0001337101629559847,
+      "loss": 0.9546,
+      "step": 8801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2772018365006181,
+      "learning_rate": 0.00013369526898581902,
+      "loss": 1.0512,
+      "step": 8802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2709787690909469,
+      "learning_rate": 0.00013368037417240807,
+      "loss": 0.994,
+      "step": 8803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2814766172965834,
+      "learning_rate": 0.0001336654785161246,
+      "loss": 1.0845,
+      "step": 8804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2853983711193067,
+      "learning_rate": 0.00013365058201734135,
+      "loss": 0.9669,
+      "step": 8805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3256691908202577,
+      "learning_rate": 0.00013363568467643117,
+      "loss": 1.0751,
+      "step": 8806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2878851657217033,
+      "learning_rate": 0.00013362078649376683,
+      "loss": 1.045,
+      "step": 8807
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30585377114209605,
+      "learning_rate": 0.00013360588746972118,
+      "loss": 1.0328,
+      "step": 8808
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2793018113302156,
+      "learning_rate": 0.00013359098760466707,
+      "loss": 1.123,
+      "step": 8809
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.25457785398487776,
+      "learning_rate": 0.0001335760868989774,
+      "loss": 1.1305,
+      "step": 8810
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.26931250840260823,
+      "learning_rate": 0.00013356118535302503,
+      "loss": 1.1142,
+      "step": 8811
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2779912568855338,
+      "learning_rate": 0.00013354628296718293,
+      "loss": 1.0774,
+      "step": 8812
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30698926954918604,
+      "learning_rate": 0.000133531379741824,
+      "loss": 1.1206,
+      "step": 8813
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28571250644527174,
+      "learning_rate": 0.0001335164756773212,
+      "loss": 1.1259,
+      "step": 8814
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30864009381335133,
+      "learning_rate": 0.00013350157077404755,
+      "loss": 0.9197,
+      "step": 8815
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2714746153198062,
+      "learning_rate": 0.00013348666503237603,
+      "loss": 0.9842,
+      "step": 8816
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.28173495612303234,
+      "learning_rate": 0.0001334717584526797,
+      "loss": 0.9923,
+      "step": 8817
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3216688395468582,
+      "learning_rate": 0.00013345685103533154,
+      "loss": 1.0964,
+      "step": 8818
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2823984993319558,
+      "learning_rate": 0.00013344194278070467,
+      "loss": 1.0409,
+      "step": 8819
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.257746108679655,
+      "learning_rate": 0.00013342703368917217,
+      "loss": 1.1301,
+      "step": 8820
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3051717861148401,
+      "learning_rate": 0.00013341212376110715,
+      "loss": 1.0672,
+      "step": 8821
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.27852339454292335,
+      "learning_rate": 0.00013339721299688272,
+      "loss": 0.9949,
+      "step": 8822
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30905296931037346,
+      "learning_rate": 0.00013338230139687206,
+      "loss": 1.0062,
+      "step": 8823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.24655148826104478,
+      "learning_rate": 0.0001333673889614483,
+      "loss": 1.001,
+      "step": 8824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30269701257489295,
+      "learning_rate": 0.00013335247569098467,
+      "loss": 1.0285,
+      "step": 8825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29912297649204367,
+      "learning_rate": 0.00013333756158585437,
+      "loss": 1.1004,
+      "step": 8826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29035499574876744,
+      "learning_rate": 0.00013332264664643067,
+      "loss": 1.053,
+      "step": 8827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2869996333200353,
+      "learning_rate": 0.00013330773087308676,
+      "loss": 1.0357,
+      "step": 8828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29232380763052546,
+      "learning_rate": 0.00013329281426619597,
+      "loss": 1.0288,
+      "step": 8829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2495577755639189,
+      "learning_rate": 0.0001332778968261316,
+      "loss": 1.0611,
+      "step": 8830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2979321424581524,
+      "learning_rate": 0.0001332629785532669,
+      "loss": 0.9719,
+      "step": 8831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2821090849292814,
+      "learning_rate": 0.0001332480594479753,
+      "loss": 1.1135,
+      "step": 8832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3016300483432541,
+      "learning_rate": 0.0001332331395106301,
+      "loss": 1.1395,
+      "step": 8833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3017945019892449,
+      "learning_rate": 0.00013321821874160472,
+      "loss": 0.9148,
+      "step": 8834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25675693104083414,
+      "learning_rate": 0.00013320329714127248,
+      "loss": 1.0208,
+      "step": 8835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30879551976248903,
+      "learning_rate": 0.0001331883747100069,
+      "loss": 1.1158,
+      "step": 8836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2910089252086533,
+      "learning_rate": 0.0001331734514481814,
+      "loss": 1.0531,
+      "step": 8837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2800038334110076,
+      "learning_rate": 0.0001331585273561694,
+      "loss": 1.0722,
+      "step": 8838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29386882904498585,
+      "learning_rate": 0.00013314360243434442,
+      "loss": 1.0914,
+      "step": 8839
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32822389242666516,
+      "learning_rate": 0.00013312867668307998,
+      "loss": 1.0664,
+      "step": 8840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28150462602342563,
+      "learning_rate": 0.00013311375010274958,
+      "loss": 1.0505,
+      "step": 8841
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28190761168490724,
+      "learning_rate": 0.00013309882269372676,
+      "loss": 1.0243,
+      "step": 8842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.324588959170944,
+      "learning_rate": 0.00013308389445638508,
+      "loss": 1.1538,
+      "step": 8843
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.313116228461298,
+      "learning_rate": 0.0001330689653910982,
+      "loss": 1.0932,
+      "step": 8844
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3028011762250344,
+      "learning_rate": 0.00013305403549823962,
+      "loss": 1.0032,
+      "step": 8845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2679913864423184,
+      "learning_rate": 0.00013303910477818306,
+      "loss": 1.0489,
+      "step": 8846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.23253134929888603,
+      "learning_rate": 0.00013302417323130214,
+      "loss": 1.0339,
+      "step": 8847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32940341276625706,
+      "learning_rate": 0.00013300924085797052,
+      "loss": 1.0542,
+      "step": 8848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32063577062079657,
+      "learning_rate": 0.0001329943076585619,
+      "loss": 1.029,
+      "step": 8849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30995947199370877,
+      "learning_rate": 0.00013297937363345,
+      "loss": 1.0355,
+      "step": 8850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25561792258910343,
+      "learning_rate": 0.00013296443878300858,
+      "loss": 1.0827,
+      "step": 8851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26801538052648793,
+      "learning_rate": 0.0001329495031076113,
+      "loss": 1.0586,
+      "step": 8852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2728699949209647,
+      "learning_rate": 0.00013293456660763204,
+      "loss": 1.033,
+      "step": 8853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2807693199525417,
+      "learning_rate": 0.00013291962928344456,
+      "loss": 1.0447,
+      "step": 8854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30144064538223736,
+      "learning_rate": 0.00013290469113542264,
+      "loss": 1.0855,
+      "step": 8855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2871386350695731,
+      "learning_rate": 0.00013288975216394015,
+      "loss": 1.045,
+      "step": 8856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2783823690348202,
+      "learning_rate": 0.00013287481236937094,
+      "loss": 1.0184,
+      "step": 8857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28082328661100076,
+      "learning_rate": 0.0001328598717520889,
+      "loss": 1.11,
+      "step": 8858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2537605984201099,
+      "learning_rate": 0.00013284493031246792,
+      "loss": 1.0461,
+      "step": 8859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29693120122561123,
+      "learning_rate": 0.00013282998805088191,
+      "loss": 0.9376,
+      "step": 8860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2736943955631886,
+      "learning_rate": 0.0001328150449677048,
+      "loss": 1.1058,
+      "step": 8861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3420821362503677,
+      "learning_rate": 0.00013280010106331058,
+      "loss": 1.1671,
+      "step": 8862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2989694017306944,
+      "learning_rate": 0.00013278515633807322,
+      "loss": 1.1161,
+      "step": 8863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24965561624077762,
+      "learning_rate": 0.00013277021079236673,
+      "loss": 0.9428,
+      "step": 8864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27560232204493745,
+      "learning_rate": 0.0001327552644265651,
+      "loss": 1.0965,
+      "step": 8865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31916956119622164,
+      "learning_rate": 0.0001327403172410424,
+      "loss": 1.1617,
+      "step": 8866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26752857028568255,
+      "learning_rate": 0.00013272536923617266,
+      "loss": 0.9989,
+      "step": 8867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2883358139977724,
+      "learning_rate": 0.00013271042041233003,
+      "loss": 0.9731,
+      "step": 8868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.33884977680481115,
+      "learning_rate": 0.00013269547076988854,
+      "loss": 1.1386,
+      "step": 8869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32546200445254864,
+      "learning_rate": 0.00013268052030922237,
+      "loss": 0.9609,
+      "step": 8870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2826007467613128,
+      "learning_rate": 0.00013266556903070563,
+      "loss": 1.0233,
+      "step": 8871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2778261537338506,
+      "learning_rate": 0.00013265061693471246,
+      "loss": 1.0323,
+      "step": 8872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26913753452272493,
+      "learning_rate": 0.00013263566402161713,
+      "loss": 0.9943,
+      "step": 8873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26448807283032155,
+      "learning_rate": 0.0001326207102917938,
+      "loss": 1.0572,
+      "step": 8874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3046847456760344,
+      "learning_rate": 0.00013260575574561666,
+      "loss": 1.0784,
+      "step": 8875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.297455531468288,
+      "learning_rate": 0.00013259080038345998,
+      "loss": 1.1246,
+      "step": 8876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31150901502574146,
+      "learning_rate": 0.0001325758442056981,
+      "loss": 1.1541,
+      "step": 8877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32499878866127807,
+      "learning_rate": 0.00013256088721270518,
+      "loss": 1.1098,
+      "step": 8878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30353184520322335,
+      "learning_rate": 0.00013254592940485562,
+      "loss": 1.1114,
+      "step": 8879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2765014106506231,
+      "learning_rate": 0.00013253097078252374,
+      "loss": 1.0607,
+      "step": 8880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3232413136845846,
+      "learning_rate": 0.00013251601134608385,
+      "loss": 1.0267,
+      "step": 8881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2576596041228671,
+      "learning_rate": 0.00013250105109591034,
+      "loss": 1.1395,
+      "step": 8882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.299087343065221,
+      "learning_rate": 0.00013248609003237762,
+      "loss": 1.0855,
+      "step": 8883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31373524675219927,
+      "learning_rate": 0.00013247112815586008,
+      "loss": 1.0168,
+      "step": 8884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29307123636308,
+      "learning_rate": 0.00013245616546673212,
+      "loss": 1.1391,
+      "step": 8885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2892716141501793,
+      "learning_rate": 0.00013244120196536825,
+      "loss": 1.0368,
+      "step": 8886
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30030230235650646,
+      "learning_rate": 0.0001324262376521429,
+      "loss": 0.9953,
+      "step": 8887
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2958420718954474,
+      "learning_rate": 0.00013241127252743056,
+      "loss": 1.0635,
+      "step": 8888
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30426355904923175,
+      "learning_rate": 0.00013239630659160577,
+      "loss": 0.9941,
+      "step": 8889
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26204874988846405,
+      "learning_rate": 0.00013238133984504305,
+      "loss": 1.0014,
+      "step": 8890
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2808476806284864,
+      "learning_rate": 0.00013236637228811695,
+      "loss": 0.9811,
+      "step": 8891
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2613165358148853,
+      "learning_rate": 0.00013235140392120202,
+      "loss": 1.0464,
+      "step": 8892
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2614569034666025,
+      "learning_rate": 0.0001323364347446729,
+      "loss": 1.0346,
+      "step": 8893
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3048725191683244,
+      "learning_rate": 0.00013232146475890415,
+      "loss": 1.1072,
+      "step": 8894
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3051855377395641,
+      "learning_rate": 0.00013230649396427048,
+      "loss": 1.2126,
+      "step": 8895
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2885176846905764,
+      "learning_rate": 0.00013229152236114646,
+      "loss": 1.1485,
+      "step": 8896
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27063874711622693,
+      "learning_rate": 0.0001322765499499068,
+      "loss": 1.0701,
+      "step": 8897
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2650234430261914,
+      "learning_rate": 0.0001322615767309262,
+      "loss": 0.9788,
+      "step": 8898
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3332394548518117,
+      "learning_rate": 0.00013224660270457937,
+      "loss": 1.0501,
+      "step": 8899
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2622983793530935,
+      "learning_rate": 0.00013223162787124104,
+      "loss": 1.0524,
+      "step": 8900
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28082192393839644,
+      "learning_rate": 0.00013221665223128593,
+      "loss": 1.2141,
+      "step": 8901
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30805730414546934,
+      "learning_rate": 0.00013220167578508892,
+      "loss": 1.073,
+      "step": 8902
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2984108714398881,
+      "learning_rate": 0.00013218669853302467,
+      "loss": 1.075,
+      "step": 8903
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.24970867197130167,
+      "learning_rate": 0.0001321717204754681,
+      "loss": 0.9129,
+      "step": 8904
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2976975973384336,
+      "learning_rate": 0.00013215674161279402,
+      "loss": 1.0718,
+      "step": 8905
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26473164124738907,
+      "learning_rate": 0.00013214176194537722,
+      "loss": 1.1786,
+      "step": 8906
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28433983890808473,
+      "learning_rate": 0.00013212678147359267,
+      "loss": 1.1131,
+      "step": 8907
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.268486990998593,
+      "learning_rate": 0.00013211180019781518,
+      "loss": 1.0515,
+      "step": 8908
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2737650744067927,
+      "learning_rate": 0.00013209681811841972,
+      "loss": 1.0995,
+      "step": 8909
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30007514688621423,
+      "learning_rate": 0.00013208183523578124,
+      "loss": 1.0624,
+      "step": 8910
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26052045873934276,
+      "learning_rate": 0.00013206685155027465,
+      "loss": 1.1553,
+      "step": 8911
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27092778505487736,
+      "learning_rate": 0.00013205186706227498,
+      "loss": 1.1842,
+      "step": 8912
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3258021626325885,
+      "learning_rate": 0.00013203688177215714,
+      "loss": 1.081,
+      "step": 8913
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2984371253963151,
+      "learning_rate": 0.00013202189568029625,
+      "loss": 1.0585,
+      "step": 8914
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28356401437541234,
+      "learning_rate": 0.00013200690878706724,
+      "loss": 0.9656,
+      "step": 8915
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27533653430669,
+      "learning_rate": 0.00013199192109284526,
+      "loss": 1.0106,
+      "step": 8916
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28571972152442254,
+      "learning_rate": 0.00013197693259800534,
+      "loss": 0.9763,
+      "step": 8917
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2560488227643934,
+      "learning_rate": 0.0001319619433029226,
+      "loss": 1.182,
+      "step": 8918
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2984853514513387,
+      "learning_rate": 0.00013194695320797214,
+      "loss": 1.1465,
+      "step": 8919
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.25687016849701755,
+      "learning_rate": 0.00013193196231352905,
+      "loss": 0.9482,
+      "step": 8920
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.31648422763460987,
+      "learning_rate": 0.00013191697061996858,
+      "loss": 1.0639,
+      "step": 8921
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2499110549279051,
+      "learning_rate": 0.00013190197812766588,
+      "loss": 1.0542,
+      "step": 8922
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26107365347955813,
+      "learning_rate": 0.00013188698483699608,
+      "loss": 1.0763,
+      "step": 8923
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2890340873975123,
+      "learning_rate": 0.00013187199074833449,
+      "loss": 1.1517,
+      "step": 8924
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.26683546682272963,
+      "learning_rate": 0.00013185699586205628,
+      "loss": 1.1019,
+      "step": 8925
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2751530786283772,
+      "learning_rate": 0.0001318420001785367,
+      "loss": 1.0552,
+      "step": 8926
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2805940758160267,
+      "learning_rate": 0.00013182700369815108,
+      "loss": 1.046,
+      "step": 8927
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2581306634911059,
+      "learning_rate": 0.00013181200642127468,
+      "loss": 0.9718,
+      "step": 8928
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2699851519716227,
+      "learning_rate": 0.00013179700834828282,
+      "loss": 1.1284,
+      "step": 8929
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2735704940152805,
+      "learning_rate": 0.00013178200947955087,
+      "loss": 0.9873,
+      "step": 8930
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.29262357118516796,
+      "learning_rate": 0.00013176700981545414,
+      "loss": 1.052,
+      "step": 8931
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3021541436530508,
+      "learning_rate": 0.00013175200935636804,
+      "loss": 0.9968,
+      "step": 8932
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2994688515381643,
+      "learning_rate": 0.0001317370081026679,
+      "loss": 1.0423,
+      "step": 8933
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.21622128251751344,
+      "learning_rate": 0.00013172200605472925,
+      "loss": 1.0596,
+      "step": 8934
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27357520538102,
+      "learning_rate": 0.00013170700321292746,
+      "loss": 1.0868,
+      "step": 8935
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.32314314647789527,
+      "learning_rate": 0.00013169199957763797,
+      "loss": 1.1866,
+      "step": 8936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27729229797866944,
+      "learning_rate": 0.00013167699514923624,
+      "loss": 1.0354,
+      "step": 8937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2689304675066156,
+      "learning_rate": 0.00013166198992809784,
+      "loss": 1.118,
+      "step": 8938
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.281878748265055,
+      "learning_rate": 0.00013164698391459823,
+      "loss": 1.0886,
+      "step": 8939
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31972214001347354,
+      "learning_rate": 0.00013163197710911294,
+      "loss": 1.0638,
+      "step": 8940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30752537186921586,
+      "learning_rate": 0.00013161696951201755,
+      "loss": 1.0808,
+      "step": 8941
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3169449014565689,
+      "learning_rate": 0.00013160196112368765,
+      "loss": 1.0815,
+      "step": 8942
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.279353465633034,
+      "learning_rate": 0.00013158695194449878,
+      "loss": 0.9856,
+      "step": 8943
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2506893330731972,
+      "learning_rate": 0.00013157194197482662,
+      "loss": 1.0427,
+      "step": 8944
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28379173348711845,
+      "learning_rate": 0.00013155693121504676,
+      "loss": 1.077,
+      "step": 8945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.35906069380401406,
+      "learning_rate": 0.00013154191966553488,
+      "loss": 1.0853,
+      "step": 8946
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2682913120576577,
+      "learning_rate": 0.0001315269073266666,
+      "loss": 1.011,
+      "step": 8947
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2686741445506597,
+      "learning_rate": 0.00013151189419881767,
+      "loss": 1.0058,
+      "step": 8948
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30034831308239196,
+      "learning_rate": 0.00013149688028236378,
+      "loss": 0.999,
+      "step": 8949
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2785651237874469,
+      "learning_rate": 0.00013148186557768065,
+      "loss": 1.0743,
+      "step": 8950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26665224113450825,
+      "learning_rate": 0.00013146685008514405,
+      "loss": 1.1498,
+      "step": 8951
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2767238879256942,
+      "learning_rate": 0.00013145183380512977,
+      "loss": 0.9933,
+      "step": 8952
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27676417086874466,
+      "learning_rate": 0.0001314368167380136,
+      "loss": 1.048,
+      "step": 8953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26331762630608485,
+      "learning_rate": 0.00013142179888417127,
+      "loss": 1.0725,
+      "step": 8954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2744742895646326,
+      "learning_rate": 0.00013140678024397876,
+      "loss": 1.0698,
+      "step": 8955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3284594696993571,
+      "learning_rate": 0.00013139176081781176,
+      "loss": 1.1246,
+      "step": 8956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.25822096562028224,
+      "learning_rate": 0.00013137674060604627,
+      "loss": 0.9168,
+      "step": 8957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26107725204700727,
+      "learning_rate": 0.0001313617196090581,
+      "loss": 1.2021,
+      "step": 8958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3179530192263277,
+      "learning_rate": 0.0001313466978272232,
+      "loss": 1.0834,
+      "step": 8959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2754800386668831,
+      "learning_rate": 0.00013133167526091746,
+      "loss": 1.0735,
+      "step": 8960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28969505946971996,
+      "learning_rate": 0.00013131665191051686,
+      "loss": 1.1824,
+      "step": 8961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3050269917318578,
+      "learning_rate": 0.0001313016277763974,
+      "loss": 1.1192,
+      "step": 8962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28997353889525096,
+      "learning_rate": 0.00013128660285893502,
+      "loss": 1.0988,
+      "step": 8963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2639957610030003,
+      "learning_rate": 0.00013127157715850572,
+      "loss": 1.095,
+      "step": 8964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.25243571206061105,
+      "learning_rate": 0.00013125655067548555,
+      "loss": 1.0326,
+      "step": 8965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27769176135258244,
+      "learning_rate": 0.00013124152341025057,
+      "loss": 1.0556,
+      "step": 8966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31119684825896593,
+      "learning_rate": 0.00013122649536317682,
+      "loss": 1.0474,
+      "step": 8967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2899032223624125,
+      "learning_rate": 0.0001312114665346404,
+      "loss": 1.1379,
+      "step": 8968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26474621772474893,
+      "learning_rate": 0.00013119643692501742,
+      "loss": 0.9737,
+      "step": 8969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.296070787125954,
+      "learning_rate": 0.000131181406534684,
+      "loss": 1.1303,
+      "step": 8970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28196083416898055,
+      "learning_rate": 0.00013116637536401626,
+      "loss": 0.9212,
+      "step": 8971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33777731349783074,
+      "learning_rate": 0.00013115134341339042,
+      "loss": 1.1114,
+      "step": 8972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2980471209168762,
+      "learning_rate": 0.00013113631068318262,
+      "loss": 1.0419,
+      "step": 8973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34067752191686806,
+      "learning_rate": 0.00013112127717376906,
+      "loss": 1.0489,
+      "step": 8974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2857921528011418,
+      "learning_rate": 0.000131106242885526,
+      "loss": 1.095,
+      "step": 8975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30595650571474364,
+      "learning_rate": 0.0001310912078188297,
+      "loss": 1.0579,
+      "step": 8976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3113974291276129,
+      "learning_rate": 0.00013107617197405632,
+      "loss": 1.1008,
+      "step": 8977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2835220329748002,
+      "learning_rate": 0.00013106113535158223,
+      "loss": 1.1758,
+      "step": 8978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30573600648132765,
+      "learning_rate": 0.00013104609795178373,
+      "loss": 1.0587,
+      "step": 8979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2724899362196218,
+      "learning_rate": 0.00013103105977503712,
+      "loss": 1.0202,
+      "step": 8980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28699936186440506,
+      "learning_rate": 0.0001310160208217187,
+      "loss": 1.1057,
+      "step": 8981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26272510332229243,
+      "learning_rate": 0.00013100098109220486,
+      "loss": 1.0776,
+      "step": 8982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27023197555661166,
+      "learning_rate": 0.00013098594058687203,
+      "loss": 1.1075,
+      "step": 8983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27849525197693725,
+      "learning_rate": 0.00013097089930609653,
+      "loss": 1.0556,
+      "step": 8984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2813628673155407,
+      "learning_rate": 0.00013095585725025481,
+      "loss": 0.9954,
+      "step": 8985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.23165959095448188,
+      "learning_rate": 0.00013094081441972333,
+      "loss": 1.0472,
+      "step": 8986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2865879802186648,
+      "learning_rate": 0.0001309257708148785,
+      "loss": 1.0757,
+      "step": 8987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28003705635045895,
+      "learning_rate": 0.00013091072643609683,
+      "loss": 1.027,
+      "step": 8988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2518027494418857,
+      "learning_rate": 0.0001308956812837548,
+      "loss": 1.074,
+      "step": 8989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2578937586135656,
+      "learning_rate": 0.0001308806353582289,
+      "loss": 1.0831,
+      "step": 8990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27260063912350424,
+      "learning_rate": 0.00013086558865989576,
+      "loss": 1.0183,
+      "step": 8991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29385408837160987,
+      "learning_rate": 0.0001308505411891318,
+      "loss": 1.0944,
+      "step": 8992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.293490265358121,
+      "learning_rate": 0.0001308354929463137,
+      "loss": 1.0724,
+      "step": 8993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2822276360093321,
+      "learning_rate": 0.00013082044393181798,
+      "loss": 1.0708,
+      "step": 8994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31600027722892315,
+      "learning_rate": 0.0001308053941460213,
+      "loss": 1.036,
+      "step": 8995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31974027288364265,
+      "learning_rate": 0.00013079034358930028,
+      "loss": 1.1244,
+      "step": 8996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2636215501400869,
+      "learning_rate": 0.00013077529226203155,
+      "loss": 0.9505,
+      "step": 8997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2732300022362931,
+      "learning_rate": 0.00013076024016459177,
+      "loss": 1.0561,
+      "step": 8998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33093910848931535,
+      "learning_rate": 0.0001307451872973577,
+      "loss": 1.0408,
+      "step": 8999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2880821403188908,
+      "learning_rate": 0.00013073013366070595,
+      "loss": 1.052,
+      "step": 9000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3061067194549959,
+      "learning_rate": 0.0001307150792550133,
+      "loss": 1.0279,
+      "step": 9001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2723856651427697,
+      "learning_rate": 0.0001307000240806565,
+      "loss": 1.0259,
+      "step": 9002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2997832873455384,
+      "learning_rate": 0.0001306849681380123,
+      "loss": 1.0171,
+      "step": 9003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.268324862068108,
+      "learning_rate": 0.00013066991142745746,
+      "loss": 1.0015,
+      "step": 9004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3313758049625676,
+      "learning_rate": 0.00013065485394936886,
+      "loss": 1.0488,
+      "step": 9005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2840867112984875,
+      "learning_rate": 0.00013063979570412324,
+      "loss": 1.0547,
+      "step": 9006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2859672508903575,
+      "learning_rate": 0.0001306247366920975,
+      "loss": 1.0588,
+      "step": 9007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29147422535829376,
+      "learning_rate": 0.00013060967691366844,
+      "loss": 1.126,
+      "step": 9008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2736559351409344,
+      "learning_rate": 0.00013059461636921298,
+      "loss": 0.9615,
+      "step": 9009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2772933874203004,
+      "learning_rate": 0.00013057955505910805,
+      "loss": 1.085,
+      "step": 9010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29810412332818575,
+      "learning_rate": 0.00013056449298373053,
+      "loss": 1.0038,
+      "step": 9011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2844185615999324,
+      "learning_rate": 0.00013054943014345732,
+      "loss": 1.1034,
+      "step": 9012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3198834730415027,
+      "learning_rate": 0.0001305343665386655,
+      "loss": 1.0527,
+      "step": 9013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29395624791864494,
+      "learning_rate": 0.00013051930216973192,
+      "loss": 1.1132,
+      "step": 9014
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28969238208023507,
+      "learning_rate": 0.0001305042370370336,
+      "loss": 0.8879,
+      "step": 9015
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2718247193165485,
+      "learning_rate": 0.0001304891711409476,
+      "loss": 0.9514,
+      "step": 9016
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2736779348162901,
+      "learning_rate": 0.00013047410448185096,
+      "loss": 1.0625,
+      "step": 9017
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3190676200490636,
+      "learning_rate": 0.00013045903706012066,
+      "loss": 1.1119,
+      "step": 9018
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2645211188163359,
+      "learning_rate": 0.00013044396887613383,
+      "loss": 1.0451,
+      "step": 9019
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27409629322747,
+      "learning_rate": 0.00013042889993026757,
+      "loss": 1.0542,
+      "step": 9020
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3057021252154314,
+      "learning_rate": 0.00013041383022289893,
+      "loss": 0.9845,
+      "step": 9021
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31255114430184533,
+      "learning_rate": 0.00013039875975440508,
+      "loss": 0.923,
+      "step": 9022
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.28633949394067004,
+      "learning_rate": 0.00013038368852516318,
+      "loss": 1.0501,
+      "step": 9023
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2635390208946298,
+      "learning_rate": 0.00013036861653555038,
+      "loss": 1.0946,
+      "step": 9024
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2938192444403693,
+      "learning_rate": 0.00013035354378594384,
+      "loss": 1.0054,
+      "step": 9025
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30536276310624666,
+      "learning_rate": 0.0001303384702767208,
+      "loss": 1.0497,
+      "step": 9026
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2790718788746024,
+      "learning_rate": 0.0001303233960082585,
+      "loss": 1.0417,
+      "step": 9027
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2816516990804407,
+      "learning_rate": 0.00013030832098093412,
+      "loss": 0.9402,
+      "step": 9028
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31401103333965236,
+      "learning_rate": 0.00013029324519512497,
+      "loss": 0.9892,
+      "step": 9029
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.26148363698940397,
+      "learning_rate": 0.00013027816865120834,
+      "loss": 1.0242,
+      "step": 9030
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2998596222624002,
+      "learning_rate": 0.0001302630913495615,
+      "loss": 1.0302,
+      "step": 9031
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29176595806240635,
+      "learning_rate": 0.00013024801329056178,
+      "loss": 1.2279,
+      "step": 9032
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3344098434780136,
+      "learning_rate": 0.00013023293447458648,
+      "loss": 1.0633,
+      "step": 9033
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.34296449020035136,
+      "learning_rate": 0.00013021785490201305,
+      "loss": 1.2539,
+      "step": 9034
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2564122283684667,
+      "learning_rate": 0.00013020277457321877,
+      "loss": 1.0903,
+      "step": 9035
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2917999377183397,
+      "learning_rate": 0.00013018769348858107,
+      "loss": 0.9977,
+      "step": 9036
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31170385260658784,
+      "learning_rate": 0.00013017261164847743,
+      "loss": 0.8905,
+      "step": 9037
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.30015787859342813,
+      "learning_rate": 0.00013015752905328514,
+      "loss": 1.0676,
+      "step": 9038
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.43157315991514567,
+      "learning_rate": 0.00013014244570338178,
+      "loss": 1.1419,
+      "step": 9039
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.25061895805404616,
+      "learning_rate": 0.0001301273615991448,
+      "loss": 0.9809,
+      "step": 9040
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2678433671294428,
+      "learning_rate": 0.00013011227674095162,
+      "loss": 0.951,
+      "step": 9041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2886930597012415,
+      "learning_rate": 0.00013009719112917978,
+      "loss": 1.0616,
+      "step": 9042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25785608296938134,
+      "learning_rate": 0.00013008210476420684,
+      "loss": 1.0351,
+      "step": 9043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25783062009926794,
+      "learning_rate": 0.0001300670176464103,
+      "loss": 1.0019,
+      "step": 9044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30369123418177635,
+      "learning_rate": 0.00013005192977616777,
+      "loss": 0.8641,
+      "step": 9045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3225428130185902,
+      "learning_rate": 0.0001300368411538568,
+      "loss": 1.1576,
+      "step": 9046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2967976819197218,
+      "learning_rate": 0.00013002175177985502,
+      "loss": 1.1703,
+      "step": 9047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3014813200032349,
+      "learning_rate": 0.00013000666165454,
+      "loss": 1.1789,
+      "step": 9048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2501779735414716,
+      "learning_rate": 0.00012999157077828944,
+      "loss": 1.105,
+      "step": 9049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32187566601028167,
+      "learning_rate": 0.000129976479151481,
+      "loss": 1.0835,
+      "step": 9050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27746438481780444,
+      "learning_rate": 0.0001299613867744923,
+      "loss": 1.0821,
+      "step": 9051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28339489212125185,
+      "learning_rate": 0.00012994629364770102,
+      "loss": 1.0247,
+      "step": 9052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27520392235897523,
+      "learning_rate": 0.00012993119977148499,
+      "loss": 0.9562,
+      "step": 9053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2575142187774334,
+      "learning_rate": 0.0001299161051462218,
+      "loss": 1.0923,
+      "step": 9054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2934030821339123,
+      "learning_rate": 0.00012990100977228934,
+      "loss": 1.0186,
+      "step": 9055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2825842909756287,
+      "learning_rate": 0.0001298859136500653,
+      "loss": 1.0244,
+      "step": 9056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2732801101975911,
+      "learning_rate": 0.0001298708167799275,
+      "loss": 1.1346,
+      "step": 9057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2835565308919408,
+      "learning_rate": 0.0001298557191622537,
+      "loss": 0.9529,
+      "step": 9058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32615560594023263,
+      "learning_rate": 0.00012984062079742181,
+      "loss": 1.1499,
+      "step": 9059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2869135232591449,
+      "learning_rate": 0.00012982552168580962,
+      "loss": 0.9194,
+      "step": 9060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3134787513394852,
+      "learning_rate": 0.000129810421827795,
+      "loss": 1.1952,
+      "step": 9061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32346472417562094,
+      "learning_rate": 0.0001297953212237558,
+      "loss": 1.023,
+      "step": 9062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3367904265131591,
+      "learning_rate": 0.00012978021987407004,
+      "loss": 1.1234,
+      "step": 9063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30432907150859745,
+      "learning_rate": 0.0001297651177791155,
+      "loss": 0.9762,
+      "step": 9064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3161449071160964,
+      "learning_rate": 0.00012975001493927018,
+      "loss": 0.9887,
+      "step": 9065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24592648677051562,
+      "learning_rate": 0.00012973491135491206,
+      "loss": 1.0069,
+      "step": 9066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26653722611873165,
+      "learning_rate": 0.00012971980702641912,
+      "loss": 1.0546,
+      "step": 9067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2862693690630182,
+      "learning_rate": 0.00012970470195416931,
+      "loss": 1.1177,
+      "step": 9068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29138542499171116,
+      "learning_rate": 0.00012968959613854063,
+      "loss": 1.1238,
+      "step": 9069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2859847406201628,
+      "learning_rate": 0.0001296744895799112,
+      "loss": 1.047,
+      "step": 9070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2807399557511384,
+      "learning_rate": 0.000129659382278659,
+      "loss": 0.9589,
+      "step": 9071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28450650670607563,
+      "learning_rate": 0.0001296442742351621,
+      "loss": 0.9999,
+      "step": 9072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31676772977788115,
+      "learning_rate": 0.0001296291654497986,
+      "loss": 0.978,
+      "step": 9073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2784851835452743,
+      "learning_rate": 0.00012961405592294665,
+      "loss": 1.0332,
+      "step": 9074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26832152402985765,
+      "learning_rate": 0.0001295989456549843,
+      "loss": 1.0811,
+      "step": 9075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29162865744847255,
+      "learning_rate": 0.00012958383464628975,
+      "loss": 1.1361,
+      "step": 9076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26672809187008667,
+      "learning_rate": 0.00012956872289724116,
+      "loss": 1.0796,
+      "step": 9077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27076549176531384,
+      "learning_rate": 0.0001295536104082167,
+      "loss": 1.1287,
+      "step": 9078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28696821632322866,
+      "learning_rate": 0.00012953849717959454,
+      "loss": 1.1795,
+      "step": 9079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2955303786908336,
+      "learning_rate": 0.00012952338321175293,
+      "loss": 1.0168,
+      "step": 9080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3039172006339761,
+      "learning_rate": 0.00012950826850507011,
+      "loss": 0.9734,
+      "step": 9081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2680076109464228,
+      "learning_rate": 0.00012949315305992433,
+      "loss": 1.1325,
+      "step": 9082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3197355029722075,
+      "learning_rate": 0.00012947803687669385,
+      "loss": 1.1474,
+      "step": 9083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.260643212258335,
+      "learning_rate": 0.00012946291995575697,
+      "loss": 1.0691,
+      "step": 9084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28094689856026667,
+      "learning_rate": 0.00012944780229749201,
+      "loss": 1.1027,
+      "step": 9085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3354262462770161,
+      "learning_rate": 0.00012943268390227727,
+      "loss": 0.9943,
+      "step": 9086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2863437127555651,
+      "learning_rate": 0.00012941756477049114,
+      "loss": 1.101,
+      "step": 9087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.24300912831153612,
+      "learning_rate": 0.00012940244490251197,
+      "loss": 0.9077,
+      "step": 9088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.259039073094172,
+      "learning_rate": 0.0001293873242987181,
+      "loss": 1.117,
+      "step": 9089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2512219349760124,
+      "learning_rate": 0.000129372202959488,
+      "loss": 0.9785,
+      "step": 9090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27273837493945713,
+      "learning_rate": 0.00012935708088520007,
+      "loss": 0.9931,
+      "step": 9091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27530538774617,
+      "learning_rate": 0.0001293419580762327,
+      "loss": 1.0153,
+      "step": 9092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2926747917554771,
+      "learning_rate": 0.0001293268345329644,
+      "loss": 1.0461,
+      "step": 9093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2500212462256167,
+      "learning_rate": 0.00012931171025577366,
+      "loss": 1.0736,
+      "step": 9094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29111956874370015,
+      "learning_rate": 0.00012929658524503894,
+      "loss": 1.0124,
+      "step": 9095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.263782277974996,
+      "learning_rate": 0.00012928145950113877,
+      "loss": 0.9819,
+      "step": 9096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2905200592972542,
+      "learning_rate": 0.00012926633302445164,
+      "loss": 1.0002,
+      "step": 9097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29496184216102744,
+      "learning_rate": 0.00012925120581535614,
+      "loss": 1.0787,
+      "step": 9098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3117510201468771,
+      "learning_rate": 0.00012923607787423085,
+      "loss": 1.0186,
+      "step": 9099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32624449591492205,
+      "learning_rate": 0.00012922094920145432,
+      "loss": 1.1453,
+      "step": 9100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.33313379774620716,
+      "learning_rate": 0.0001292058197974052,
+      "loss": 1.1103,
+      "step": 9101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30476232541112913,
+      "learning_rate": 0.0001291906896624621,
+      "loss": 1.0197,
+      "step": 9102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2966527557371042,
+      "learning_rate": 0.00012917555879700358,
+      "loss": 1.0609,
+      "step": 9103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3050234189493314,
+      "learning_rate": 0.0001291604272014084,
+      "loss": 1.098,
+      "step": 9104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27720296715762255,
+      "learning_rate": 0.0001291452948760552,
+      "loss": 1.0512,
+      "step": 9105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3042006172159329,
+      "learning_rate": 0.00012913016182132268,
+      "loss": 0.9907,
+      "step": 9106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3068804882948529,
+      "learning_rate": 0.00012911502803758954,
+      "loss": 1.1273,
+      "step": 9107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25437915672597605,
+      "learning_rate": 0.00012909989352523455,
+      "loss": 1.0053,
+      "step": 9108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2588485696040596,
+      "learning_rate": 0.00012908475828463643,
+      "loss": 1.069,
+      "step": 9109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.288489014040999,
+      "learning_rate": 0.00012906962231617396,
+      "loss": 1.1162,
+      "step": 9110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2902499883183751,
+      "learning_rate": 0.00012905448562022592,
+      "loss": 0.959,
+      "step": 9111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2885389820207888,
+      "learning_rate": 0.00012903934819717108,
+      "loss": 1.0781,
+      "step": 9112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2434151493288888,
+      "learning_rate": 0.00012902421004738833,
+      "loss": 1.1026,
+      "step": 9113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27724613447577373,
+      "learning_rate": 0.0001290090711712565,
+      "loss": 0.9479,
+      "step": 9114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2638723940275514,
+      "learning_rate": 0.00012899393156915438,
+      "loss": 1.0193,
+      "step": 9115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3146091758887233,
+      "learning_rate": 0.00012897879124146094,
+      "loss": 1.0415,
+      "step": 9116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2781231516386979,
+      "learning_rate": 0.00012896365018855502,
+      "loss": 1.1148,
+      "step": 9117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2814021450123451,
+      "learning_rate": 0.00012894850841081555,
+      "loss": 0.9947,
+      "step": 9118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2951572978095963,
+      "learning_rate": 0.0001289333659086215,
+      "loss": 1.0666,
+      "step": 9119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2789547266879171,
+      "learning_rate": 0.00012891822268235175,
+      "loss": 1.1119,
+      "step": 9120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2944178727397588,
+      "learning_rate": 0.0001289030787323853,
+      "loss": 1.0622,
+      "step": 9121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3002264804462691,
+      "learning_rate": 0.00012888793405910117,
+      "loss": 1.1301,
+      "step": 9122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29596892828518445,
+      "learning_rate": 0.0001288727886628783,
+      "loss": 0.9955,
+      "step": 9123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28581897006530516,
+      "learning_rate": 0.00012885764254409577,
+      "loss": 1.0991,
+      "step": 9124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2848802047710548,
+      "learning_rate": 0.0001288424957031326,
+      "loss": 1.1973,
+      "step": 9125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.30273121467305963,
+      "learning_rate": 0.00012882734814036783,
+      "loss": 1.1433,
+      "step": 9126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.26901216004525036,
+      "learning_rate": 0.00012881219985618058,
+      "loss": 0.9651,
+      "step": 9127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29357748445834536,
+      "learning_rate": 0.0001287970508509499,
+      "loss": 1.0533,
+      "step": 9128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28302245682622934,
+      "learning_rate": 0.00012878190112505496,
+      "loss": 1.1715,
+      "step": 9129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.32220788805164724,
+      "learning_rate": 0.0001287667506788748,
+      "loss": 1.037,
+      "step": 9130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2838656933293558,
+      "learning_rate": 0.00012875159951278867,
+      "loss": 0.9909,
+      "step": 9131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31812680763347256,
+      "learning_rate": 0.0001287364476271757,
+      "loss": 1.0599,
+      "step": 9132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2729154895834484,
+      "learning_rate": 0.00012872129502241502,
+      "loss": 1.0977,
+      "step": 9133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3125996237172212,
+      "learning_rate": 0.0001287061416988859,
+      "loss": 1.0514,
+      "step": 9134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2807631918763277,
+      "learning_rate": 0.00012869098765696757,
+      "loss": 1.051,
+      "step": 9135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31219828383217524,
+      "learning_rate": 0.0001286758328970392,
+      "loss": 1.0464,
+      "step": 9136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2639358482520139,
+      "learning_rate": 0.0001286606774194801,
+      "loss": 1.0578,
+      "step": 9137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2974661876294612,
+      "learning_rate": 0.00012864552122466956,
+      "loss": 0.9595,
+      "step": 9138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2926598180262982,
+      "learning_rate": 0.00012863036431298684,
+      "loss": 1.1708,
+      "step": 9139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.25929024999451794,
+      "learning_rate": 0.00012861520668481122,
+      "loss": 0.9799,
+      "step": 9140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29654960563184474,
+      "learning_rate": 0.0001286000483405221,
+      "loss": 1.1523,
+      "step": 9141
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27691848301002864,
+      "learning_rate": 0.00012858488928049882,
+      "loss": 1.0334,
+      "step": 9142
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29295271269701484,
+      "learning_rate": 0.00012856972950512068,
+      "loss": 1.0546,
+      "step": 9143
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2803146245701744,
+      "learning_rate": 0.00012855456901476712,
+      "loss": 0.9898,
+      "step": 9144
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2863989751594063,
+      "learning_rate": 0.00012853940780981751,
+      "loss": 1.1504,
+      "step": 9145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2739972569019825,
+      "learning_rate": 0.00012852424589065132,
+      "loss": 0.923,
+      "step": 9146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25052615701520137,
+      "learning_rate": 0.0001285090832576479,
+      "loss": 1.0061,
+      "step": 9147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3030142194521633,
+      "learning_rate": 0.00012849391991118683,
+      "loss": 1.0694,
+      "step": 9148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2714696456433213,
+      "learning_rate": 0.00012847875585164745,
+      "loss": 1.0009,
+      "step": 9149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28875446135458116,
+      "learning_rate": 0.00012846359107940931,
+      "loss": 1.1649,
+      "step": 9150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3150764287272924,
+      "learning_rate": 0.00012844842559485192,
+      "loss": 0.9091,
+      "step": 9151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27701993372597217,
+      "learning_rate": 0.00012843325939835483,
+      "loss": 1.083,
+      "step": 9152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29338798312190123,
+      "learning_rate": 0.00012841809249029747,
+      "loss": 0.9318,
+      "step": 9153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30076117068870556,
+      "learning_rate": 0.00012840292487105955,
+      "loss": 1.0577,
+      "step": 9154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26916632631906984,
+      "learning_rate": 0.0001283877565410206,
+      "loss": 1.0901,
+      "step": 9155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2863676566010998,
+      "learning_rate": 0.00012837258750056016,
+      "loss": 1.0916,
+      "step": 9156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.311513288826919,
+      "learning_rate": 0.0001283574177500579,
+      "loss": 0.9895,
+      "step": 9157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2788747044437777,
+      "learning_rate": 0.00012834224728989344,
+      "loss": 1.0058,
+      "step": 9158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.31246123148504296,
+      "learning_rate": 0.00012832707612044642,
+      "loss": 1.1208,
+      "step": 9159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3153487374039508,
+      "learning_rate": 0.00012831190424209655,
+      "loss": 1.0619,
+      "step": 9160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2979832645966896,
+      "learning_rate": 0.00012829673165522343,
+      "loss": 1.0788,
+      "step": 9161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27761648021394475,
+      "learning_rate": 0.00012828155836020687,
+      "loss": 1.0676,
+      "step": 9162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2966190620374098,
+      "learning_rate": 0.00012826638435742654,
+      "loss": 1.127,
+      "step": 9163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2677449239886606,
+      "learning_rate": 0.0001282512096472621,
+      "loss": 1.1182,
+      "step": 9164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2955895529683899,
+      "learning_rate": 0.00012823603423009347,
+      "loss": 0.8931,
+      "step": 9165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2864122311531748,
+      "learning_rate": 0.0001282208581063003,
+      "loss": 1.1078,
+      "step": 9166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2901266851305628,
+      "learning_rate": 0.00012820568127626242,
+      "loss": 1.0239,
+      "step": 9167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3049079996092775,
+      "learning_rate": 0.00012819050374035962,
+      "loss": 1.0681,
+      "step": 9168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27371968074107056,
+      "learning_rate": 0.0001281753254989718,
+      "loss": 1.0393,
+      "step": 9169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.31873095311089905,
+      "learning_rate": 0.0001281601465524787,
+      "loss": 1.0495,
+      "step": 9170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33430466740169595,
+      "learning_rate": 0.00012814496690126027,
+      "loss": 1.0828,
+      "step": 9171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2793091890823801,
+      "learning_rate": 0.00012812978654569635,
+      "loss": 1.0,
+      "step": 9172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28127672160759526,
+      "learning_rate": 0.00012811460548616682,
+      "loss": 1.1068,
+      "step": 9173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2822256164444625,
+      "learning_rate": 0.00012809942372305164,
+      "loss": 1.1039,
+      "step": 9174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2989368506283751,
+      "learning_rate": 0.0001280842412567307,
+      "loss": 1.0661,
+      "step": 9175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2787834185347091,
+      "learning_rate": 0.000128069058087584,
+      "loss": 1.0754,
+      "step": 9176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26459335862211264,
+      "learning_rate": 0.00012805387421599144,
+      "loss": 1.0921,
+      "step": 9177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2717879411221796,
+      "learning_rate": 0.0001280386896423331,
+      "loss": 1.0256,
+      "step": 9178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3392781499730835,
+      "learning_rate": 0.00012802350436698888,
+      "loss": 0.963,
+      "step": 9179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3008890143060967,
+      "learning_rate": 0.0001280083183903389,
+      "loss": 1.0453,
+      "step": 9180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2859398503777074,
+      "learning_rate": 0.00012799313171276308,
+      "loss": 1.095,
+      "step": 9181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.269640842666525,
+      "learning_rate": 0.0001279779443346416,
+      "loss": 1.0724,
+      "step": 9182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28038845119609257,
+      "learning_rate": 0.0001279627562563545,
+      "loss": 1.0738,
+      "step": 9183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2887167863295914,
+      "learning_rate": 0.00012794756747828179,
+      "loss": 1.0641,
+      "step": 9184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29441549775626563,
+      "learning_rate": 0.00012793237800080365,
+      "loss": 1.1956,
+      "step": 9185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2614449210833528,
+      "learning_rate": 0.00012791718782430024,
+      "loss": 0.9941,
+      "step": 9186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27898220786899924,
+      "learning_rate": 0.00012790199694915163,
+      "loss": 1.0125,
+      "step": 9187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2623293441843174,
+      "learning_rate": 0.000127886805375738,
+      "loss": 1.018,
+      "step": 9188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24474445860858335,
+      "learning_rate": 0.00012787161310443958,
+      "loss": 0.9925,
+      "step": 9189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26691886103805196,
+      "learning_rate": 0.0001278564201356365,
+      "loss": 1.0262,
+      "step": 9190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2996480479285695,
+      "learning_rate": 0.000127841226469709,
+      "loss": 1.1404,
+      "step": 9191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2831005610084299,
+      "learning_rate": 0.0001278260321070373,
+      "loss": 1.0199,
+      "step": 9192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3016403069097424,
+      "learning_rate": 0.00012781083704800167,
+      "loss": 1.1395,
+      "step": 9193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.258599684442815,
+      "learning_rate": 0.00012779564129298233,
+      "loss": 1.0072,
+      "step": 9194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28981820294114957,
+      "learning_rate": 0.00012778044484235964,
+      "loss": 1.0872,
+      "step": 9195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2944671368276984,
+      "learning_rate": 0.0001277652476965139,
+      "loss": 0.9736,
+      "step": 9196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.32250052614896707,
+      "learning_rate": 0.0001277500498558253,
+      "loss": 1.0861,
+      "step": 9197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.33413812354501804,
+      "learning_rate": 0.00012773485132067428,
+      "loss": 0.9882,
+      "step": 9198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2396696634982347,
+      "learning_rate": 0.00012771965209144122,
+      "loss": 0.999,
+      "step": 9199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2871461793248707,
+      "learning_rate": 0.00012770445216850638,
+      "loss": 0.9985,
+      "step": 9200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27681646471608834,
+      "learning_rate": 0.00012768925155225025,
+      "loss": 0.9664,
+      "step": 9201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27957278139129027,
+      "learning_rate": 0.00012767405024305322,
+      "loss": 1.0305,
+      "step": 9202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28753086690595353,
+      "learning_rate": 0.00012765884824129565,
+      "loss": 1.0237,
+      "step": 9203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3111440689789316,
+      "learning_rate": 0.000127643645547358,
+      "loss": 1.2017,
+      "step": 9204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2921443717707693,
+      "learning_rate": 0.0001276284421616208,
+      "loss": 1.029,
+      "step": 9205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30261904265294104,
+      "learning_rate": 0.00012761323808446447,
+      "loss": 1.0758,
+      "step": 9206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25772792131049205,
+      "learning_rate": 0.00012759803331626948,
+      "loss": 1.0139,
+      "step": 9207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3046831790519165,
+      "learning_rate": 0.00012758282785741638,
+      "loss": 1.0825,
+      "step": 9208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3183296731093666,
+      "learning_rate": 0.00012756762170828566,
+      "loss": 0.9899,
+      "step": 9209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2817939566711674,
+      "learning_rate": 0.0001275524148692579,
+      "loss": 0.9752,
+      "step": 9210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2665373598422484,
+      "learning_rate": 0.0001275372073407136,
+      "loss": 1.0575,
+      "step": 9211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2725700632449814,
+      "learning_rate": 0.00012752199912303345,
+      "loss": 1.0651,
+      "step": 9212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29400308946907394,
+      "learning_rate": 0.00012750679021659794,
+      "loss": 1.0696,
+      "step": 9213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.297641959611062,
+      "learning_rate": 0.00012749158062178769,
+      "loss": 1.0051,
+      "step": 9214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24521868429916097,
+      "learning_rate": 0.0001274763703389834,
+      "loss": 0.9303,
+      "step": 9215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29297299579392244,
+      "learning_rate": 0.00012746115936856564,
+      "loss": 1.1431,
+      "step": 9216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26851527665671227,
+      "learning_rate": 0.00012744594771091513,
+      "loss": 1.0198,
+      "step": 9217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2819204613169668,
+      "learning_rate": 0.0001274307353664125,
+      "loss": 1.1456,
+      "step": 9218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3038986733258005,
+      "learning_rate": 0.00012741552233543852,
+      "loss": 0.9887,
+      "step": 9219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3047117203847732,
+      "learning_rate": 0.0001274003086183738,
+      "loss": 1.069,
+      "step": 9220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2920407974764749,
+      "learning_rate": 0.0001273850942155992,
+      "loss": 1.113,
+      "step": 9221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28997161723843756,
+      "learning_rate": 0.0001273698791274954,
+      "loss": 0.9928,
+      "step": 9222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2601062102739676,
+      "learning_rate": 0.00012735466335444314,
+      "loss": 1.0277,
+      "step": 9223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24196462571085023,
+      "learning_rate": 0.00012733944689682325,
+      "loss": 1.0118,
+      "step": 9224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27692067285133026,
+      "learning_rate": 0.00012732422975501653,
+      "loss": 1.1095,
+      "step": 9225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25760809183221073,
+      "learning_rate": 0.0001273090119294038,
+      "loss": 0.9548,
+      "step": 9226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.25859993613149374,
+      "learning_rate": 0.00012729379342036587,
+      "loss": 0.9637,
+      "step": 9227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2540462826516634,
+      "learning_rate": 0.00012727857422828359,
+      "loss": 1.1307,
+      "step": 9228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2921227707514754,
+      "learning_rate": 0.00012726335435353785,
+      "loss": 1.1008,
+      "step": 9229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2984915373574552,
+      "learning_rate": 0.00012724813379650954,
+      "loss": 1.0848,
+      "step": 9230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3095681979770164,
+      "learning_rate": 0.00012723291255757957,
+      "loss": 1.1033,
+      "step": 9231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28318769682527445,
+      "learning_rate": 0.00012721769063712884,
+      "loss": 1.0971,
+      "step": 9232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27525707454107234,
+      "learning_rate": 0.00012720246803553828,
+      "loss": 1.0534,
+      "step": 9233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28825651516813144,
+      "learning_rate": 0.0001271872447531889,
+      "loss": 1.1621,
+      "step": 9234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3010094105143729,
+      "learning_rate": 0.0001271720207904616,
+      "loss": 1.072,
+      "step": 9235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.24956098318442146,
+      "learning_rate": 0.00012715679614773738,
+      "loss": 1.2222,
+      "step": 9236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2764944338239069,
+      "learning_rate": 0.00012714157082539733,
+      "loss": 1.0995,
+      "step": 9237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.27634606116433363,
+      "learning_rate": 0.00012712634482382238,
+      "loss": 1.036,
+      "step": 9238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2794870761616325,
+      "learning_rate": 0.0001271111181433936,
+      "loss": 1.1254,
+      "step": 9239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2601345404562058,
+      "learning_rate": 0.00012709589078449204,
+      "loss": 1.1053,
+      "step": 9240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2622935878916725,
+      "learning_rate": 0.0001270806627474988,
+      "loss": 1.0455,
+      "step": 9241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29883490193917805,
+      "learning_rate": 0.00012706543403279497,
+      "loss": 0.9762,
+      "step": 9242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.22869752618561462,
+      "learning_rate": 0.0001270502046407616,
+      "loss": 0.9916,
+      "step": 9243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29444348606081083,
+      "learning_rate": 0.00012703497457177988,
+      "loss": 0.9847,
+      "step": 9244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2938071041973925,
+      "learning_rate": 0.00012701974382623094,
+      "loss": 1.0545,
+      "step": 9245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26894685253793144,
+      "learning_rate": 0.00012700451240449593,
+      "loss": 1.0102,
+      "step": 9246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3295306409371883,
+      "learning_rate": 0.00012698928030695602,
+      "loss": 1.1135,
+      "step": 9247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.29519953492147355,
+      "learning_rate": 0.0001269740475339924,
+      "loss": 1.1716,
+      "step": 9248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.31644779195394845,
+      "learning_rate": 0.0001269588140859863,
+      "loss": 1.0338,
+      "step": 9249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28862661343334506,
+      "learning_rate": 0.00012694357996331893,
+      "loss": 0.9805,
+      "step": 9250
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29109092097690287,
+      "learning_rate": 0.00012692834516637156,
+      "loss": 1.0948,
+      "step": 9251
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2755208136905377,
+      "learning_rate": 0.00012691310969552538,
+      "loss": 1.0859,
+      "step": 9252
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2947036722853116,
+      "learning_rate": 0.00012689787355116177,
+      "loss": 1.0358,
+      "step": 9253
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2871135631256799,
+      "learning_rate": 0.00012688263673366195,
+      "loss": 1.0249,
+      "step": 9254
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2562542263019815,
+      "learning_rate": 0.0001268673992434072,
+      "loss": 1.0182,
+      "step": 9255
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31659182199731395,
+      "learning_rate": 0.00012685216108077895,
+      "loss": 1.0291,
+      "step": 9256
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2933670124167318,
+      "learning_rate": 0.0001268369222461585,
+      "loss": 1.0631,
+      "step": 9257
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2657225784590937,
+      "learning_rate": 0.0001268216827399272,
+      "loss": 1.119,
+      "step": 9258
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2797518168945411,
+      "learning_rate": 0.00012680644256246642,
+      "loss": 0.9506,
+      "step": 9259
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3020240632022724,
+      "learning_rate": 0.00012679120171415757,
+      "loss": 0.9763,
+      "step": 9260
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2637706589429481,
+      "learning_rate": 0.00012677596019538206,
+      "loss": 1.0194,
+      "step": 9261
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31418144937011505,
+      "learning_rate": 0.0001267607180065213,
+      "loss": 1.1327,
+      "step": 9262
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2952874649931668,
+      "learning_rate": 0.00012674547514795675,
+      "loss": 1.0817,
+      "step": 9263
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3150374662852804,
+      "learning_rate": 0.00012673023162006989,
+      "loss": 1.0925,
+      "step": 9264
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3198349522638215,
+      "learning_rate": 0.0001267149874232422,
+      "loss": 1.1409,
+      "step": 9265
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25854203157035266,
+      "learning_rate": 0.00012669974255785516,
+      "loss": 1.138,
+      "step": 9266
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2608990015480636,
+      "learning_rate": 0.00012668449702429028,
+      "loss": 1.0271,
+      "step": 9267
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29909886728393076,
+      "learning_rate": 0.0001266692508229291,
+      "loss": 1.0969,
+      "step": 9268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.23909051250886207,
+      "learning_rate": 0.0001266540039541531,
+      "loss": 0.9744,
+      "step": 9269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28087485327219586,
+      "learning_rate": 0.00012663875641834394,
+      "loss": 1.1004,
+      "step": 9270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2552216763566098,
+      "learning_rate": 0.0001266235082158832,
+      "loss": 0.9802,
+      "step": 9271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28763301871999136,
+      "learning_rate": 0.00012660825934715235,
+      "loss": 1.0887,
+      "step": 9272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29630933468752285,
+      "learning_rate": 0.00012659300981253315,
+      "loss": 1.1228,
+      "step": 9273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29194868948582664,
+      "learning_rate": 0.00012657775961240713,
+      "loss": 1.0174,
+      "step": 9274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2617941239283223,
+      "learning_rate": 0.000126562508747156,
+      "loss": 1.0721,
+      "step": 9275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28279751520459484,
+      "learning_rate": 0.00012654725721716138,
+      "loss": 1.0261,
+      "step": 9276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30103442627581245,
+      "learning_rate": 0.00012653200502280498,
+      "loss": 1.1181,
+      "step": 9277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27808862986365707,
+      "learning_rate": 0.00012651675216446848,
+      "loss": 1.0745,
+      "step": 9278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29656831264442357,
+      "learning_rate": 0.00012650149864253357,
+      "loss": 1.1474,
+      "step": 9279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27503893982251604,
+      "learning_rate": 0.000126486244457382,
+      "loss": 1.0173,
+      "step": 9280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2657056986153241,
+      "learning_rate": 0.00012647098960939554,
+      "loss": 0.9806,
+      "step": 9281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25653477829913407,
+      "learning_rate": 0.0001264557340989559,
+      "loss": 1.0358,
+      "step": 9282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29619037014025956,
+      "learning_rate": 0.0001264404779264449,
+      "loss": 1.1536,
+      "step": 9283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2822157801841811,
+      "learning_rate": 0.00012642522109224434,
+      "loss": 1.1011,
+      "step": 9284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31455283414554064,
+      "learning_rate": 0.000126409963596736,
+      "loss": 0.9813,
+      "step": 9285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2775822826731524,
+      "learning_rate": 0.0001263947054403017,
+      "loss": 0.951,
+      "step": 9286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2575236283104343,
+      "learning_rate": 0.00012637944662332332,
+      "loss": 0.9371,
+      "step": 9287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24149231079738087,
+      "learning_rate": 0.00012636418714618273,
+      "loss": 0.9715,
+      "step": 9288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2831753975004831,
+      "learning_rate": 0.00012634892700926178,
+      "loss": 1.0039,
+      "step": 9289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29422901876914576,
+      "learning_rate": 0.00012633366621294238,
+      "loss": 1.1093,
+      "step": 9290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27392177332876777,
+      "learning_rate": 0.00012631840475760644,
+      "loss": 1.1547,
+      "step": 9291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27799764214008904,
+      "learning_rate": 0.00012630314264363584,
+      "loss": 1.1274,
+      "step": 9292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31072988212098307,
+      "learning_rate": 0.0001262878798714126,
+      "loss": 1.0469,
+      "step": 9293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31492709765160437,
+      "learning_rate": 0.00012627261644131862,
+      "loss": 0.9578,
+      "step": 9294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2528035347256275,
+      "learning_rate": 0.00012625735235373593,
+      "loss": 1.0668,
+      "step": 9295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3155214475191695,
+      "learning_rate": 0.00012624208760904647,
+      "loss": 0.9916,
+      "step": 9296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.302087235196872,
+      "learning_rate": 0.00012622682220763228,
+      "loss": 1.0884,
+      "step": 9297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2990095102194339,
+      "learning_rate": 0.00012621155614987538,
+      "loss": 1.0101,
+      "step": 9298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26706525825653776,
+      "learning_rate": 0.00012619628943615782,
+      "loss": 1.0278,
+      "step": 9299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27756280255006816,
+      "learning_rate": 0.00012618102206686166,
+      "loss": 1.0099,
+      "step": 9300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30586196916112135,
+      "learning_rate": 0.00012616575404236899,
+      "loss": 1.0915,
+      "step": 9301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.25272225815414157,
+      "learning_rate": 0.0001261504853630618,
+      "loss": 1.0799,
+      "step": 9302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3258028723785073,
+      "learning_rate": 0.00012613521602932237,
+      "loss": 0.9907,
+      "step": 9303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3105133570818196,
+      "learning_rate": 0.00012611994604153269,
+      "loss": 1.0494,
+      "step": 9304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2853288951774398,
+      "learning_rate": 0.00012610467540007494,
+      "loss": 1.0367,
+      "step": 9305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30136724902716244,
+      "learning_rate": 0.00012608940410533127,
+      "loss": 1.0905,
+      "step": 9306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27766552230271924,
+      "learning_rate": 0.00012607413215768388,
+      "loss": 1.0067,
+      "step": 9307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2839527667018618,
+      "learning_rate": 0.00012605885955751497,
+      "loss": 0.9767,
+      "step": 9308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30687156353062944,
+      "learning_rate": 0.0001260435863052067,
+      "loss": 1.0725,
+      "step": 9309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3156040190026411,
+      "learning_rate": 0.0001260283124011413,
+      "loss": 1.0457,
+      "step": 9310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3104920835342392,
+      "learning_rate": 0.00012601303784570106,
+      "loss": 1.1438,
+      "step": 9311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27017582473335783,
+      "learning_rate": 0.0001259977626392682,
+      "loss": 1.052,
+      "step": 9312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27144499750698986,
+      "learning_rate": 0.00012598248678222498,
+      "loss": 1.0559,
+      "step": 9313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2675880100657242,
+      "learning_rate": 0.0001259672102749537,
+      "loss": 0.9763,
+      "step": 9314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2955864081602817,
+      "learning_rate": 0.00012595193311783665,
+      "loss": 0.9867,
+      "step": 9315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2813400142046753,
+      "learning_rate": 0.00012593665531125615,
+      "loss": 1.0049,
+      "step": 9316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2831320746009813,
+      "learning_rate": 0.00012592137685559458,
+      "loss": 1.029,
+      "step": 9317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29831547442685147,
+      "learning_rate": 0.00012590609775123426,
+      "loss": 1.116,
+      "step": 9318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2936254006184292,
+      "learning_rate": 0.00012589081799855756,
+      "loss": 1.0433,
+      "step": 9319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30202822877591445,
+      "learning_rate": 0.00012587553759794683,
+      "loss": 1.067,
+      "step": 9320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31153441957615685,
+      "learning_rate": 0.00012586025654978458,
+      "loss": 1.0185,
+      "step": 9321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.31241827225447677,
+      "learning_rate": 0.0001258449748544531,
+      "loss": 1.0975,
+      "step": 9322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27730753920433476,
+      "learning_rate": 0.0001258296925123349,
+      "loss": 1.069,
+      "step": 9323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2897016648466612,
+      "learning_rate": 0.00012581440952381243,
+      "loss": 1.067,
+      "step": 9324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.24652987673338808,
+      "learning_rate": 0.0001257991258892681,
+      "loss": 1.1338,
+      "step": 9325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3071238167163373,
+      "learning_rate": 0.00012578384160908445,
+      "loss": 1.1402,
+      "step": 9326
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29112654720110537,
+      "learning_rate": 0.00012576855668364396,
+      "loss": 1.0772,
+      "step": 9327
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28108564003717257,
+      "learning_rate": 0.00012575327111332912,
+      "loss": 1.0895,
+      "step": 9328
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2697345251089792,
+      "learning_rate": 0.00012573798489852253,
+      "loss": 1.1332,
+      "step": 9329
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.304121790273392,
+      "learning_rate": 0.00012572269803960665,
+      "loss": 1.0902,
+      "step": 9330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28853126394768475,
+      "learning_rate": 0.00012570741053696412,
+      "loss": 1.0721,
+      "step": 9331
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2851724014266135,
+      "learning_rate": 0.0001256921223909775,
+      "loss": 1.1567,
+      "step": 9332
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28817110546138114,
+      "learning_rate": 0.0001256768336020293,
+      "loss": 1.1432,
+      "step": 9333
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2674528738842632,
+      "learning_rate": 0.00012566154417050225,
+      "loss": 1.1189,
+      "step": 9334
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.27322750722188155,
+      "learning_rate": 0.00012564625409677895,
+      "loss": 1.0305,
+      "step": 9335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29977545197889227,
+      "learning_rate": 0.000125630963381242,
+      "loss": 1.1117,
+      "step": 9336
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29111294479338623,
+      "learning_rate": 0.00012561567202427407,
+      "loss": 0.9916,
+      "step": 9337
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3036268811633583,
+      "learning_rate": 0.00012560038002625788,
+      "loss": 1.0984,
+      "step": 9338
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2549172822007168,
+      "learning_rate": 0.00012558508738757604,
+      "loss": 1.1815,
+      "step": 9339
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2675936118833171,
+      "learning_rate": 0.00012556979410861135,
+      "loss": 0.962,
+      "step": 9340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2773919289873,
+      "learning_rate": 0.00012555450018974647,
+      "loss": 1.1229,
+      "step": 9341
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2813050855847403,
+      "learning_rate": 0.00012553920563136418,
+      "loss": 1.1681,
+      "step": 9342
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2798799129707208,
+      "learning_rate": 0.00012552391043384718,
+      "loss": 1.1568,
+      "step": 9343
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30424213764742,
+      "learning_rate": 0.00012550861459757835,
+      "loss": 1.1593,
+      "step": 9344
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32261170593030764,
+      "learning_rate": 0.00012549331812294033,
+      "loss": 1.0733,
+      "step": 9345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28695317695752437,
+      "learning_rate": 0.00012547802101031604,
+      "loss": 1.0534,
+      "step": 9346
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.297270830182022,
+      "learning_rate": 0.00012546272326008828,
+      "loss": 0.9348,
+      "step": 9347
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.32798136328441224,
+      "learning_rate": 0.00012544742487263983,
+      "loss": 1.2133,
+      "step": 9348
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.276438841709617,
+      "learning_rate": 0.00012543212584835363,
+      "loss": 0.9188,
+      "step": 9349
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30713210860480056,
+      "learning_rate": 0.00012541682618761243,
+      "loss": 0.936,
+      "step": 9350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3058264322272426,
+      "learning_rate": 0.00012540152589079922,
+      "loss": 1.0953,
+      "step": 9351
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3086810532233868,
+      "learning_rate": 0.00012538622495829687,
+      "loss": 1.0574,
+      "step": 9352
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.26492015598732366,
+      "learning_rate": 0.00012537092339048829,
+      "loss": 1.0913,
+      "step": 9353
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2613494790949844,
+      "learning_rate": 0.00012535562118775638,
+      "loss": 1.0103,
+      "step": 9354
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2745257568238265,
+      "learning_rate": 0.00012534031835048412,
+      "loss": 1.016,
+      "step": 9355
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32547504865315513,
+      "learning_rate": 0.00012532501487905447,
+      "loss": 1.0429,
+      "step": 9356
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25947151074133556,
+      "learning_rate": 0.0001253097107738504,
+      "loss": 1.0684,
+      "step": 9357
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2749564028335704,
+      "learning_rate": 0.00012529440603525495,
+      "loss": 1.0026,
+      "step": 9358
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26444889751740525,
+      "learning_rate": 0.00012527910066365108,
+      "loss": 1.1251,
+      "step": 9359
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24760688189734706,
+      "learning_rate": 0.00012526379465942179,
+      "loss": 1.0365,
+      "step": 9360
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2952587878180295,
+      "learning_rate": 0.00012524848802295018,
+      "loss": 1.1582,
+      "step": 9361
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27540696444772433,
+      "learning_rate": 0.0001252331807546193,
+      "loss": 1.1061,
+      "step": 9362
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27595222755388815,
+      "learning_rate": 0.00012521787285481222,
+      "loss": 1.0053,
+      "step": 9363
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3078665527994498,
+      "learning_rate": 0.00012520256432391197,
+      "loss": 1.0609,
+      "step": 9364
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30912162354934847,
+      "learning_rate": 0.00012518725516230176,
+      "loss": 1.092,
+      "step": 9365
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24850758690847607,
+      "learning_rate": 0.00012517194537036463,
+      "loss": 0.9636,
+      "step": 9366
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25462201249484734,
+      "learning_rate": 0.00012515663494848378,
+      "loss": 1.0612,
+      "step": 9367
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30057529749822187,
+      "learning_rate": 0.0001251413238970423,
+      "loss": 1.1008,
+      "step": 9368
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28606400293322526,
+      "learning_rate": 0.00012512601221642338,
+      "loss": 0.9646,
+      "step": 9369
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2718761895079578,
+      "learning_rate": 0.00012511069990701022,
+      "loss": 1.1127,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26749842342835667,
+      "learning_rate": 0.00012509538696918606,
+      "loss": 1.0016,
+      "step": 9371
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24422845567032508,
+      "learning_rate": 0.00012508007340333402,
+      "loss": 1.0075,
+      "step": 9372
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26485906439128654,
+      "learning_rate": 0.00012506475920983742,
+      "loss": 1.0334,
+      "step": 9373
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.23407388212556063,
+      "learning_rate": 0.00012504944438907945,
+      "loss": 0.9974,
+      "step": 9374
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2948257650153299,
+      "learning_rate": 0.00012503412894144337,
+      "loss": 1.0004,
+      "step": 9375
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2624520239691215,
+      "learning_rate": 0.0001250188128673125,
+      "loss": 1.113,
+      "step": 9376
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28609464800481466,
+      "learning_rate": 0.00012500349616707013,
+      "loss": 0.9897,
+      "step": 9377
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2647060271102632,
+      "learning_rate": 0.0001249881788410995,
+      "loss": 1.1142,
+      "step": 9378
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3319227336941671,
+      "learning_rate": 0.00012497286088978407,
+      "loss": 1.1371,
+      "step": 9379
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.290890609063919,
+      "learning_rate": 0.00012495754231350704,
+      "loss": 0.9454,
+      "step": 9380
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.275011083994787,
+      "learning_rate": 0.00012494222311265185,
+      "loss": 1.0498,
+      "step": 9381
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29328964264372026,
+      "learning_rate": 0.00012492690328760184,
+      "loss": 1.2401,
+      "step": 9382
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2849134201992309,
+      "learning_rate": 0.00012491158283874042,
+      "loss": 1.0384,
+      "step": 9383
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2644889309340255,
+      "learning_rate": 0.00012489626176645098,
+      "loss": 1.0729,
+      "step": 9384
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2671846138345653,
+      "learning_rate": 0.00012488094007111694,
+      "loss": 1.0643,
+      "step": 9385
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28342755899615574,
+      "learning_rate": 0.00012486561775312176,
+      "loss": 1.0652,
+      "step": 9386
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2708718366533827,
+      "learning_rate": 0.00012485029481284883,
+      "loss": 1.0131,
+      "step": 9387
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2893275903976366,
+      "learning_rate": 0.00012483497125068168,
+      "loss": 1.1254,
+      "step": 9388
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29651203987564034,
+      "learning_rate": 0.00012481964706700374,
+      "loss": 1.0964,
+      "step": 9389
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2668243860412548,
+      "learning_rate": 0.00012480432226219857,
+      "loss": 1.0646,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2607022670832829,
+      "learning_rate": 0.0001247889968366496,
+      "loss": 1.094,
+      "step": 9391
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3032946090686901,
+      "learning_rate": 0.00012477367079074045,
+      "loss": 1.0649,
+      "step": 9392
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31332018726986793,
+      "learning_rate": 0.0001247583441248546,
+      "loss": 1.0273,
+      "step": 9393
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25149055452189184,
+      "learning_rate": 0.00012474301683937562,
+      "loss": 1.0565,
+      "step": 9394
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31857711281294315,
+      "learning_rate": 0.00012472768893468712,
+      "loss": 1.1011,
+      "step": 9395
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27584446547396924,
+      "learning_rate": 0.00012471236041117263,
+      "loss": 1.1056,
+      "step": 9396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2710593279658963,
+      "learning_rate": 0.00012469703126921582,
+      "loss": 1.0088,
+      "step": 9397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2924314964891264,
+      "learning_rate": 0.00012468170150920028,
+      "loss": 1.0443,
+      "step": 9398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3100262243859713,
+      "learning_rate": 0.00012466637113150964,
+      "loss": 1.068,
+      "step": 9399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3203008665877059,
+      "learning_rate": 0.00012465104013652755,
+      "loss": 1.0662,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28159144815769815,
+      "learning_rate": 0.00012463570852463767,
+      "loss": 1.0416,
+      "step": 9401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2670293148269375,
+      "learning_rate": 0.00012462037629622374,
+      "loss": 1.0158,
+      "step": 9402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28251780180802666,
+      "learning_rate": 0.00012460504345166942,
+      "loss": 1.0844,
+      "step": 9403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2762339006289285,
+      "learning_rate": 0.00012458970999135839,
+      "loss": 1.0557,
+      "step": 9404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26743346835091364,
+      "learning_rate": 0.00012457437591567442,
+      "loss": 1.0952,
+      "step": 9405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28037210231320464,
+      "learning_rate": 0.00012455904122500128,
+      "loss": 0.9932,
+      "step": 9406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2629358216186376,
+      "learning_rate": 0.00012454370591972268,
+      "loss": 0.9753,
+      "step": 9407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2749296239864377,
+      "learning_rate": 0.0001245283700002224,
+      "loss": 1.0294,
+      "step": 9408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3005194418484203,
+      "learning_rate": 0.00012451303346688424,
+      "loss": 0.9774,
+      "step": 9409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2994528395818742,
+      "learning_rate": 0.00012449769632009205,
+      "loss": 1.0247,
+      "step": 9410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2491825962382407,
+      "learning_rate": 0.00012448235856022958,
+      "loss": 1.0416,
+      "step": 9411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24946941218248858,
+      "learning_rate": 0.0001244670201876807,
+      "loss": 1.0701,
+      "step": 9412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30337869321931366,
+      "learning_rate": 0.0001244516812028293,
+      "loss": 1.1049,
+      "step": 9413
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30488539436890616,
+      "learning_rate": 0.00012443634160605918,
+      "loss": 1.0473,
+      "step": 9414
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.307492284380008,
+      "learning_rate": 0.00012442100139775425,
+      "loss": 1.0344,
+      "step": 9415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3091356412029321,
+      "learning_rate": 0.00012440566057829843,
+      "loss": 1.0037,
+      "step": 9416
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27659203912070673,
+      "learning_rate": 0.0001243903191480756,
+      "loss": 0.9788,
+      "step": 9417
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.295213821134966,
+      "learning_rate": 0.00012437497710746974,
+      "loss": 1.0793,
+      "step": 9418
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2921351405650045,
+      "learning_rate": 0.00012435963445686472,
+      "loss": 1.0692,
+      "step": 9419
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32103907136316706,
+      "learning_rate": 0.00012434429119664457,
+      "loss": 1.1415,
+      "step": 9420
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30821488009357795,
+      "learning_rate": 0.0001243289473271932,
+      "loss": 1.1574,
+      "step": 9421
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3020517741494569,
+      "learning_rate": 0.00012431360284889464,
+      "loss": 1.1252,
+      "step": 9422
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30111749627739215,
+      "learning_rate": 0.0001242982577621329,
+      "loss": 1.0815,
+      "step": 9423
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2654478388360173,
+      "learning_rate": 0.000124282912067292,
+      "loss": 1.0287,
+      "step": 9424
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2845065279316126,
+      "learning_rate": 0.00012426756576475593,
+      "loss": 1.0703,
+      "step": 9425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.24604429019373228,
+      "learning_rate": 0.00012425221885490882,
+      "loss": 0.982,
+      "step": 9426
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3204878522592504,
+      "learning_rate": 0.00012423687133813466,
+      "loss": 0.9874,
+      "step": 9427
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2698848880583595,
+      "learning_rate": 0.00012422152321481754,
+      "loss": 1.1232,
+      "step": 9428
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2956334280898767,
+      "learning_rate": 0.00012420617448534162,
+      "loss": 1.1245,
+      "step": 9429
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28215348025828507,
+      "learning_rate": 0.00012419082515009093,
+      "loss": 1.0598,
+      "step": 9430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29800328339448867,
+      "learning_rate": 0.00012417547520944967,
+      "loss": 0.9549,
+      "step": 9431
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2758307830968502,
+      "learning_rate": 0.00012416012466380194,
+      "loss": 1.0299,
+      "step": 9432
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.27789615338120255,
+      "learning_rate": 0.00012414477351353192,
+      "loss": 0.955,
+      "step": 9433
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28216151632355363,
+      "learning_rate": 0.00012412942175902376,
+      "loss": 1.0228,
+      "step": 9434
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3646066899027749,
+      "learning_rate": 0.00012411406940066163,
+      "loss": 1.2245,
+      "step": 9435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28097256754194205,
+      "learning_rate": 0.0001240987164388298,
+      "loss": 1.0332,
+      "step": 9436
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.25589371187045934,
+      "learning_rate": 0.00012408336287391243,
+      "loss": 1.0571,
+      "step": 9437
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2924108280699943,
+      "learning_rate": 0.00012406800870629373,
+      "loss": 1.0727,
+      "step": 9438
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26309668552925775,
+      "learning_rate": 0.00012405265393635804,
+      "loss": 0.9953,
+      "step": 9439
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3055036695706396,
+      "learning_rate": 0.00012403729856448956,
+      "loss": 1.1705,
+      "step": 9440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26404553824001786,
+      "learning_rate": 0.00012402194259107256,
+      "loss": 0.8035,
+      "step": 9441
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.263456979982817,
+      "learning_rate": 0.00012400658601649135,
+      "loss": 1.0515,
+      "step": 9442
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2697295571066528,
+      "learning_rate": 0.00012399122884113024,
+      "loss": 1.0783,
+      "step": 9443
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3032996814397787,
+      "learning_rate": 0.00012397587106537355,
+      "loss": 1.107,
+      "step": 9444
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.30209747282548055,
+      "learning_rate": 0.0001239605126896056,
+      "loss": 1.0748,
+      "step": 9445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2541798522215149,
+      "learning_rate": 0.0001239451537142108,
+      "loss": 0.9973,
+      "step": 9446
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26055885474176194,
+      "learning_rate": 0.0001239297941395735,
+      "loss": 0.9793,
+      "step": 9447
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2937502415391966,
+      "learning_rate": 0.00012391443396607798,
+      "loss": 1.0501,
+      "step": 9448
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2484096340317221,
+      "learning_rate": 0.00012389907319410877,
+      "loss": 0.9737,
+      "step": 9449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34708042996433486,
+      "learning_rate": 0.00012388371182405023,
+      "loss": 1.0665,
+      "step": 9450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2870344403909073,
+      "learning_rate": 0.0001238683498562868,
+      "loss": 1.0346,
+      "step": 9451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.32368230775180656,
+      "learning_rate": 0.00012385298729120287,
+      "loss": 0.9915,
+      "step": 9452
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2902156260509134,
+      "learning_rate": 0.00012383762412918297,
+      "loss": 1.0838,
+      "step": 9453
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26483538433540343,
+      "learning_rate": 0.00012382226037061157,
+      "loss": 1.0152,
+      "step": 9454
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31506688541873007,
+      "learning_rate": 0.0001238068960158731,
+      "loss": 1.0498,
+      "step": 9455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31606368328175294,
+      "learning_rate": 0.00012379153106535212,
+      "loss": 1.0568,
+      "step": 9456
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.26749250980306544,
+      "learning_rate": 0.00012377616551943312,
+      "loss": 1.1441,
+      "step": 9457
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2805411852480236,
+      "learning_rate": 0.0001237607993785006,
+      "loss": 1.0273,
+      "step": 9458
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29590637384614576,
+      "learning_rate": 0.0001237454326429392,
+      "loss": 1.2318,
+      "step": 9459
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.277157897803464,
+      "learning_rate": 0.00012373006531313338,
+      "loss": 1.0606,
+      "step": 9460
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2820422335962146,
+      "learning_rate": 0.0001237146973894678,
+      "loss": 1.0724,
+      "step": 9461
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28970884694413745,
+      "learning_rate": 0.00012369932887232695,
+      "loss": 1.0501,
+      "step": 9462
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26821460214493453,
+      "learning_rate": 0.00012368395976209554,
+      "loss": 1.1002,
+      "step": 9463
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3020120162788039,
+      "learning_rate": 0.00012366859005915817,
+      "loss": 1.2129,
+      "step": 9464
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25290616792649784,
+      "learning_rate": 0.00012365321976389942,
+      "loss": 0.9608,
+      "step": 9465
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3118979180099709,
+      "learning_rate": 0.000123637848876704,
+      "loss": 1.0709,
+      "step": 9466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31154007441669707,
+      "learning_rate": 0.00012362247739795658,
+      "loss": 1.0235,
+      "step": 9467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28995095468089827,
+      "learning_rate": 0.00012360710532804178,
+      "loss": 1.201,
+      "step": 9468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3110913941186855,
+      "learning_rate": 0.00012359173266734435,
+      "loss": 1.1567,
+      "step": 9469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27514587325982237,
+      "learning_rate": 0.00012357635941624898,
+      "loss": 1.0559,
+      "step": 9470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30517981996469423,
+      "learning_rate": 0.00012356098557514037,
+      "loss": 1.0068,
+      "step": 9471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2725388181887218,
+      "learning_rate": 0.00012354561114440334,
+      "loss": 0.9899,
+      "step": 9472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.297899718417481,
+      "learning_rate": 0.00012353023612442254,
+      "loss": 1.0983,
+      "step": 9473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2848396496021322,
+      "learning_rate": 0.00012351486051558283,
+      "loss": 1.0539,
+      "step": 9474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2852186312128984,
+      "learning_rate": 0.00012349948431826895,
+      "loss": 1.0044,
+      "step": 9475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29012020789960563,
+      "learning_rate": 0.0001234841075328657,
+      "loss": 0.9673,
+      "step": 9476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24672860443444639,
+      "learning_rate": 0.0001234687301597579,
+      "loss": 1.0161,
+      "step": 9477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2726877358325203,
+      "learning_rate": 0.0001234533521993304,
+      "loss": 1.0621,
+      "step": 9478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28948332424230283,
+      "learning_rate": 0.00012343797365196797,
+      "loss": 0.9913,
+      "step": 9479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26969250387705407,
+      "learning_rate": 0.00012342259451805557,
+      "loss": 1.1202,
+      "step": 9480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3258916039185605,
+      "learning_rate": 0.000123407214797978,
+      "loss": 1.0168,
+      "step": 9481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25973605245424924,
+      "learning_rate": 0.00012339183449212017,
+      "loss": 1.0874,
+      "step": 9482
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2676443845035316,
+      "learning_rate": 0.00012337645360086698,
+      "loss": 0.9667,
+      "step": 9483
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30203903279304206,
+      "learning_rate": 0.00012336107212460338,
+      "loss": 1.0537,
+      "step": 9484
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27476191579196646,
+      "learning_rate": 0.00012334569006371422,
+      "loss": 1.0224,
+      "step": 9485
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26328678406410266,
+      "learning_rate": 0.0001233303074185845,
+      "loss": 1.1169,
+      "step": 9486
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29307038009665376,
+      "learning_rate": 0.0001233149241895992,
+      "loss": 1.0547,
+      "step": 9487
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3029822127610837,
+      "learning_rate": 0.00012329954037714326,
+      "loss": 1.0455,
+      "step": 9488
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31316815984313306,
+      "learning_rate": 0.00012328415598160167,
+      "loss": 1.0461,
+      "step": 9489
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26416179108140025,
+      "learning_rate": 0.00012326877100335946,
+      "loss": 0.9973,
+      "step": 9490
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27917647838809806,
+      "learning_rate": 0.0001232533854428016,
+      "loss": 1.1014,
+      "step": 9491
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2880924083809222,
+      "learning_rate": 0.00012323799930031318,
+      "loss": 1.1198,
+      "step": 9492
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2738883771003005,
+      "learning_rate": 0.0001232226125762792,
+      "loss": 0.9753,
+      "step": 9493
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28505026809880435,
+      "learning_rate": 0.00012320722527108476,
+      "loss": 1.0131,
+      "step": 9494
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2700388433926024,
+      "learning_rate": 0.00012319183738511495,
+      "loss": 1.0627,
+      "step": 9495
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.21724216250627967,
+      "learning_rate": 0.0001231764489187548,
+      "loss": 0.969,
+      "step": 9496
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30567642004976847,
+      "learning_rate": 0.00012316105987238946,
+      "loss": 1.1282,
+      "step": 9497
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2960952448251748,
+      "learning_rate": 0.00012314567024640405,
+      "loss": 1.0719,
+      "step": 9498
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2679330934652599,
+      "learning_rate": 0.00012313028004118368,
+      "loss": 1.0671,
+      "step": 9499
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30977923520733813,
+      "learning_rate": 0.00012311488925711352,
+      "loss": 1.0649,
+      "step": 9500
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2877059133094227,
+      "learning_rate": 0.00012309949789457872,
+      "loss": 0.99,
+      "step": 9501
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2747597630168646,
+      "learning_rate": 0.0001230841059539645,
+      "loss": 1.0708,
+      "step": 9502
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2849182513539573,
+      "learning_rate": 0.00012306871343565598,
+      "loss": 1.1457,
+      "step": 9503
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2816422560556322,
+      "learning_rate": 0.00012305332034003843,
+      "loss": 1.0653,
+      "step": 9504
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2449196404919519,
+      "learning_rate": 0.00012303792666749704,
+      "loss": 1.0025,
+      "step": 9505
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28277588561340145,
+      "learning_rate": 0.00012302253241841705,
+      "loss": 1.0981,
+      "step": 9506
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28881985968739604,
+      "learning_rate": 0.00012300713759318374,
+      "loss": 0.8564,
+      "step": 9507
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2658935819955603,
+      "learning_rate": 0.00012299174219218236,
+      "loss": 0.9685,
+      "step": 9508
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.33536541922122687,
+      "learning_rate": 0.00012297634621579815,
+      "loss": 1.0644,
+      "step": 9509
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2883666884244152,
+      "learning_rate": 0.00012296094966441644,
+      "loss": 1.0163,
+      "step": 9510
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2874022639714324,
+      "learning_rate": 0.00012294555253842258,
+      "loss": 1.0547,
+      "step": 9511
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28615282013866633,
+      "learning_rate": 0.0001229301548382018,
+      "loss": 1.0267,
+      "step": 9512
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31559288068892855,
+      "learning_rate": 0.0001229147565641395,
+      "loss": 1.0427,
+      "step": 9513
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29067613727477026,
+      "learning_rate": 0.000122899357716621,
+      "loss": 1.0771,
+      "step": 9514
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3091379944841265,
+      "learning_rate": 0.00012288395829603168,
+      "loss": 1.1228,
+      "step": 9515
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34288634284416575,
+      "learning_rate": 0.0001228685583027569,
+      "loss": 1.07,
+      "step": 9516
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.25806670007511917,
+      "learning_rate": 0.00012285315773718215,
+      "loss": 1.177,
+      "step": 9517
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28617035042389005,
+      "learning_rate": 0.00012283775659969272,
+      "loss": 0.9902,
+      "step": 9518
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2699590057863018,
+      "learning_rate": 0.00012282235489067406,
+      "loss": 1.0114,
+      "step": 9519
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2668076542427271,
+      "learning_rate": 0.00012280695261051168,
+      "loss": 1.0479,
+      "step": 9520
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27069952546890097,
+      "learning_rate": 0.00012279154975959093,
+      "loss": 0.9784,
+      "step": 9521
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26601560597027246,
+      "learning_rate": 0.00012277614633829736,
+      "loss": 0.9275,
+      "step": 9522
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.34915490477463507,
+      "learning_rate": 0.00012276074234701637,
+      "loss": 1.0747,
+      "step": 9523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2963938117016832,
+      "learning_rate": 0.00012274533778613354,
+      "loss": 0.968,
+      "step": 9524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28627744365026025,
+      "learning_rate": 0.00012272993265603432,
+      "loss": 1.0285,
+      "step": 9525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2543047213053924,
+      "learning_rate": 0.00012271452695710423,
+      "loss": 0.9785,
+      "step": 9526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30369638655565206,
+      "learning_rate": 0.00012269912068972887,
+      "loss": 1.0088,
+      "step": 9527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30048051465973585,
+      "learning_rate": 0.00012268371385429371,
+      "loss": 1.0032,
+      "step": 9528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24363694434494604,
+      "learning_rate": 0.0001226683064511844,
+      "loss": 1.083,
+      "step": 9529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27300145834174794,
+      "learning_rate": 0.0001226528984807864,
+      "loss": 1.056,
+      "step": 9530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2840057926179831,
+      "learning_rate": 0.00012263748994348543,
+      "loss": 1.0906,
+      "step": 9531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2867767195608796,
+      "learning_rate": 0.00012262208083966707,
+      "loss": 1.0343,
+      "step": 9532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30097203291437813,
+      "learning_rate": 0.00012260667116971687,
+      "loss": 1.1535,
+      "step": 9533
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26057797605093236,
+      "learning_rate": 0.0001225912609340205,
+      "loss": 0.9956,
+      "step": 9534
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2722485625219177,
+      "learning_rate": 0.00012257585013296368,
+      "loss": 0.9609,
+      "step": 9535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31674933662782234,
+      "learning_rate": 0.00012256043876693199,
+      "loss": 1.1484,
+      "step": 9536
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24732657574619976,
+      "learning_rate": 0.00012254502683631114,
+      "loss": 1.11,
+      "step": 9537
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.302444859564973,
+      "learning_rate": 0.00012252961434148685,
+      "loss": 1.1601,
+      "step": 9538
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3039631031521198,
+      "learning_rate": 0.0001225142012828448,
+      "loss": 1.1319,
+      "step": 9539
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2833615843410249,
+      "learning_rate": 0.0001224987876607707,
+      "loss": 1.1211,
+      "step": 9540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24087751384697798,
+      "learning_rate": 0.0001224833734756503,
+      "loss": 0.9828,
+      "step": 9541
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3022341870078711,
+      "learning_rate": 0.00012246795872786938,
+      "loss": 1.0778,
+      "step": 9542
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3125008887305215,
+      "learning_rate": 0.00012245254341781362,
+      "loss": 0.9715,
+      "step": 9543
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26961102066302006,
+      "learning_rate": 0.0001224371275458689,
+      "loss": 1.0219,
+      "step": 9544
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30552221121015993,
+      "learning_rate": 0.00012242171111242093,
+      "loss": 0.9941,
+      "step": 9545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30365825120398526,
+      "learning_rate": 0.00012240629411785557,
+      "loss": 0.9531,
+      "step": 9546
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27544518032809634,
+      "learning_rate": 0.0001223908765625586,
+      "loss": 1.0634,
+      "step": 9547
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2469345335356712,
+      "learning_rate": 0.00012237545844691585,
+      "loss": 1.0724,
+      "step": 9548
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2956105861133632,
+      "learning_rate": 0.00012236003977131324,
+      "loss": 1.1523,
+      "step": 9549
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2968897322282517,
+      "learning_rate": 0.00012234462053613653,
+      "loss": 1.07,
+      "step": 9550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.26572173065636373,
+      "learning_rate": 0.0001223292007417717,
+      "loss": 1.0555,
+      "step": 9551
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24837107366849048,
+      "learning_rate": 0.00012231378038860455,
+      "loss": 1.1674,
+      "step": 9552
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.27292429243707583,
+      "learning_rate": 0.00012229835947702103,
+      "loss": 1.1317,
+      "step": 9553
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.31026913454307414,
+      "learning_rate": 0.00012228293800740705,
+      "loss": 1.1446,
+      "step": 9554
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3072735487423723,
+      "learning_rate": 0.00012226751598014854,
+      "loss": 1.1987,
+      "step": 9555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24847902065517485,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 1.1041,
+      "step": 9556
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2924020829498602,
+      "learning_rate": 0.00012223667025424172,
+      "loss": 1.0721,
+      "step": 9557
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2613223633837016,
+      "learning_rate": 0.00012222124655636538,
+      "loss": 1.1266,
+      "step": 9558
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2894905328595956,
+      "learning_rate": 0.00012220582230238839,
+      "loss": 1.0732,
+      "step": 9559
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28681410216353886,
+      "learning_rate": 0.00012219039749269668,
+      "loss": 1.0915,
+      "step": 9560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2943586385473255,
+      "learning_rate": 0.00012217497212767636,
+      "loss": 1.0788,
+      "step": 9561
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.285929389887662,
+      "learning_rate": 0.00012215954620771344,
+      "loss": 1.1005,
+      "step": 9562
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3149078861825459,
+      "learning_rate": 0.00012214411973319396,
+      "loss": 1.0933,
+      "step": 9563
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.24280112314190866,
+      "learning_rate": 0.0001221286927045039,
+      "loss": 0.954,
+      "step": 9564
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24225580439211833,
+      "learning_rate": 0.00012211326512202945,
+      "loss": 0.9918,
+      "step": 9565
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3066920417822219,
+      "learning_rate": 0.00012209783698615665,
+      "loss": 1.0828,
+      "step": 9566
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3112335184826233,
+      "learning_rate": 0.00012208240829727156,
+      "loss": 1.1426,
+      "step": 9567
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.263752238595472,
+      "learning_rate": 0.00012206697905576034,
+      "loss": 1.0959,
+      "step": 9568
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3013509779853578,
+      "learning_rate": 0.0001220515492620091,
+      "loss": 1.0237,
+      "step": 9569
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29198852460496516,
+      "learning_rate": 0.00012203611891640398,
+      "loss": 1.1627,
+      "step": 9570
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26850931222903324,
+      "learning_rate": 0.00012202068801933112,
+      "loss": 1.1425,
+      "step": 9571
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30251674022231434,
+      "learning_rate": 0.00012200525657117673,
+      "loss": 1.1273,
+      "step": 9572
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3018015700917101,
+      "learning_rate": 0.00012198982457232698,
+      "loss": 0.9809,
+      "step": 9573
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.273598131245597,
+      "learning_rate": 0.000121974392023168,
+      "loss": 1.0571,
+      "step": 9574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30277174215701025,
+      "learning_rate": 0.00012195895892408609,
+      "loss": 0.9031,
+      "step": 9575
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28490196205477425,
+      "learning_rate": 0.00012194352527546739,
+      "loss": 1.0131,
+      "step": 9576
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2715100350623435,
+      "learning_rate": 0.0001219280910776982,
+      "loss": 1.0771,
+      "step": 9577
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2948896697832583,
+      "learning_rate": 0.00012191265633116473,
+      "loss": 1.0838,
+      "step": 9578
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2570476008819864,
+      "learning_rate": 0.00012189722103625332,
+      "loss": 1.0897,
+      "step": 9579
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2923182653513327,
+      "learning_rate": 0.00012188178519335014,
+      "loss": 1.063,
+      "step": 9580
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24843565354969446,
+      "learning_rate": 0.00012186634880284155,
+      "loss": 1.0832,
+      "step": 9581
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30405197780168924,
+      "learning_rate": 0.00012185091186511383,
+      "loss": 1.0964,
+      "step": 9582
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28552359904155467,
+      "learning_rate": 0.00012183547438055334,
+      "loss": 1.0048,
+      "step": 9583
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25623311580886565,
+      "learning_rate": 0.00012182003634954635,
+      "loss": 1.1127,
+      "step": 9584
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28043711484669764,
+      "learning_rate": 0.00012180459777247924,
+      "loss": 0.9836,
+      "step": 9585
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29306608514635557,
+      "learning_rate": 0.00012178915864973839,
+      "loss": 1.0981,
+      "step": 9586
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2509994097045669,
+      "learning_rate": 0.00012177371898171011,
+      "loss": 1.0955,
+      "step": 9587
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25595385108043606,
+      "learning_rate": 0.00012175827876878085,
+      "loss": 1.0905,
+      "step": 9588
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29970697558699483,
+      "learning_rate": 0.00012174283801133701,
+      "loss": 1.08,
+      "step": 9589
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25132811288192697,
+      "learning_rate": 0.00012172739670976497,
+      "loss": 0.9933,
+      "step": 9590
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2795967397200255,
+      "learning_rate": 0.00012171195486445115,
+      "loss": 1.0812,
+      "step": 9591
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27861509817589664,
+      "learning_rate": 0.00012169651247578205,
+      "loss": 1.0368,
+      "step": 9592
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2916254831937168,
+      "learning_rate": 0.00012168106954414406,
+      "loss": 1.0809,
+      "step": 9593
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2922945494777507,
+      "learning_rate": 0.00012166562606992368,
+      "loss": 1.1979,
+      "step": 9594
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27637344653802837,
+      "learning_rate": 0.0001216501820535074,
+      "loss": 0.957,
+      "step": 9595
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2890101477293797,
+      "learning_rate": 0.00012163473749528169,
+      "loss": 1.0743,
+      "step": 9596
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27375555736611096,
+      "learning_rate": 0.0001216192923956331,
+      "loss": 1.0294,
+      "step": 9597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3020964878540027,
+      "learning_rate": 0.00012160384675494811,
+      "loss": 1.117,
+      "step": 9598
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31564811381117275,
+      "learning_rate": 0.00012158840057361332,
+      "loss": 1.1155,
+      "step": 9599
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30570724996592125,
+      "learning_rate": 0.00012157295385201522,
+      "loss": 1.0954,
+      "step": 9600
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2772559483117768,
+      "learning_rate": 0.00012155750659054035,
+      "loss": 1.0597,
+      "step": 9601
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27996328612745197,
+      "learning_rate": 0.00012154205878957539,
+      "loss": 1.0686,
+      "step": 9602
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2862321218720333,
+      "learning_rate": 0.00012152661044950684,
+      "loss": 0.9934,
+      "step": 9603
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2853452333425892,
+      "learning_rate": 0.00012151116157072132,
+      "loss": 1.1348,
+      "step": 9604
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2617628600038958,
+      "learning_rate": 0.00012149571215360547,
+      "loss": 0.97,
+      "step": 9605
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.25696280278663025,
+      "learning_rate": 0.00012148026219854594,
+      "loss": 1.0261,
+      "step": 9606
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.310031454374293,
+      "learning_rate": 0.00012146481170592933,
+      "loss": 1.0158,
+      "step": 9607
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2790355327820539,
+      "learning_rate": 0.0001214493606761423,
+      "loss": 1.1757,
+      "step": 9608
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30770923269815564,
+      "learning_rate": 0.00012143390910957157,
+      "loss": 1.1518,
+      "step": 9609
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30029358695910546,
+      "learning_rate": 0.00012141845700660379,
+      "loss": 1.1968,
+      "step": 9610
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3098726943709629,
+      "learning_rate": 0.00012140300436762564,
+      "loss": 1.0352,
+      "step": 9611
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2959419907143475,
+      "learning_rate": 0.00012138755119302388,
+      "loss": 1.0186,
+      "step": 9612
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27785698197656766,
+      "learning_rate": 0.00012137209748318521,
+      "loss": 1.0375,
+      "step": 9613
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2930400342362595,
+      "learning_rate": 0.00012135664323849634,
+      "loss": 0.9418,
+      "step": 9614
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28085631013272555,
+      "learning_rate": 0.00012134118845934405,
+      "loss": 1.0465,
+      "step": 9615
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2541256790591151,
+      "learning_rate": 0.00012132573314611516,
+      "loss": 1.0293,
+      "step": 9616
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28908641994102546,
+      "learning_rate": 0.00012131027729919633,
+      "loss": 1.078,
+      "step": 9617
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2772893250241189,
+      "learning_rate": 0.00012129482091897446,
+      "loss": 1.1151,
+      "step": 9618
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.327164314875272,
+      "learning_rate": 0.00012127936400583629,
+      "loss": 1.0591,
+      "step": 9619
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2886231869799765,
+      "learning_rate": 0.00012126390656016866,
+      "loss": 1.0113,
+      "step": 9620
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.283680263633664,
+      "learning_rate": 0.0001212484485823584,
+      "loss": 1.0249,
+      "step": 9621
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26111321954434336,
+      "learning_rate": 0.00012123299007279238,
+      "loss": 1.0342,
+      "step": 9622
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2955195301556354,
+      "learning_rate": 0.00012121753103185745,
+      "loss": 0.9461,
+      "step": 9623
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.27409190677236495,
+      "learning_rate": 0.00012120207145994045,
+      "loss": 1.0637,
+      "step": 9624
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26924614303903166,
+      "learning_rate": 0.00012118661135742828,
+      "loss": 0.9253,
+      "step": 9625
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2692477485784918,
+      "learning_rate": 0.00012117115072470788,
+      "loss": 1.0827,
+      "step": 9626
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26774970205249055,
+      "learning_rate": 0.0001211556895621661,
+      "loss": 1.0528,
+      "step": 9627
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29867507929874776,
+      "learning_rate": 0.00012114022787018988,
+      "loss": 1.0436,
+      "step": 9628
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3208070747287111,
+      "learning_rate": 0.00012112476564916622,
+      "loss": 1.1365,
+      "step": 9629
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2972966476878782,
+      "learning_rate": 0.000121109302899482,
+      "loss": 1.1186,
+      "step": 9630
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30544954734093066,
+      "learning_rate": 0.00012109383962152416,
+      "loss": 1.1047,
+      "step": 9631
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2967657674672275,
+      "learning_rate": 0.00012107837581567977,
+      "loss": 1.072,
+      "step": 9632
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3033096939499795,
+      "learning_rate": 0.00012106291148233579,
+      "loss": 1.0254,
+      "step": 9633
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26626395824315857,
+      "learning_rate": 0.00012104744662187922,
+      "loss": 0.9941,
+      "step": 9634
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2895890185705288,
+      "learning_rate": 0.00012103198123469704,
+      "loss": 1.0531,
+      "step": 9635
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2957215592529329,
+      "learning_rate": 0.00012101651532117632,
+      "loss": 1.0199,
+      "step": 9636
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2966423558454665,
+      "learning_rate": 0.00012100104888170407,
+      "loss": 0.9559,
+      "step": 9637
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3146917874326988,
+      "learning_rate": 0.00012098558191666742,
+      "loss": 1.1623,
+      "step": 9638
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33430522861396766,
+      "learning_rate": 0.00012097011442645337,
+      "loss": 1.0295,
+      "step": 9639
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2725271575959663,
+      "learning_rate": 0.00012095464641144902,
+      "loss": 1.0745,
+      "step": 9640
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.32836523657529365,
+      "learning_rate": 0.00012093917787204148,
+      "loss": 1.0739,
+      "step": 9641
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2905477137452141,
+      "learning_rate": 0.00012092370880861786,
+      "loss": 0.9218,
+      "step": 9642
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28312434963987126,
+      "learning_rate": 0.00012090823922156526,
+      "loss": 1.1302,
+      "step": 9643
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31119878698668474,
+      "learning_rate": 0.00012089276911127088,
+      "loss": 1.1067,
+      "step": 9644
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.24999334472319124,
+      "learning_rate": 0.00012087729847812176,
+      "loss": 1.0741,
+      "step": 9645
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2543284979471403,
+      "learning_rate": 0.00012086182732250517,
+      "loss": 0.9579,
+      "step": 9646
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2594464862081111,
+      "learning_rate": 0.00012084635564480824,
+      "loss": 1.0921,
+      "step": 9647
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3158864173446626,
+      "learning_rate": 0.00012083088344541813,
+      "loss": 1.1075,
+      "step": 9648
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29277153816070817,
+      "learning_rate": 0.00012081541072472208,
+      "loss": 1.0819,
+      "step": 9649
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.285551700825859,
+      "learning_rate": 0.00012079993748310729,
+      "loss": 1.1289,
+      "step": 9650
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26578837779844633,
+      "learning_rate": 0.00012078446372096102,
+      "loss": 1.0786,
+      "step": 9651
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2770368943423453,
+      "learning_rate": 0.00012076898943867046,
+      "loss": 1.2504,
+      "step": 9652
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2915791206912121,
+      "learning_rate": 0.0001207535146366229,
+      "loss": 1.0691,
+      "step": 9653
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29701485898418156,
+      "learning_rate": 0.00012073803931520557,
+      "loss": 1.0343,
+      "step": 9654
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3164931927910207,
+      "learning_rate": 0.0001207225634748058,
+      "loss": 1.0555,
+      "step": 9655
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.26381700408017583,
+      "learning_rate": 0.00012070708711581083,
+      "loss": 1.1291,
+      "step": 9656
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29962327944839096,
+      "learning_rate": 0.00012069161023860802,
+      "loss": 1.039,
+      "step": 9657
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3144393601015309,
+      "learning_rate": 0.00012067613284358461,
+      "loss": 1.1407,
+      "step": 9658
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29485729590956794,
+      "learning_rate": 0.00012066065493112803,
+      "loss": 0.9772,
+      "step": 9659
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3317695176709224,
+      "learning_rate": 0.00012064517650162555,
+      "loss": 1.0321,
+      "step": 9660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2795257128963298,
+      "learning_rate": 0.00012062969755546456,
+      "loss": 1.1162,
+      "step": 9661
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3308060051457206,
+      "learning_rate": 0.00012061421809303241,
+      "loss": 0.9852,
+      "step": 9662
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.284967097310632,
+      "learning_rate": 0.00012059873811471651,
+      "loss": 0.9865,
+      "step": 9663
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.31468502602583576,
+      "learning_rate": 0.00012058325762090426,
+      "loss": 1.0112,
+      "step": 9664
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2760079699322398,
+      "learning_rate": 0.00012056777661198301,
+      "loss": 0.9597,
+      "step": 9665
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2911837084304875,
+      "learning_rate": 0.00012055229508834027,
+      "loss": 1.0258,
+      "step": 9666
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.30981871374584724,
+      "learning_rate": 0.00012053681305036342,
+      "loss": 1.1091,
+      "step": 9667
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28199493834041306,
+      "learning_rate": 0.00012052133049843992,
+      "loss": 0.9613,
+      "step": 9668
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2833626576757252,
+      "learning_rate": 0.00012050584743295718,
+      "loss": 1.0918,
+      "step": 9669
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32179313490464345,
+      "learning_rate": 0.00012049036385430277,
+      "loss": 1.0589,
+      "step": 9670
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3529820688530236,
+      "learning_rate": 0.0001204748797628641,
+      "loss": 1.0392,
+      "step": 9671
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.316182685666834,
+      "learning_rate": 0.00012045939515902872,
+      "loss": 1.091,
+      "step": 9672
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3132427307808237,
+      "learning_rate": 0.00012044391004318409,
+      "loss": 1.1227,
+      "step": 9673
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2997545510559344,
+      "learning_rate": 0.0001204284244157178,
+      "loss": 1.0994,
+      "step": 9674
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2536805673335262,
+      "learning_rate": 0.00012041293827701729,
+      "loss": 0.9636,
+      "step": 9675
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2876440010059215,
+      "learning_rate": 0.00012039745162747022,
+      "loss": 1.1536,
+      "step": 9676
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2691609443880876,
+      "learning_rate": 0.0001203819644674641,
+      "loss": 1.1113,
+      "step": 9677
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25701727543705255,
+      "learning_rate": 0.00012036647679738649,
+      "loss": 0.9819,
+      "step": 9678
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29817078458954743,
+      "learning_rate": 0.00012035098861762502,
+      "loss": 0.9931,
+      "step": 9679
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2642831901167002,
+      "learning_rate": 0.00012033549992856726,
+      "loss": 1.0036,
+      "step": 9680
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29959073213056514,
+      "learning_rate": 0.00012032001073060082,
+      "loss": 1.1449,
+      "step": 9681
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32204374587710666,
+      "learning_rate": 0.00012030452102411333,
+      "loss": 0.9079,
+      "step": 9682
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30732832828204176,
+      "learning_rate": 0.00012028903080949248,
+      "loss": 0.8944,
+      "step": 9683
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27203865429830504,
+      "learning_rate": 0.00012027354008712588,
+      "loss": 0.9959,
+      "step": 9684
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31723507213333874,
+      "learning_rate": 0.0001202580488574012,
+      "loss": 1.0008,
+      "step": 9685
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2753881383976869,
+      "learning_rate": 0.00012024255712070607,
+      "loss": 0.9831,
+      "step": 9686
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2905972632173138,
+      "learning_rate": 0.00012022706487742827,
+      "loss": 0.9718,
+      "step": 9687
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3441351333384633,
+      "learning_rate": 0.00012021157212795544,
+      "loss": 1.1038,
+      "step": 9688
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30829573394192783,
+      "learning_rate": 0.00012019607887267532,
+      "loss": 1.0747,
+      "step": 9689
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3116599969382794,
+      "learning_rate": 0.00012018058511197563,
+      "loss": 0.9906,
+      "step": 9690
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.32153118671256326,
+      "learning_rate": 0.00012016509084624413,
+      "loss": 1.0987,
+      "step": 9691
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30020496331906704,
+      "learning_rate": 0.00012014959607586853,
+      "loss": 0.9893,
+      "step": 9692
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2488961913424749,
+      "learning_rate": 0.00012013410080123666,
+      "loss": 1.0128,
+      "step": 9693
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2894864987548495,
+      "learning_rate": 0.00012011860502273625,
+      "loss": 1.0898,
+      "step": 9694
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.339459814596149,
+      "learning_rate": 0.0001201031087407551,
+      "loss": 0.972,
+      "step": 9695
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28936642655855416,
+      "learning_rate": 0.00012008761195568101,
+      "loss": 1.1054,
+      "step": 9696
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29865336998518743,
+      "learning_rate": 0.00012007211466790183,
+      "loss": 1.0518,
+      "step": 9697
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29410810287938477,
+      "learning_rate": 0.00012005661687780537,
+      "loss": 1.0348,
+      "step": 9698
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28715006026168827,
+      "learning_rate": 0.00012004111858577942,
+      "loss": 1.0277,
+      "step": 9699
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3246035439472215,
+      "learning_rate": 0.00012002561979221191,
+      "loss": 1.0978,
+      "step": 9700
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28714503427033067,
+      "learning_rate": 0.00012001012049749067,
+      "loss": 1.0528,
+      "step": 9701
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2999877264540504,
+      "learning_rate": 0.0001199946207020036,
+      "loss": 1.0535,
+      "step": 9702
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2831842528289797,
+      "learning_rate": 0.00011997912040613856,
+      "loss": 1.0861,
+      "step": 9703
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27863916487179596,
+      "learning_rate": 0.00011996361961028351,
+      "loss": 1.0172,
+      "step": 9704
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3076428593870626,
+      "learning_rate": 0.00011994811831482633,
+      "loss": 1.154,
+      "step": 9705
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3178394621331127,
+      "learning_rate": 0.00011993261652015493,
+      "loss": 1.0459,
+      "step": 9706
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2834780610000245,
+      "learning_rate": 0.00011991711422665728,
+      "loss": 0.8991,
+      "step": 9707
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28248971237831144,
+      "learning_rate": 0.00011990161143472134,
+      "loss": 1.0412,
+      "step": 9708
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.37293406629713854,
+      "learning_rate": 0.00011988610814473504,
+      "loss": 1.1236,
+      "step": 9709
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29203647947031824,
+      "learning_rate": 0.00011987060435708643,
+      "loss": 0.972,
+      "step": 9710
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2784065820475021,
+      "learning_rate": 0.00011985510007216343,
+      "loss": 1.0975,
+      "step": 9711
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2967922411536282,
+      "learning_rate": 0.0001198395952903541,
+      "loss": 1.0979,
+      "step": 9712
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29200788752529727,
+      "learning_rate": 0.00011982409001204637,
+      "loss": 1.0726,
+      "step": 9713
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2692696205517818,
+      "learning_rate": 0.00011980858423762837,
+      "loss": 1.0631,
+      "step": 9714
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2965913502544387,
+      "learning_rate": 0.00011979307796748811,
+      "loss": 1.1334,
+      "step": 9715
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28309793277521694,
+      "learning_rate": 0.0001197775712020136,
+      "loss": 1.1148,
+      "step": 9716
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2842362819659992,
+      "learning_rate": 0.00011976206394159297,
+      "loss": 1.089,
+      "step": 9717
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30869020384451756,
+      "learning_rate": 0.00011974655618661425,
+      "loss": 1.1634,
+      "step": 9718
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2919947030319265,
+      "learning_rate": 0.00011973104793746554,
+      "loss": 0.9976,
+      "step": 9719
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28107687880463733,
+      "learning_rate": 0.000119715539194535,
+      "loss": 1.0656,
+      "step": 9720
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2538194528442572,
+      "learning_rate": 0.00011970002995821069,
+      "loss": 0.9763,
+      "step": 9721
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3042314896103001,
+      "learning_rate": 0.0001196845202288807,
+      "loss": 1.1705,
+      "step": 9722
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2572188222835515,
+      "learning_rate": 0.00011966901000693325,
+      "loss": 1.0154,
+      "step": 9723
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28736803000564753,
+      "learning_rate": 0.00011965349929275646,
+      "loss": 0.9974,
+      "step": 9724
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31110944588178774,
+      "learning_rate": 0.00011963798808673852,
+      "loss": 1.0147,
+      "step": 9725
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2875577964788188,
+      "learning_rate": 0.00011962247638926755,
+      "loss": 1.0678,
+      "step": 9726
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2700897281383975,
+      "learning_rate": 0.00011960696420073181,
+      "loss": 1.1177,
+      "step": 9727
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2806117435648341,
+      "learning_rate": 0.00011959145152151947,
+      "loss": 1.012,
+      "step": 9728
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31209826283604314,
+      "learning_rate": 0.00011957593835201875,
+      "loss": 0.9591,
+      "step": 9729
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2750239320324656,
+      "learning_rate": 0.00011956042469261781,
+      "loss": 1.1095,
+      "step": 9730
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28264832045438565,
+      "learning_rate": 0.000119544910543705,
+      "loss": 1.037,
+      "step": 9731
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3300651441181935,
+      "learning_rate": 0.00011952939590566852,
+      "loss": 1.0304,
+      "step": 9732
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2859521337438915,
+      "learning_rate": 0.0001195138807788966,
+      "loss": 0.969,
+      "step": 9733
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2828356666065811,
+      "learning_rate": 0.00011949836516377759,
+      "loss": 1.004,
+      "step": 9734
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2696647965723336,
+      "learning_rate": 0.00011948284906069974,
+      "loss": 1.089,
+      "step": 9735
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2708560032776781,
+      "learning_rate": 0.00011946733247005131,
+      "loss": 0.9534,
+      "step": 9736
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3142091199851108,
+      "learning_rate": 0.00011945181539222065,
+      "loss": 1.1159,
+      "step": 9737
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2580939496791897,
+      "learning_rate": 0.00011943629782759611,
+      "loss": 0.9785,
+      "step": 9738
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.29871072295327306,
+      "learning_rate": 0.00011942077977656601,
+      "loss": 1.072,
+      "step": 9739
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28134297529949587,
+      "learning_rate": 0.00011940526123951865,
+      "loss": 1.0097,
+      "step": 9740
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27030245416409626,
+      "learning_rate": 0.00011938974221684248,
+      "loss": 1.0362,
+      "step": 9741
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27891127006754407,
+      "learning_rate": 0.00011937422270892578,
+      "loss": 1.1503,
+      "step": 9742
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2527819203001018,
+      "learning_rate": 0.00011935870271615701,
+      "loss": 1.0682,
+      "step": 9743
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25930692949225115,
+      "learning_rate": 0.00011934318223892451,
+      "loss": 1.062,
+      "step": 9744
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2875000837745984,
+      "learning_rate": 0.00011932766127761675,
+      "loss": 0.9477,
+      "step": 9745
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.325246715534001,
+      "learning_rate": 0.00011931213983262211,
+      "loss": 1.0165,
+      "step": 9746
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.25662864976415845,
+      "learning_rate": 0.00011929661790432903,
+      "loss": 1.0786,
+      "step": 9747
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.260785274906786,
+      "learning_rate": 0.00011928109549312596,
+      "loss": 1.0819,
+      "step": 9748
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.24971660647552188,
+      "learning_rate": 0.00011926557259940137,
+      "loss": 1.1172,
+      "step": 9749
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2634144133215548,
+      "learning_rate": 0.00011925004922354368,
+      "loss": 1.0457,
+      "step": 9750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28481353084008365,
+      "learning_rate": 0.00011923452536594144,
+      "loss": 1.0667,
+      "step": 9751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2480420939966222,
+      "learning_rate": 0.00011921900102698312,
+      "loss": 1.0387,
+      "step": 9752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.290303015413711,
+      "learning_rate": 0.00011920347620705719,
+      "loss": 0.9624,
+      "step": 9753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.30582174621304614,
+      "learning_rate": 0.00011918795090655221,
+      "loss": 0.982,
+      "step": 9754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2655028218340249,
+      "learning_rate": 0.00011917242512585674,
+      "loss": 1.0536,
+      "step": 9755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27477226526624493,
+      "learning_rate": 0.00011915689886535923,
+      "loss": 0.9856,
+      "step": 9756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28682754344560973,
+      "learning_rate": 0.00011914137212544831,
+      "loss": 1.1707,
+      "step": 9757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26153050639845216,
+      "learning_rate": 0.00011912584490651253,
+      "loss": 1.0497,
+      "step": 9758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28229038067976947,
+      "learning_rate": 0.00011911031720894046,
+      "loss": 0.9965,
+      "step": 9759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28006029083616607,
+      "learning_rate": 0.00011909478903312066,
+      "loss": 1.0668,
+      "step": 9760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27740453645729457,
+      "learning_rate": 0.00011907926037944179,
+      "loss": 1.0366,
+      "step": 9761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2384257527120478,
+      "learning_rate": 0.00011906373124829244,
+      "loss": 1.0827,
+      "step": 9762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.293678354692542,
+      "learning_rate": 0.0001190482016400612,
+      "loss": 1.0645,
+      "step": 9763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2938241245866292,
+      "learning_rate": 0.00011903267155513677,
+      "loss": 1.055,
+      "step": 9764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26615653569909886,
+      "learning_rate": 0.00011901714099390777,
+      "loss": 1.0972,
+      "step": 9765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28171980966420157,
+      "learning_rate": 0.00011900160995676288,
+      "loss": 0.9422,
+      "step": 9766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.26740119824230185,
+      "learning_rate": 0.00011898607844409073,
+      "loss": 1.0326,
+      "step": 9767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.27023112187042453,
+      "learning_rate": 0.00011897054645628005,
+      "loss": 0.9777,
+      "step": 9768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3200885822581613,
+      "learning_rate": 0.00011895501399371953,
+      "loss": 1.0587,
+      "step": 9769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.31831634758810756,
+      "learning_rate": 0.00011893948105679787,
+      "loss": 1.0764,
+      "step": 9770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.3081776019474058,
+      "learning_rate": 0.00011892394764590378,
+      "loss": 1.0104,
+      "step": 9771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28957521753453463,
+      "learning_rate": 0.00011890841376142603,
+      "loss": 1.0592,
+      "step": 9772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28207411325497395,
+      "learning_rate": 0.00011889287940375334,
+      "loss": 1.104,
+      "step": 9773
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.36180324747623,
+      "learning_rate": 0.00011887734457327443,
+      "loss": 1.0525,
+      "step": 9774
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2910045433010645,
+      "learning_rate": 0.00011886180927037815,
+      "loss": 1.087,
+      "step": 9775
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28543697105123744,
+      "learning_rate": 0.00011884627349545323,
+      "loss": 1.0952,
+      "step": 9776
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2849455865427251,
+      "learning_rate": 0.00011883073724888844,
+      "loss": 1.0036,
+      "step": 9777
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3445968201656461,
+      "learning_rate": 0.00011881520053107267,
+      "loss": 1.0498,
+      "step": 9778
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2506321633298875,
+      "learning_rate": 0.00011879966334239466,
+      "loss": 1.1624,
+      "step": 9779
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29162103318787663,
+      "learning_rate": 0.00011878412568324322,
+      "loss": 1.0233,
+      "step": 9780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29500101419963837,
+      "learning_rate": 0.00011876858755400728,
+      "loss": 1.0922,
+      "step": 9781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3009884592875106,
+      "learning_rate": 0.00011875304895507562,
+      "loss": 1.1165,
+      "step": 9782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27565694136640945,
+      "learning_rate": 0.00011873750988683712,
+      "loss": 1.0688,
+      "step": 9783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3273050375332514,
+      "learning_rate": 0.00011872197034968067,
+      "loss": 1.1118,
+      "step": 9784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29960996186748606,
+      "learning_rate": 0.00011870643034399514,
+      "loss": 1.2125,
+      "step": 9785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2964779408152491,
+      "learning_rate": 0.00011869088987016943,
+      "loss": 1.1393,
+      "step": 9786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2858838476728597,
+      "learning_rate": 0.00011867534892859244,
+      "loss": 1.0267,
+      "step": 9787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2852488900654838,
+      "learning_rate": 0.00011865980751965313,
+      "loss": 1.1121,
+      "step": 9788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29119397522264384,
+      "learning_rate": 0.00011864426564374043,
+      "loss": 1.0825,
+      "step": 9789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2594508280784169,
+      "learning_rate": 0.00011862872330124324,
+      "loss": 1.154,
+      "step": 9790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2975677876895368,
+      "learning_rate": 0.00011861318049255052,
+      "loss": 1.0173,
+      "step": 9791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2648318009564742,
+      "learning_rate": 0.00011859763721805128,
+      "loss": 1.0818,
+      "step": 9792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2475611229418435,
+      "learning_rate": 0.00011858209347813449,
+      "loss": 0.9974,
+      "step": 9793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24368772705519215,
+      "learning_rate": 0.00011856654927318914,
+      "loss": 1.0076,
+      "step": 9794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2589222900696365,
+      "learning_rate": 0.0001185510046036042,
+      "loss": 1.0072,
+      "step": 9795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2518421468766372,
+      "learning_rate": 0.00011853545946976874,
+      "loss": 1.1606,
+      "step": 9796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.295292430219631,
+      "learning_rate": 0.00011851991387207171,
+      "loss": 0.98,
+      "step": 9797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29245516327817817,
+      "learning_rate": 0.00011850436781090223,
+      "loss": 1.0599,
+      "step": 9798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2914565208305866,
+      "learning_rate": 0.00011848882128664933,
+      "loss": 0.9911,
+      "step": 9799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3078689389856212,
+      "learning_rate": 0.00011847327429970203,
+      "loss": 0.9396,
+      "step": 9800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3127304428982239,
+      "learning_rate": 0.00011845772685044945,
+      "loss": 1.0645,
+      "step": 9801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2873883647576085,
+      "learning_rate": 0.00011844217893928064,
+      "loss": 0.9597,
+      "step": 9802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29524076369164753,
+      "learning_rate": 0.00011842663056658471,
+      "loss": 1.054,
+      "step": 9803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30216797407479606,
+      "learning_rate": 0.00011841108173275078,
+      "loss": 1.0366,
+      "step": 9804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2683352824632881,
+      "learning_rate": 0.00011839553243816794,
+      "loss": 1.1457,
+      "step": 9805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3088671649045264,
+      "learning_rate": 0.00011837998268322535,
+      "loss": 0.9811,
+      "step": 9806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3127798071067639,
+      "learning_rate": 0.00011836443246831215,
+      "loss": 1.0054,
+      "step": 9807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2958330910482311,
+      "learning_rate": 0.00011834888179381746,
+      "loss": 0.9717,
+      "step": 9808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.24632414148720827,
+      "learning_rate": 0.00011833333066013051,
+      "loss": 1.0858,
+      "step": 9809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2723921866158821,
+      "learning_rate": 0.00011831777906764044,
+      "loss": 1.0244,
+      "step": 9810
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2975337869324926,
+      "learning_rate": 0.00011830222701673639,
+      "loss": 1.0093,
+      "step": 9811
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28650604212889696,
+      "learning_rate": 0.00011828667450780764,
+      "loss": 1.069,
+      "step": 9812
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27777382822296665,
+      "learning_rate": 0.00011827112154124338,
+      "loss": 1.1207,
+      "step": 9813
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2870567816741802,
+      "learning_rate": 0.00011825556811743279,
+      "loss": 1.1142,
+      "step": 9814
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30386027251939157,
+      "learning_rate": 0.00011824001423676513,
+      "loss": 1.1055,
+      "step": 9815
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2705336512672914,
+      "learning_rate": 0.00011822445989962969,
+      "loss": 1.1277,
+      "step": 9816
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29969456459847577,
+      "learning_rate": 0.00011820890510641567,
+      "loss": 1.0124,
+      "step": 9817
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30298340488413317,
+      "learning_rate": 0.00011819334985751233,
+      "loss": 1.0302,
+      "step": 9818
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27778928676842507,
+      "learning_rate": 0.00011817779415330901,
+      "loss": 1.0885,
+      "step": 9819
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.300820511952422,
+      "learning_rate": 0.00011816223799419497,
+      "loss": 1.0825,
+      "step": 9820
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3444116599675863,
+      "learning_rate": 0.00011814668138055947,
+      "loss": 1.098,
+      "step": 9821
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27136608383916344,
+      "learning_rate": 0.0001181311243127919,
+      "loss": 1.1083,
+      "step": 9822
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29143092481382066,
+      "learning_rate": 0.00011811556679128153,
+      "loss": 1.0637,
+      "step": 9823
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25973448401962235,
+      "learning_rate": 0.00011810000881641771,
+      "loss": 1.0963,
+      "step": 9824
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2774775599714097,
+      "learning_rate": 0.00011808445038858982,
+      "loss": 1.0984,
+      "step": 9825
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.263532587397457,
+      "learning_rate": 0.00011806889150818716,
+      "loss": 0.9681,
+      "step": 9826
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3141288079300534,
+      "learning_rate": 0.00011805333217559918,
+      "loss": 1.0527,
+      "step": 9827
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31147150343587526,
+      "learning_rate": 0.00011803777239121516,
+      "loss": 1.0198,
+      "step": 9828
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2732581866158368,
+      "learning_rate": 0.00011802221215542459,
+      "loss": 1.0903,
+      "step": 9829
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28505766654611703,
+      "learning_rate": 0.00011800665146861683,
+      "loss": 1.0171,
+      "step": 9830
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28825531298858104,
+      "learning_rate": 0.00011799109033118127,
+      "loss": 1.1042,
+      "step": 9831
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28681592008334156,
+      "learning_rate": 0.00011797552874350739,
+      "loss": 1.0476,
+      "step": 9832
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3123668239526817,
+      "learning_rate": 0.00011795996670598462,
+      "loss": 1.0743,
+      "step": 9833
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2418704079745117,
+      "learning_rate": 0.0001179444042190024,
+      "loss": 1.0915,
+      "step": 9834
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3233001919313393,
+      "learning_rate": 0.00011792884128295014,
+      "loss": 1.1112,
+      "step": 9835
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.23543282622618317,
+      "learning_rate": 0.0001179132778982174,
+      "loss": 0.9408,
+      "step": 9836
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27953787864369156,
+      "learning_rate": 0.00011789771406519361,
+      "loss": 1.0951,
+      "step": 9837
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.26888014225579154,
+      "learning_rate": 0.00011788214978426827,
+      "loss": 1.0421,
+      "step": 9838
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28757566533635587,
+      "learning_rate": 0.0001178665850558309,
+      "loss": 1.1287,
+      "step": 9839
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3053049659715436,
+      "learning_rate": 0.00011785101988027103,
+      "loss": 1.0709,
+      "step": 9840
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2693832593847435,
+      "learning_rate": 0.00011783545425797813,
+      "loss": 1.178,
+      "step": 9841
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29150202296705147,
+      "learning_rate": 0.0001178198881893418,
+      "loss": 1.1276,
+      "step": 9842
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28026413298185754,
+      "learning_rate": 0.00011780432167475157,
+      "loss": 1.1026,
+      "step": 9843
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2883237105659363,
+      "learning_rate": 0.00011778875471459703,
+      "loss": 1.029,
+      "step": 9844
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2882413609766784,
+      "learning_rate": 0.00011777318730926768,
+      "loss": 1.1348,
+      "step": 9845
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3024147348726725,
+      "learning_rate": 0.00011775761945915315,
+      "loss": 1.1326,
+      "step": 9846
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27230068738152696,
+      "learning_rate": 0.00011774205116464304,
+      "loss": 1.1274,
+      "step": 9847
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2671829447086581,
+      "learning_rate": 0.00011772648242612694,
+      "loss": 0.999,
+      "step": 9848
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.27183496557565345,
+      "learning_rate": 0.00011771091324399447,
+      "loss": 0.9818,
+      "step": 9849
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30565783107867467,
+      "learning_rate": 0.0001176953436186353,
+      "loss": 1.0334,
+      "step": 9850
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2672952617830373,
+      "learning_rate": 0.00011767977355043902,
+      "loss": 1.014,
+      "step": 9851
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.25977834030458224,
+      "learning_rate": 0.00011766420303979528,
+      "loss": 1.0847,
+      "step": 9852
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29226617360753704,
+      "learning_rate": 0.00011764863208709378,
+      "loss": 1.1198,
+      "step": 9853
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2710445438667288,
+      "learning_rate": 0.00011763306069272415,
+      "loss": 1.1502,
+      "step": 9854
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2683767291753003,
+      "learning_rate": 0.00011761748885707611,
+      "loss": 1.0218,
+      "step": 9855
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30996151508397074,
+      "learning_rate": 0.00011760191658053933,
+      "loss": 1.0196,
+      "step": 9856
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2836262926909511,
+      "learning_rate": 0.00011758634386350353,
+      "loss": 0.9859,
+      "step": 9857
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2835130303902405,
+      "learning_rate": 0.00011757077070635842,
+      "loss": 0.9518,
+      "step": 9858
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29179997921864015,
+      "learning_rate": 0.00011755519710949375,
+      "loss": 1.0851,
+      "step": 9859
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2580784420137259,
+      "learning_rate": 0.0001175396230732992,
+      "loss": 1.0792,
+      "step": 9860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2980701381253143,
+      "learning_rate": 0.00011752404859816459,
+      "loss": 1.1523,
+      "step": 9861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30650598237537346,
+      "learning_rate": 0.00011750847368447963,
+      "loss": 1.0509,
+      "step": 9862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2829225311103543,
+      "learning_rate": 0.00011749289833263413,
+      "loss": 1.0698,
+      "step": 9863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3150147271464283,
+      "learning_rate": 0.00011747732254301786,
+      "loss": 1.0618,
+      "step": 9864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.26617589803804353,
+      "learning_rate": 0.00011746174631602059,
+      "loss": 0.9886,
+      "step": 9865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28821178558985444,
+      "learning_rate": 0.00011744616965203214,
+      "loss": 0.9826,
+      "step": 9866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.264781657352801,
+      "learning_rate": 0.00011743059255144233,
+      "loss": 0.8746,
+      "step": 9867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.29886418377952234,
+      "learning_rate": 0.000117415015014641,
+      "loss": 1.0563,
+      "step": 9868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3002368289446441,
+      "learning_rate": 0.00011739943704201796,
+      "loss": 1.0341,
+      "step": 9869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28512524718120663,
+      "learning_rate": 0.00011738385863396311,
+      "loss": 0.9591,
+      "step": 9870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2900458256329998,
+      "learning_rate": 0.00011736827979086625,
+      "loss": 1.0893,
+      "step": 9871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28440813574018564,
+      "learning_rate": 0.00011735270051311724,
+      "loss": 1.052,
+      "step": 9872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2469565537916454,
+      "learning_rate": 0.00011733712080110603,
+      "loss": 0.857,
+      "step": 9873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2501421742546667,
+      "learning_rate": 0.00011732154065522247,
+      "loss": 1.1517,
+      "step": 9874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2943711294105304,
+      "learning_rate": 0.00011730596007585646,
+      "loss": 1.0131,
+      "step": 9875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2759019579633036,
+      "learning_rate": 0.00011729037906339795,
+      "loss": 1.0827,
+      "step": 9876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2734166501679395,
+      "learning_rate": 0.00011727479761823683,
+      "loss": 1.017,
+      "step": 9877
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2945759953073775,
+      "learning_rate": 0.00011725921574076305,
+      "loss": 1.048,
+      "step": 9878
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2847423179857022,
+      "learning_rate": 0.00011724363343136651,
+      "loss": 0.9387,
+      "step": 9879
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2731998133242859,
+      "learning_rate": 0.00011722805069043724,
+      "loss": 1.1269,
+      "step": 9880
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31097755473044353,
+      "learning_rate": 0.00011721246751836514,
+      "loss": 1.1183,
+      "step": 9881
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24077282326706342,
+      "learning_rate": 0.00011719688391554024,
+      "loss": 1.1315,
+      "step": 9882
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28290762006495423,
+      "learning_rate": 0.00011718129988235251,
+      "loss": 1.0301,
+      "step": 9883
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28799605220582924,
+      "learning_rate": 0.00011716571541919197,
+      "loss": 1.1275,
+      "step": 9884
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2506151300260542,
+      "learning_rate": 0.00011715013052644859,
+      "loss": 1.0488,
+      "step": 9885
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3003785975679057,
+      "learning_rate": 0.00011713454520451243,
+      "loss": 1.1445,
+      "step": 9886
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2871681991986448,
+      "learning_rate": 0.00011711895945377351,
+      "loss": 1.1341,
+      "step": 9887
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.263469877916956,
+      "learning_rate": 0.00011710337327462186,
+      "loss": 1.0684,
+      "step": 9888
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25031317343613296,
+      "learning_rate": 0.00011708778666744756,
+      "loss": 1.0692,
+      "step": 9889
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27978745157205426,
+      "learning_rate": 0.00011707219963264063,
+      "loss": 0.9031,
+      "step": 9890
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2771417917801048,
+      "learning_rate": 0.00011705661217059121,
+      "loss": 1.0672,
+      "step": 9891
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27046659772554993,
+      "learning_rate": 0.00011704102428168931,
+      "loss": 1.1261,
+      "step": 9892
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2627203026681411,
+      "learning_rate": 0.00011702543596632512,
+      "loss": 1.0424,
+      "step": 9893
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24842638162917163,
+      "learning_rate": 0.00011700984722488865,
+      "loss": 1.0639,
+      "step": 9894
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.297596295779575,
+      "learning_rate": 0.00011699425805777008,
+      "loss": 1.1003,
+      "step": 9895
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27759923733902636,
+      "learning_rate": 0.00011697866846535953,
+      "loss": 1.1192,
+      "step": 9896
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30285981266810275,
+      "learning_rate": 0.00011696307844804713,
+      "loss": 1.1918,
+      "step": 9897
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31450365765502614,
+      "learning_rate": 0.00011694748800622301,
+      "loss": 1.0818,
+      "step": 9898
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3141057225598429,
+      "learning_rate": 0.00011693189714027737,
+      "loss": 0.9055,
+      "step": 9899
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2718320654630044,
+      "learning_rate": 0.00011691630585060036,
+      "loss": 1.182,
+      "step": 9900
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2642413261096164,
+      "learning_rate": 0.00011690071413758217,
+      "loss": 1.0622,
+      "step": 9901
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2781872909054998,
+      "learning_rate": 0.00011688512200161297,
+      "loss": 1.0917,
+      "step": 9902
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2933665399467418,
+      "learning_rate": 0.00011686952944308298,
+      "loss": 1.0172,
+      "step": 9903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2986798343224596,
+      "learning_rate": 0.00011685393646238243,
+      "loss": 1.0515,
+      "step": 9904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27008347335950034,
+      "learning_rate": 0.00011683834305990154,
+      "loss": 1.0519,
+      "step": 9905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27146126457125797,
+      "learning_rate": 0.00011682274923603049,
+      "loss": 1.0363,
+      "step": 9906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24730678500742276,
+      "learning_rate": 0.00011680715499115959,
+      "loss": 1.0349,
+      "step": 9907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.274923895199804,
+      "learning_rate": 0.00011679156032567911,
+      "loss": 1.0628,
+      "step": 9908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28568564007828456,
+      "learning_rate": 0.00011677596523997922,
+      "loss": 1.0989,
+      "step": 9909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26786733880651786,
+      "learning_rate": 0.00011676036973445028,
+      "loss": 0.9306,
+      "step": 9910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30754840140433903,
+      "learning_rate": 0.00011674477380948255,
+      "loss": 1.0696,
+      "step": 9911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25852789773425133,
+      "learning_rate": 0.00011672917746546634,
+      "loss": 0.8838,
+      "step": 9912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2558247705235392,
+      "learning_rate": 0.00011671358070279193,
+      "loss": 0.99,
+      "step": 9913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29138329913400457,
+      "learning_rate": 0.00011669798352184968,
+      "loss": 0.9954,
+      "step": 9914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30260586220267804,
+      "learning_rate": 0.0001166823859230299,
+      "loss": 1.0312,
+      "step": 9915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2827165548781267,
+      "learning_rate": 0.0001166667879067229,
+      "loss": 1.131,
+      "step": 9916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29392517958616593,
+      "learning_rate": 0.0001166511894733191,
+      "loss": 1.0896,
+      "step": 9917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3260991950351522,
+      "learning_rate": 0.00011663559062320878,
+      "loss": 1.0871,
+      "step": 9918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2561723519059954,
+      "learning_rate": 0.00011661999135678237,
+      "loss": 1.1483,
+      "step": 9919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.309722846552505,
+      "learning_rate": 0.00011660439167443022,
+      "loss": 1.0889,
+      "step": 9920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25190002272893053,
+      "learning_rate": 0.00011658879157654276,
+      "loss": 1.0626,
+      "step": 9921
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3259479008437906,
+      "learning_rate": 0.00011657319106351035,
+      "loss": 1.0441,
+      "step": 9922
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26095427143336863,
+      "learning_rate": 0.0001165575901357234,
+      "loss": 1.108,
+      "step": 9923
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3334009293399356,
+      "learning_rate": 0.00011654198879357236,
+      "loss": 0.9941,
+      "step": 9924
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2479075214464415,
+      "learning_rate": 0.00011652638703744769,
+      "loss": 0.9997,
+      "step": 9925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2699529303379684,
+      "learning_rate": 0.00011651078486773974,
+      "loss": 1.0008,
+      "step": 9926
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2793875562565832,
+      "learning_rate": 0.00011649518228483907,
+      "loss": 1.1141,
+      "step": 9927
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30640720124070897,
+      "learning_rate": 0.00011647957928913606,
+      "loss": 1.088,
+      "step": 9928
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29450119862900453,
+      "learning_rate": 0.00011646397588102123,
+      "loss": 1.0208,
+      "step": 9929
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2986882197988017,
+      "learning_rate": 0.00011644837206088508,
+      "loss": 1.1405,
+      "step": 9930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28289114316471997,
+      "learning_rate": 0.00011643276782911805,
+      "loss": 1.0624,
+      "step": 9931
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25290685088339804,
+      "learning_rate": 0.0001164171631861107,
+      "loss": 0.9819,
+      "step": 9932
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29613853648276806,
+      "learning_rate": 0.00011640155813225348,
+      "loss": 0.9768,
+      "step": 9933
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30470164680746314,
+      "learning_rate": 0.00011638595266793701,
+      "loss": 1.0121,
+      "step": 9934
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26617140467629113,
+      "learning_rate": 0.00011637034679355176,
+      "loss": 1.0911,
+      "step": 9935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29756347865505667,
+      "learning_rate": 0.00011635474050948829,
+      "loss": 1.0299,
+      "step": 9936
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2496000722355499,
+      "learning_rate": 0.00011633913381613717,
+      "loss": 0.9958,
+      "step": 9937
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28395690278817887,
+      "learning_rate": 0.00011632352671388898,
+      "loss": 1.0131,
+      "step": 9938
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2922841640955166,
+      "learning_rate": 0.00011630791920313425,
+      "loss": 1.0615,
+      "step": 9939
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3033319129166104,
+      "learning_rate": 0.00011629231128426356,
+      "loss": 1.1185,
+      "step": 9940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2696093697556044,
+      "learning_rate": 0.00011627670295766759,
+      "loss": 0.9108,
+      "step": 9941
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.275903786765533,
+      "learning_rate": 0.00011626109422373688,
+      "loss": 0.9584,
+      "step": 9942
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2593587234240088,
+      "learning_rate": 0.00011624548508286206,
+      "loss": 1.0215,
+      "step": 9943
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.25466792159457646,
+      "learning_rate": 0.00011622987553543376,
+      "loss": 1.0309,
+      "step": 9944
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28160232777966615,
+      "learning_rate": 0.00011621426558184265,
+      "loss": 1.036,
+      "step": 9945
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2665942842206432,
+      "learning_rate": 0.00011619865522247933,
+      "loss": 1.0767,
+      "step": 9946
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29079511088912224,
+      "learning_rate": 0.00011618304445773451,
+      "loss": 1.0276,
+      "step": 9947
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2803873259698023,
+      "learning_rate": 0.00011616743328799881,
+      "loss": 1.0198,
+      "step": 9948
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2555819038764869,
+      "learning_rate": 0.00011615182171366297,
+      "loss": 1.0222,
+      "step": 9949
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31734739800334083,
+      "learning_rate": 0.00011613620973511758,
+      "loss": 1.1547,
+      "step": 9950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28917459887517855,
+      "learning_rate": 0.00011612059735275342,
+      "loss": 1.1102,
+      "step": 9951
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.24841835752111638,
+      "learning_rate": 0.00011610498456696119,
+      "loss": 1.1191,
+      "step": 9952
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29654024864058454,
+      "learning_rate": 0.00011608937137813161,
+      "loss": 0.9583,
+      "step": 9953
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3147918475594126,
+      "learning_rate": 0.00011607375778665536,
+      "loss": 1.1377,
+      "step": 9954
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2772820949372226,
+      "learning_rate": 0.00011605814379292325,
+      "loss": 1.0474,
+      "step": 9955
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30109169932300395,
+      "learning_rate": 0.00011604252939732601,
+      "loss": 1.0939,
+      "step": 9956
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2868717018893174,
+      "learning_rate": 0.00011602691460025437,
+      "loss": 0.9908,
+      "step": 9957
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2675493227671909,
+      "learning_rate": 0.00011601129940209911,
+      "loss": 1.0693,
+      "step": 9958
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31375505859098296,
+      "learning_rate": 0.00011599568380325106,
+      "loss": 1.1187,
+      "step": 9959
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.295148232115398,
+      "learning_rate": 0.00011598006780410091,
+      "loss": 1.0427,
+      "step": 9960
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28936388619727343,
+      "learning_rate": 0.00011596445140503957,
+      "loss": 0.9698,
+      "step": 9961
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28216120171352943,
+      "learning_rate": 0.0001159488346064578,
+      "loss": 1.0421,
+      "step": 9962
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28268688737805364,
+      "learning_rate": 0.00011593321740874639,
+      "loss": 1.0307,
+      "step": 9963
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27612823655418073,
+      "learning_rate": 0.00011591759981229622,
+      "loss": 0.9663,
+      "step": 9964
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2740428689355272,
+      "learning_rate": 0.00011590198181749811,
+      "loss": 1.0961,
+      "step": 9965
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2499724504263894,
+      "learning_rate": 0.0001158863634247429,
+      "loss": 1.0651,
+      "step": 9966
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2757709129720417,
+      "learning_rate": 0.00011587074463442147,
+      "loss": 1.0643,
+      "step": 9967
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2985207172650015,
+      "learning_rate": 0.00011585512544692467,
+      "loss": 1.1078,
+      "step": 9968
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3045737397007342,
+      "learning_rate": 0.00011583950586264343,
+      "loss": 1.1594,
+      "step": 9969
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27637988509985273,
+      "learning_rate": 0.00011582388588196855,
+      "loss": 1.0231,
+      "step": 9970
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2821414694181937,
+      "learning_rate": 0.000115808265505291,
+      "loss": 1.1147,
+      "step": 9971
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29155806162009856,
+      "learning_rate": 0.00011579264473300167,
+      "loss": 1.0431,
+      "step": 9972
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27636257246936596,
+      "learning_rate": 0.00011577702356549149,
+      "loss": 1.0565,
+      "step": 9973
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31102867461907796,
+      "learning_rate": 0.00011576140200315135,
+      "loss": 1.1162,
+      "step": 9974
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.28686875083828667,
+      "learning_rate": 0.00011574578004637226,
+      "loss": 0.9938,
+      "step": 9975
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2829922774553789,
+      "learning_rate": 0.00011573015769554512,
+      "loss": 1.043,
+      "step": 9976
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30378521566991484,
+      "learning_rate": 0.00011571453495106086,
+      "loss": 1.125,
+      "step": 9977
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2882399502995974,
+      "learning_rate": 0.00011569891181331054,
+      "loss": 1.0316,
+      "step": 9978
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2546929646781789,
+      "learning_rate": 0.00011568328828268506,
+      "loss": 1.0347,
+      "step": 9979
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.27885757954603946,
+      "learning_rate": 0.00011566766435957541,
+      "loss": 1.0919,
+      "step": 9980
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.26559169992251663,
+      "learning_rate": 0.00011565204004437267,
+      "loss": 1.0969,
+      "step": 9981
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.33065813751668216,
+      "learning_rate": 0.00011563641533746774,
+      "loss": 1.0411,
+      "step": 9982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.33089614226385505,
+      "learning_rate": 0.00011562079023925172,
+      "loss": 1.0129,
+      "step": 9983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29760832729104647,
+      "learning_rate": 0.00011560516475011558,
+      "loss": 1.0865,
+      "step": 9984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2982860380062189,
+      "learning_rate": 0.00011558953887045041,
+      "loss": 1.0879,
+      "step": 9985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3089962049085526,
+      "learning_rate": 0.00011557391260064723,
+      "loss": 1.062,
+      "step": 9986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2907036025220388,
+      "learning_rate": 0.00011555828594109707,
+      "loss": 1.0863,
+      "step": 9987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31013815730431993,
+      "learning_rate": 0.00011554265889219106,
+      "loss": 1.0049,
+      "step": 9988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29883777425014585,
+      "learning_rate": 0.00011552703145432025,
+      "loss": 1.0613,
+      "step": 9989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29975907234168436,
+      "learning_rate": 0.0001155114036278757,
+      "loss": 1.0005,
+      "step": 9990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30684771167447394,
+      "learning_rate": 0.0001154957754132485,
+      "loss": 1.1158,
+      "step": 9991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26335079161415853,
+      "learning_rate": 0.00011548014681082981,
+      "loss": 1.021,
+      "step": 9992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27648270125908575,
+      "learning_rate": 0.00011546451782101071,
+      "loss": 1.1223,
+      "step": 9993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3475752040212275,
+      "learning_rate": 0.00011544888844418233,
+      "loss": 1.1233,
+      "step": 9994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30111765274598085,
+      "learning_rate": 0.0001154332586807358,
+      "loss": 1.0963,
+      "step": 9995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2970605582481724,
+      "learning_rate": 0.0001154176285310623,
+      "loss": 0.99,
+      "step": 9996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.305235597238056,
+      "learning_rate": 0.00011540199799555294,
+      "loss": 1.132,
+      "step": 9997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2670713296870493,
+      "learning_rate": 0.00011538636707459889,
+      "loss": 1.0476,
+      "step": 9998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2677367396231706,
+      "learning_rate": 0.00011537073576859136,
+      "loss": 1.069,
+      "step": 9999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2525746570305202,
+      "learning_rate": 0.00011535510407792149,
+      "loss": 0.9385,
+      "step": 10000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2758414503350628,
+      "learning_rate": 0.0001153394720029805,
+      "loss": 1.0631,
+      "step": 10001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3047038095467878,
+      "learning_rate": 0.00011532383954415957,
+      "loss": 1.0311,
+      "step": 10002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28111478399862794,
+      "learning_rate": 0.00011530820670184995,
+      "loss": 1.0573,
+      "step": 10003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28036170451809095,
+      "learning_rate": 0.0001152925734764428,
+      "loss": 1.1742,
+      "step": 10004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2963156258676252,
+      "learning_rate": 0.00011527693986832942,
+      "loss": 1.016,
+      "step": 10005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2815524009206448,
+      "learning_rate": 0.000115261305877901,
+      "loss": 1.0306,
+      "step": 10006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29474398852141115,
+      "learning_rate": 0.00011524567150554881,
+      "loss": 1.0829,
+      "step": 10007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29358155011217263,
+      "learning_rate": 0.00011523003675166411,
+      "loss": 1.0714,
+      "step": 10008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2844867269668011,
+      "learning_rate": 0.00011521440161663819,
+      "loss": 1.0348,
+      "step": 10009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.255280580277318,
+      "learning_rate": 0.00011519876610086229,
+      "loss": 1.1012,
+      "step": 10010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31413925882198807,
+      "learning_rate": 0.00011518313020472768,
+      "loss": 1.0744,
+      "step": 10011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31994016396189484,
+      "learning_rate": 0.00011516749392862576,
+      "loss": 0.9883,
+      "step": 10012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27390552772942023,
+      "learning_rate": 0.00011515185727294771,
+      "loss": 0.9983,
+      "step": 10013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2853534170678965,
+      "learning_rate": 0.00011513622023808495,
+      "loss": 1.0904,
+      "step": 10014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25465697705481827,
+      "learning_rate": 0.00011512058282442874,
+      "loss": 0.922,
+      "step": 10015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26802643575831,
+      "learning_rate": 0.00011510494503237046,
+      "loss": 1.0313,
+      "step": 10016
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3193729552335535,
+      "learning_rate": 0.00011508930686230146,
+      "loss": 0.9854,
+      "step": 10017
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2879367749732901,
+      "learning_rate": 0.00011507366831461302,
+      "loss": 1.1051,
+      "step": 10018
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30097969407853326,
+      "learning_rate": 0.0001150580293896966,
+      "loss": 1.0425,
+      "step": 10019
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30624544517915264,
+      "learning_rate": 0.0001150423900879435,
+      "loss": 1.0311,
+      "step": 10020
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2767941099814116,
+      "learning_rate": 0.00011502675040974516,
+      "loss": 0.9427,
+      "step": 10021
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28490801229190077,
+      "learning_rate": 0.00011501111035549295,
+      "loss": 1.1947,
+      "step": 10022
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2784277522206461,
+      "learning_rate": 0.00011499546992557826,
+      "loss": 0.9624,
+      "step": 10023
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2666766109775799,
+      "learning_rate": 0.00011497982912039249,
+      "loss": 0.9757,
+      "step": 10024
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32891260560151236,
+      "learning_rate": 0.00011496418794032711,
+      "loss": 1.1256,
+      "step": 10025
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3086769971151652,
+      "learning_rate": 0.0001149485463857735,
+      "loss": 0.9429,
+      "step": 10026
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25897740279910364,
+      "learning_rate": 0.00011493290445712315,
+      "loss": 0.8941,
+      "step": 10027
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27367523231686985,
+      "learning_rate": 0.00011491726215476746,
+      "loss": 0.961,
+      "step": 10028
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2902717009917645,
+      "learning_rate": 0.0001149016194790979,
+      "loss": 1.0828,
+      "step": 10029
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2767681614076228,
+      "learning_rate": 0.00011488597643050598,
+      "loss": 1.1453,
+      "step": 10030
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2857144403267616,
+      "learning_rate": 0.0001148703330093831,
+      "loss": 1.1737,
+      "step": 10031
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28150370452068746,
+      "learning_rate": 0.00011485468921612084,
+      "loss": 1.1734,
+      "step": 10032
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3290209532225155,
+      "learning_rate": 0.00011483904505111063,
+      "loss": 1.1331,
+      "step": 10033
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3138396953516879,
+      "learning_rate": 0.00011482340051474396,
+      "loss": 1.0148,
+      "step": 10034
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30110096644908,
+      "learning_rate": 0.00011480775560741239,
+      "loss": 1.0134,
+      "step": 10035
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3203655127160484,
+      "learning_rate": 0.00011479211032950743,
+      "loss": 1.0475,
+      "step": 10036
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2772142324547969,
+      "learning_rate": 0.00011477646468142062,
+      "loss": 1.0685,
+      "step": 10037
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2917042525821615,
+      "learning_rate": 0.0001147608186635435,
+      "loss": 1.1421,
+      "step": 10038
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26698256397886955,
+      "learning_rate": 0.00011474517227626762,
+      "loss": 1.1066,
+      "step": 10039
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2675194389510324,
+      "learning_rate": 0.00011472952551998452,
+      "loss": 0.9672,
+      "step": 10040
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2925787442554594,
+      "learning_rate": 0.0001147138783950858,
+      "loss": 1.0619,
+      "step": 10041
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2925652633706256,
+      "learning_rate": 0.00011469823090196303,
+      "loss": 1.067,
+      "step": 10042
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26146199790712404,
+      "learning_rate": 0.00011468258304100779,
+      "loss": 1.0669,
+      "step": 10043
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29387442287441945,
+      "learning_rate": 0.00011466693481261168,
+      "loss": 1.1529,
+      "step": 10044
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30566339364004746,
+      "learning_rate": 0.0001146512862171663,
+      "loss": 1.0711,
+      "step": 10045
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2623345882172563,
+      "learning_rate": 0.00011463563725506328,
+      "loss": 1.1218,
+      "step": 10046
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3090594432747929,
+      "learning_rate": 0.00011461998792669426,
+      "loss": 1.0833,
+      "step": 10047
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3168446002514091,
+      "learning_rate": 0.0001146043382324508,
+      "loss": 1.0041,
+      "step": 10048
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26118984316659555,
+      "learning_rate": 0.00011458868817272465,
+      "loss": 1.075,
+      "step": 10049
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3084194238337441,
+      "learning_rate": 0.0001145730377479074,
+      "loss": 0.9865,
+      "step": 10050
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2916812390686762,
+      "learning_rate": 0.00011455738695839071,
+      "loss": 1.1019,
+      "step": 10051
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2853685064498807,
+      "learning_rate": 0.00011454173580456627,
+      "loss": 1.1053,
+      "step": 10052
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28381729559022867,
+      "learning_rate": 0.00011452608428682574,
+      "loss": 1.0872,
+      "step": 10053
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.35555285136726095,
+      "learning_rate": 0.0001145104324055608,
+      "loss": 1.0321,
+      "step": 10054
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29725262158547283,
+      "learning_rate": 0.00011449478016116322,
+      "loss": 1.0607,
+      "step": 10055
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.32988647361187395,
+      "learning_rate": 0.00011447912755402463,
+      "loss": 1.0194,
+      "step": 10056
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2986055841309832,
+      "learning_rate": 0.00011446347458453677,
+      "loss": 1.0673,
+      "step": 10057
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2696942404700563,
+      "learning_rate": 0.00011444782125309137,
+      "loss": 1.0442,
+      "step": 10058
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3064172789572965,
+      "learning_rate": 0.00011443216756008017,
+      "loss": 0.9705,
+      "step": 10059
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29986340116379595,
+      "learning_rate": 0.00011441651350589493,
+      "loss": 1.042,
+      "step": 10060
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29196996307249695,
+      "learning_rate": 0.00011440085909092735,
+      "loss": 1.1514,
+      "step": 10061
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26305531388271913,
+      "learning_rate": 0.00011438520431556923,
+      "loss": 1.0125,
+      "step": 10062
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28626327759234815,
+      "learning_rate": 0.00011436954918021232,
+      "loss": 1.1407,
+      "step": 10063
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27422258071461836,
+      "learning_rate": 0.00011435389368524842,
+      "loss": 1.055,
+      "step": 10064
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29139299897756993,
+      "learning_rate": 0.0001143382378310693,
+      "loss": 1.015,
+      "step": 10065
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2590303991609533,
+      "learning_rate": 0.0001143225816180668,
+      "loss": 0.9926,
+      "step": 10066
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28910079655703913,
+      "learning_rate": 0.00011430692504663265,
+      "loss": 0.9727,
+      "step": 10067
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3312654664415843,
+      "learning_rate": 0.00011429126811715872,
+      "loss": 1.101,
+      "step": 10068
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30000869026615357,
+      "learning_rate": 0.00011427561083003683,
+      "loss": 1.0738,
+      "step": 10069
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28159858071588617,
+      "learning_rate": 0.00011425995318565883,
+      "loss": 1.0995,
+      "step": 10070
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3123705444016819,
+      "learning_rate": 0.00011424429518441653,
+      "loss": 0.9907,
+      "step": 10071
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.25560829274875024,
+      "learning_rate": 0.00011422863682670176,
+      "loss": 1.0387,
+      "step": 10072
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3047604879614061,
+      "learning_rate": 0.00011421297811290643,
+      "loss": 1.0803,
+      "step": 10073
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27464921456414265,
+      "learning_rate": 0.0001141973190434224,
+      "loss": 1.1457,
+      "step": 10074
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3015661166736289,
+      "learning_rate": 0.00011418165961864151,
+      "loss": 0.9435,
+      "step": 10075
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3288482467287445,
+      "learning_rate": 0.0001141659998389557,
+      "loss": 1.0268,
+      "step": 10076
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2880879802413768,
+      "learning_rate": 0.00011415033970475682,
+      "loss": 1.1155,
+      "step": 10077
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3046710592105044,
+      "learning_rate": 0.00011413467921643681,
+      "loss": 1.0854,
+      "step": 10078
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.23189332771051346,
+      "learning_rate": 0.00011411901837438757,
+      "loss": 0.8959,
+      "step": 10079
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2860860545736034,
+      "learning_rate": 0.00011410335717900102,
+      "loss": 1.004,
+      "step": 10080
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29819011560467884,
+      "learning_rate": 0.0001140876956306691,
+      "loss": 1.0271,
+      "step": 10081
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2961548882133075,
+      "learning_rate": 0.00011407203372978372,
+      "loss": 1.0964,
+      "step": 10082
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.27643172524881576,
+      "learning_rate": 0.00011405637147673688,
+      "loss": 1.0848,
+      "step": 10083
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.26601937387043634,
+      "learning_rate": 0.00011404070887192051,
+      "loss": 1.0771,
+      "step": 10084
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.29423767837815973,
+      "learning_rate": 0.00011402504591572656,
+      "loss": 1.1087,
+      "step": 10085
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31957243246751704,
+      "learning_rate": 0.00011400938260854703,
+      "loss": 1.1154,
+      "step": 10086
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34871076842626053,
+      "learning_rate": 0.00011399371895077389,
+      "loss": 1.0691,
+      "step": 10087
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30378790346074774,
+      "learning_rate": 0.00011397805494279916,
+      "loss": 1.096,
+      "step": 10088
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25772589878682645,
+      "learning_rate": 0.00011396239058501476,
+      "loss": 1.0342,
+      "step": 10089
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2823080995120186,
+      "learning_rate": 0.00011394672587781284,
+      "loss": 1.1017,
+      "step": 10090
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.312629120636968,
+      "learning_rate": 0.0001139310608215853,
+      "loss": 1.0579,
+      "step": 10091
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3054424461560633,
+      "learning_rate": 0.00011391539541672418,
+      "loss": 1.0553,
+      "step": 10092
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28281060096224714,
+      "learning_rate": 0.00011389972966362159,
+      "loss": 1.0612,
+      "step": 10093
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27166297516714194,
+      "learning_rate": 0.00011388406356266951,
+      "loss": 1.0524,
+      "step": 10094
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31743027731547635,
+      "learning_rate": 0.00011386839711426003,
+      "loss": 1.0024,
+      "step": 10095
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32534625159202174,
+      "learning_rate": 0.00011385273031878516,
+      "loss": 1.0885,
+      "step": 10096
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30049574798069184,
+      "learning_rate": 0.00011383706317663705,
+      "loss": 1.0135,
+      "step": 10097
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3266379229812977,
+      "learning_rate": 0.00011382139568820771,
+      "loss": 1.0434,
+      "step": 10098
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3039958500677989,
+      "learning_rate": 0.00011380572785388923,
+      "loss": 1.1338,
+      "step": 10099
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2820873927102382,
+      "learning_rate": 0.0001137900596740738,
+      "loss": 0.9901,
+      "step": 10100
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25895906852181655,
+      "learning_rate": 0.00011377439114915343,
+      "loss": 1.0197,
+      "step": 10101
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2736675525144609,
+      "learning_rate": 0.00011375872227952024,
+      "loss": 1.0666,
+      "step": 10102
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27812994654877704,
+      "learning_rate": 0.00011374305306556641,
+      "loss": 0.9667,
+      "step": 10103
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24944483327099637,
+      "learning_rate": 0.00011372738350768404,
+      "loss": 1.1377,
+      "step": 10104
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3052298238193706,
+      "learning_rate": 0.00011371171360626528,
+      "loss": 1.0432,
+      "step": 10105
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27306031709563633,
+      "learning_rate": 0.00011369604336170221,
+      "loss": 1.0682,
+      "step": 10106
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3167325272734246,
+      "learning_rate": 0.0001136803727743871,
+      "loss": 1.1381,
+      "step": 10107
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3310696902768376,
+      "learning_rate": 0.00011366470184471206,
+      "loss": 0.9888,
+      "step": 10108
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2920470605038157,
+      "learning_rate": 0.00011364903057306923,
+      "loss": 1.0723,
+      "step": 10109
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28721162596130884,
+      "learning_rate": 0.00011363335895985087,
+      "loss": 1.0795,
+      "step": 10110
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3123155431147055,
+      "learning_rate": 0.00011361768700544915,
+      "loss": 1.0195,
+      "step": 10111
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31432924829965664,
+      "learning_rate": 0.00011360201471025625,
+      "loss": 1.0262,
+      "step": 10112
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30371743385151373,
+      "learning_rate": 0.00011358634207466434,
+      "loss": 1.1198,
+      "step": 10113
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30399866249564966,
+      "learning_rate": 0.0001135706690990657,
+      "loss": 1.0732,
+      "step": 10114
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32184020587696033,
+      "learning_rate": 0.00011355499578385256,
+      "loss": 1.1721,
+      "step": 10115
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30947844684370757,
+      "learning_rate": 0.00011353932212941709,
+      "loss": 0.9335,
+      "step": 10116
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30689326685860674,
+      "learning_rate": 0.00011352364813615159,
+      "loss": 1.154,
+      "step": 10117
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.35179570677545263,
+      "learning_rate": 0.0001135079738044483,
+      "loss": 1.1623,
+      "step": 10118
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2895264616854555,
+      "learning_rate": 0.00011349229913469948,
+      "loss": 1.1446,
+      "step": 10119
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31453418960821145,
+      "learning_rate": 0.00011347662412729738,
+      "loss": 1.1001,
+      "step": 10120
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2761763344790832,
+      "learning_rate": 0.00011346094878263431,
+      "loss": 0.9588,
+      "step": 10121
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30740807652927366,
+      "learning_rate": 0.00011344527310110256,
+      "loss": 1.0145,
+      "step": 10122
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2852359412921041,
+      "learning_rate": 0.00011342959708309435,
+      "loss": 0.9789,
+      "step": 10123
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2870738764698151,
+      "learning_rate": 0.00011341392072900205,
+      "loss": 1.0349,
+      "step": 10124
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3079179973519736,
+      "learning_rate": 0.00011339824403921797,
+      "loss": 1.0653,
+      "step": 10125
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.256710403425938,
+      "learning_rate": 0.0001133825670141344,
+      "loss": 0.9875,
+      "step": 10126
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3171962319601306,
+      "learning_rate": 0.00011336688965414369,
+      "loss": 1.1135,
+      "step": 10127
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2964396038679801,
+      "learning_rate": 0.00011335121195963813,
+      "loss": 1.0385,
+      "step": 10128
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2750463815164073,
+      "learning_rate": 0.00011333553393101013,
+      "loss": 0.9907,
+      "step": 10129
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3245742501930826,
+      "learning_rate": 0.00011331985556865201,
+      "loss": 1.0089,
+      "step": 10130
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2976235078728115,
+      "learning_rate": 0.00011330417687295614,
+      "loss": 1.0739,
+      "step": 10131
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30953794415640545,
+      "learning_rate": 0.00011328849784431488,
+      "loss": 1.1504,
+      "step": 10132
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2779345999485772,
+      "learning_rate": 0.00011327281848312059,
+      "loss": 1.1356,
+      "step": 10133
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2802541009558856,
+      "learning_rate": 0.0001132571387897657,
+      "loss": 1.0917,
+      "step": 10134
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27849384342581196,
+      "learning_rate": 0.00011324145876464259,
+      "loss": 1.0923,
+      "step": 10135
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3318350576277318,
+      "learning_rate": 0.00011322577840814361,
+      "loss": 0.9506,
+      "step": 10136
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3459815664705523,
+      "learning_rate": 0.00011321009772066124,
+      "loss": 1.0129,
+      "step": 10137
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3037533943998662,
+      "learning_rate": 0.00011319441670258788,
+      "loss": 1.0418,
+      "step": 10138
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2520746414367993,
+      "learning_rate": 0.00011317873535431591,
+      "loss": 1.0216,
+      "step": 10139
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2661292266112255,
+      "learning_rate": 0.00011316305367623785,
+      "loss": 1.1706,
+      "step": 10140
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3063821264146261,
+      "learning_rate": 0.00011314737166874607,
+      "loss": 1.0417,
+      "step": 10141
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26323311620303635,
+      "learning_rate": 0.00011313168933223306,
+      "loss": 1.0204,
+      "step": 10142
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27238653921733935,
+      "learning_rate": 0.00011311600666709126,
+      "loss": 1.0871,
+      "step": 10143
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3013746324023429,
+      "learning_rate": 0.00011310032367371317,
+      "loss": 1.0356,
+      "step": 10144
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2574291995601369,
+      "learning_rate": 0.00011308464035249125,
+      "loss": 1.0812,
+      "step": 10145
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2869306171965447,
+      "learning_rate": 0.00011306895670381797,
+      "loss": 1.0173,
+      "step": 10146
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.23922368061607324,
+      "learning_rate": 0.00011305327272808583,
+      "loss": 1.0289,
+      "step": 10147
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.25159221628276673,
+      "learning_rate": 0.00011303758842568735,
+      "loss": 0.999,
+      "step": 10148
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2868191888678923,
+      "learning_rate": 0.00011302190379701503,
+      "loss": 1.0263,
+      "step": 10149
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.361397853429607,
+      "learning_rate": 0.00011300621884246136,
+      "loss": 1.1683,
+      "step": 10150
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2826280133952878,
+      "learning_rate": 0.00011299053356241891,
+      "loss": 1.0963,
+      "step": 10151
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28972955414613405,
+      "learning_rate": 0.00011297484795728019,
+      "loss": 1.0244,
+      "step": 10152
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3320877897903177,
+      "learning_rate": 0.00011295916202743773,
+      "loss": 1.0985,
+      "step": 10153
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2928253376533602,
+      "learning_rate": 0.00011294347577328412,
+      "loss": 1.0557,
+      "step": 10154
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33688760633249853,
+      "learning_rate": 0.00011292778919521189,
+      "loss": 1.149,
+      "step": 10155
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27676103323848317,
+      "learning_rate": 0.00011291210229361362,
+      "loss": 1.1117,
+      "step": 10156
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29026256750352997,
+      "learning_rate": 0.00011289641506888182,
+      "loss": 1.0704,
+      "step": 10157
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29334540208045223,
+      "learning_rate": 0.0001128807275214092,
+      "loss": 0.9988,
+      "step": 10158
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2818605763391759,
+      "learning_rate": 0.00011286503965158822,
+      "loss": 1.0227,
+      "step": 10159
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27801097090219445,
+      "learning_rate": 0.00011284935145981157,
+      "loss": 0.9915,
+      "step": 10160
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2805850444759864,
+      "learning_rate": 0.0001128336629464718,
+      "loss": 1.0897,
+      "step": 10161
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3032491771748229,
+      "learning_rate": 0.00011281797411196156,
+      "loss": 1.0666,
+      "step": 10162
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.300948028153539,
+      "learning_rate": 0.00011280228495667346,
+      "loss": 1.1215,
+      "step": 10163
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.32505655549938,
+      "learning_rate": 0.00011278659548100015,
+      "loss": 1.0754,
+      "step": 10164
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2830441409159092,
+      "learning_rate": 0.00011277090568533424,
+      "loss": 1.1184,
+      "step": 10165
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27235511959421765,
+      "learning_rate": 0.0001127552155700684,
+      "loss": 1.0373,
+      "step": 10166
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3112142074976755,
+      "learning_rate": 0.00011273952513559525,
+      "loss": 1.0028,
+      "step": 10167
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27216215210937883,
+      "learning_rate": 0.0001127238343823075,
+      "loss": 1.0513,
+      "step": 10168
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2953527746729821,
+      "learning_rate": 0.0001127081433105978,
+      "loss": 1.0803,
+      "step": 10169
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2897815242186968,
+      "learning_rate": 0.0001126924519208588,
+      "loss": 1.0248,
+      "step": 10170
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28105812749428083,
+      "learning_rate": 0.00011267676021348323,
+      "loss": 1.013,
+      "step": 10171
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30722992978249025,
+      "learning_rate": 0.00011266106818886377,
+      "loss": 1.0101,
+      "step": 10172
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3033918428966205,
+      "learning_rate": 0.00011264537584739314,
+      "loss": 1.0618,
+      "step": 10173
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28043324409698095,
+      "learning_rate": 0.00011262968318946398,
+      "loss": 0.9475,
+      "step": 10174
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2609838079910845,
+      "learning_rate": 0.00011261399021546912,
+      "loss": 0.9683,
+      "step": 10175
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3252941464289039,
+      "learning_rate": 0.00011259829692580119,
+      "loss": 1.0946,
+      "step": 10176
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3102405322061509,
+      "learning_rate": 0.00011258260332085298,
+      "loss": 1.0548,
+      "step": 10177
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27928723327298144,
+      "learning_rate": 0.0001125669094010172,
+      "loss": 1.0014,
+      "step": 10178
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3043342368498864,
+      "learning_rate": 0.00011255121516668663,
+      "loss": 1.0738,
+      "step": 10179
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28720036803685933,
+      "learning_rate": 0.00011253552061825398,
+      "loss": 1.0355,
+      "step": 10180
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.284815493784025,
+      "learning_rate": 0.00011251982575611209,
+      "loss": 1.0894,
+      "step": 10181
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2556934414454461,
+      "learning_rate": 0.00011250413058065365,
+      "loss": 0.9576,
+      "step": 10182
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.26083285136088175,
+      "learning_rate": 0.00011248843509227152,
+      "loss": 0.945,
+      "step": 10183
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28387550518261473,
+      "learning_rate": 0.00011247273929135841,
+      "loss": 1.1312,
+      "step": 10184
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.33228217976195246,
+      "learning_rate": 0.00011245704317830721,
+      "loss": 1.0897,
+      "step": 10185
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.31237530100881616,
+      "learning_rate": 0.00011244134675351066,
+      "loss": 1.1166,
+      "step": 10186
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30031596382794623,
+      "learning_rate": 0.00011242565001736159,
+      "loss": 1.0411,
+      "step": 10187
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3052672362577054,
+      "learning_rate": 0.00011240995297025281,
+      "loss": 0.9902,
+      "step": 10188
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.306885450267344,
+      "learning_rate": 0.00011239425561257717,
+      "loss": 1.0137,
+      "step": 10189
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29453938744454133,
+      "learning_rate": 0.00011237855794472748,
+      "loss": 1.0604,
+      "step": 10190
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.24569112604498705,
+      "learning_rate": 0.00011236285996709659,
+      "loss": 1.128,
+      "step": 10191
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30552330478731954,
+      "learning_rate": 0.00011234716168007737,
+      "loss": 1.0256,
+      "step": 10192
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29514963250865506,
+      "learning_rate": 0.00011233146308406268,
+      "loss": 1.1758,
+      "step": 10193
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2895811885576239,
+      "learning_rate": 0.00011231576417944536,
+      "loss": 1.0462,
+      "step": 10194
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2674722074074362,
+      "learning_rate": 0.00011230006496661831,
+      "loss": 1.2296,
+      "step": 10195
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31631409210970807,
+      "learning_rate": 0.00011228436544597442,
+      "loss": 1.0523,
+      "step": 10196
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2383251012221512,
+      "learning_rate": 0.00011226866561790653,
+      "loss": 0.9241,
+      "step": 10197
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3245079733703714,
+      "learning_rate": 0.00011225296548280759,
+      "loss": 1.0727,
+      "step": 10198
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28495612155493694,
+      "learning_rate": 0.0001122372650410705,
+      "loss": 0.9841,
+      "step": 10199
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30013506380723926,
+      "learning_rate": 0.00011222156429308812,
+      "loss": 0.9932,
+      "step": 10200
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2786075519347718,
+      "learning_rate": 0.00011220586323925346,
+      "loss": 1.0858,
+      "step": 10201
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3138110381506769,
+      "learning_rate": 0.00011219016187995937,
+      "loss": 1.1049,
+      "step": 10202
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27581534819480813,
+      "learning_rate": 0.00011217446021559883,
+      "loss": 0.9373,
+      "step": 10203
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3292494819868315,
+      "learning_rate": 0.00011215875824656477,
+      "loss": 0.9851,
+      "step": 10204
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2655547214014843,
+      "learning_rate": 0.00011214305597325015,
+      "loss": 0.9125,
+      "step": 10205
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2747966640064946,
+      "learning_rate": 0.00011212735339604792,
+      "loss": 0.9829,
+      "step": 10206
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26757047246677396,
+      "learning_rate": 0.00011211165051535104,
+      "loss": 0.9451,
+      "step": 10207
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27347785909550587,
+      "learning_rate": 0.00011209594733155251,
+      "loss": 1.0511,
+      "step": 10208
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.250251681777428,
+      "learning_rate": 0.00011208024384504527,
+      "loss": 1.0769,
+      "step": 10209
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2977263384929927,
+      "learning_rate": 0.00011206454005622237,
+      "loss": 1.0488,
+      "step": 10210
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26128674737229673,
+      "learning_rate": 0.00011204883596547676,
+      "loss": 0.9721,
+      "step": 10211
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2864777874901107,
+      "learning_rate": 0.00011203313157320146,
+      "loss": 0.9903,
+      "step": 10212
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2930793218779365,
+      "learning_rate": 0.00011201742687978946,
+      "loss": 1.0835,
+      "step": 10213
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2726110420231643,
+      "learning_rate": 0.0001120017218856338,
+      "loss": 1.0418,
+      "step": 10214
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26500863939608205,
+      "learning_rate": 0.00011198601659112753,
+      "loss": 0.9941,
+      "step": 10215
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.35922537571617147,
+      "learning_rate": 0.00011197031099666366,
+      "loss": 1.1,
+      "step": 10216
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26403707584086666,
+      "learning_rate": 0.00011195460510263523,
+      "loss": 1.0841,
+      "step": 10217
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30153285229911025,
+      "learning_rate": 0.00011193889890943528,
+      "loss": 0.9745,
+      "step": 10218
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27058298329981373,
+      "learning_rate": 0.0001119231924174569,
+      "loss": 1.0014,
+      "step": 10219
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2890597865454572,
+      "learning_rate": 0.00011190748562709314,
+      "loss": 1.1932,
+      "step": 10220
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2559758601582075,
+      "learning_rate": 0.00011189177853873705,
+      "loss": 1.035,
+      "step": 10221
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2675198324495206,
+      "learning_rate": 0.00011187607115278173,
+      "loss": 1.0852,
+      "step": 10222
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2984560752244984,
+      "learning_rate": 0.00011186036346962025,
+      "loss": 1.0197,
+      "step": 10223
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2746333508360344,
+      "learning_rate": 0.00011184465548964575,
+      "loss": 1.1096,
+      "step": 10224
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2911661232884761,
+      "learning_rate": 0.00011182894721325128,
+      "loss": 1.0876,
+      "step": 10225
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2545901819238632,
+      "learning_rate": 0.00011181323864082999,
+      "loss": 1.008,
+      "step": 10226
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27438906716939765,
+      "learning_rate": 0.00011179752977277498,
+      "loss": 0.9276,
+      "step": 10227
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30116616901446114,
+      "learning_rate": 0.00011178182060947935,
+      "loss": 0.9543,
+      "step": 10228
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2752406915494225,
+      "learning_rate": 0.00011176611115133628,
+      "loss": 1.0355,
+      "step": 10229
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27687088339188193,
+      "learning_rate": 0.00011175040139873889,
+      "loss": 0.9992,
+      "step": 10230
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2650331819262323,
+      "learning_rate": 0.00011173469135208028,
+      "loss": 0.9954,
+      "step": 10231
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29487040103252415,
+      "learning_rate": 0.00011171898101175369,
+      "loss": 0.9526,
+      "step": 10232
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2784410390172419,
+      "learning_rate": 0.0001117032703781522,
+      "loss": 1.1314,
+      "step": 10233
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34321818169181995,
+      "learning_rate": 0.00011168755945166905,
+      "loss": 1.0451,
+      "step": 10234
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2934212795606079,
+      "learning_rate": 0.00011167184823269735,
+      "loss": 0.9916,
+      "step": 10235
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27720714282535736,
+      "learning_rate": 0.00011165613672163032,
+      "loss": 1.052,
+      "step": 10236
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2827387556355499,
+      "learning_rate": 0.00011164042491886115,
+      "loss": 1.0031,
+      "step": 10237
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3020361414751645,
+      "learning_rate": 0.00011162471282478299,
+      "loss": 1.1295,
+      "step": 10238
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2926727806203999,
+      "learning_rate": 0.00011160900043978915,
+      "loss": 1.0201,
+      "step": 10239
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29327346670931664,
+      "learning_rate": 0.00011159328776427274,
+      "loss": 1.0045,
+      "step": 10240
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3197321134655832,
+      "learning_rate": 0.00011157757479862701,
+      "loss": 0.9832,
+      "step": 10241
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2834245023692689,
+      "learning_rate": 0.00011156186154324522,
+      "loss": 0.9674,
+      "step": 10242
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2925004499767554,
+      "learning_rate": 0.00011154614799852055,
+      "loss": 1.0646,
+      "step": 10243
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25673872599009095,
+      "learning_rate": 0.0001115304341648463,
+      "loss": 1.04,
+      "step": 10244
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2808493957291415,
+      "learning_rate": 0.00011151472004261565,
+      "loss": 1.0743,
+      "step": 10245
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2908844238513358,
+      "learning_rate": 0.00011149900563222193,
+      "loss": 1.0376,
+      "step": 10246
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2887693925794027,
+      "learning_rate": 0.00011148329093405836,
+      "loss": 1.018,
+      "step": 10247
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29234813333297677,
+      "learning_rate": 0.0001114675759485182,
+      "loss": 1.0377,
+      "step": 10248
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28841549577391123,
+      "learning_rate": 0.00011145186067599478,
+      "loss": 1.1507,
+      "step": 10249
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31377105859709736,
+      "learning_rate": 0.00011143614511688132,
+      "loss": 0.9281,
+      "step": 10250
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27520050582114036,
+      "learning_rate": 0.00011142042927157114,
+      "loss": 1.0736,
+      "step": 10251
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2836489124077,
+      "learning_rate": 0.00011140471314045755,
+      "loss": 1.1397,
+      "step": 10252
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3196462341558277,
+      "learning_rate": 0.00011138899672393386,
+      "loss": 1.1109,
+      "step": 10253
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2919022639490787,
+      "learning_rate": 0.00011137328002239335,
+      "loss": 1.0623,
+      "step": 10254
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2668993177956516,
+      "learning_rate": 0.00011135756303622937,
+      "loss": 0.9756,
+      "step": 10255
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29098900383647164,
+      "learning_rate": 0.00011134184576583525,
+      "loss": 0.981,
+      "step": 10256
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2741127564871344,
+      "learning_rate": 0.00011132612821160428,
+      "loss": 1.0042,
+      "step": 10257
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3179337291974993,
+      "learning_rate": 0.00011131041037392984,
+      "loss": 0.9453,
+      "step": 10258
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2903023541952086,
+      "learning_rate": 0.00011129469225320527,
+      "loss": 1.1199,
+      "step": 10259
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2762878923470942,
+      "learning_rate": 0.00011127897384982396,
+      "loss": 1.0547,
+      "step": 10260
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29633017173721615,
+      "learning_rate": 0.00011126325516417921,
+      "loss": 1.0946,
+      "step": 10261
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30069391214448493,
+      "learning_rate": 0.00011124753619666441,
+      "loss": 1.1047,
+      "step": 10262
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30197537398471846,
+      "learning_rate": 0.000111231816947673,
+      "loss": 1.0716,
+      "step": 10263
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.22716259808522002,
+      "learning_rate": 0.00011121609741759824,
+      "loss": 1.1495,
+      "step": 10264
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.33447762653923324,
+      "learning_rate": 0.00011120037760683364,
+      "loss": 1.0615,
+      "step": 10265
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2677195592128131,
+      "learning_rate": 0.00011118465751577254,
+      "loss": 1.0661,
+      "step": 10266
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.25916514691347986,
+      "learning_rate": 0.00011116893714480836,
+      "loss": 1.0618,
+      "step": 10267
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2735392836561538,
+      "learning_rate": 0.0001111532164943345,
+      "loss": 1.1209,
+      "step": 10268
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.306277634485208,
+      "learning_rate": 0.0001111374955647444,
+      "loss": 1.0481,
+      "step": 10269
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.24999171813245338,
+      "learning_rate": 0.00011112177435643147,
+      "loss": 0.9234,
+      "step": 10270
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26942078573492634,
+      "learning_rate": 0.00011110605286978914,
+      "loss": 1.0864,
+      "step": 10271
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26245411051479745,
+      "learning_rate": 0.00011109033110521086,
+      "loss": 1.1104,
+      "step": 10272
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31836586363546904,
+      "learning_rate": 0.00011107460906309008,
+      "loss": 0.9678,
+      "step": 10273
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3164102794109734,
+      "learning_rate": 0.00011105888674382025,
+      "loss": 1.063,
+      "step": 10274
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3168927687531968,
+      "learning_rate": 0.0001110431641477948,
+      "loss": 1.0617,
+      "step": 10275
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2901612455077041,
+      "learning_rate": 0.00011102744127540728,
+      "loss": 1.0798,
+      "step": 10276
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2808843978397793,
+      "learning_rate": 0.0001110117181270511,
+      "loss": 1.083,
+      "step": 10277
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28149992419883424,
+      "learning_rate": 0.00011099599470311972,
+      "loss": 1.0127,
+      "step": 10278
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31449595403015373,
+      "learning_rate": 0.00011098027100400667,
+      "loss": 1.1207,
+      "step": 10279
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.243827419121298,
+      "learning_rate": 0.00011096454703010546,
+      "loss": 0.9526,
+      "step": 10280
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27945107442928213,
+      "learning_rate": 0.00011094882278180956,
+      "loss": 0.9607,
+      "step": 10281
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27291199800336474,
+      "learning_rate": 0.00011093309825951245,
+      "loss": 1.0538,
+      "step": 10282
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30765998442794057,
+      "learning_rate": 0.00011091737346360773,
+      "loss": 1.1567,
+      "step": 10283
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28547060490679954,
+      "learning_rate": 0.00011090164839448887,
+      "loss": 0.9414,
+      "step": 10284
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28516651031906826,
+      "learning_rate": 0.00011088592305254939,
+      "loss": 1.1272,
+      "step": 10285
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28976084124588714,
+      "learning_rate": 0.00011087019743818288,
+      "loss": 1.067,
+      "step": 10286
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30715911284891306,
+      "learning_rate": 0.00011085447155178279,
+      "loss": 1.0638,
+      "step": 10287
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.26407524082018785,
+      "learning_rate": 0.00011083874539374277,
+      "loss": 1.0787,
+      "step": 10288
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2694386589995852,
+      "learning_rate": 0.00011082301896445633,
+      "loss": 1.0979,
+      "step": 10289
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27217656363948856,
+      "learning_rate": 0.00011080729226431703,
+      "loss": 1.141,
+      "step": 10290
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2794911223689666,
+      "learning_rate": 0.00011079156529371846,
+      "loss": 1.049,
+      "step": 10291
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.27195359614242276,
+      "learning_rate": 0.00011077583805305418,
+      "loss": 0.9813,
+      "step": 10292
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.29038357957473354,
+      "learning_rate": 0.00011076011054271778,
+      "loss": 1.1996,
+      "step": 10293
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.28199425499385034,
+      "learning_rate": 0.00011074438276310287,
+      "loss": 1.1444,
+      "step": 10294
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.30433530645019496,
+      "learning_rate": 0.00011072865471460301,
+      "loss": 1.0179,
+      "step": 10295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26891004746639385,
+      "learning_rate": 0.00011071292639761181,
+      "loss": 1.0334,
+      "step": 10296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29402614889194606,
+      "learning_rate": 0.0001106971978125229,
+      "loss": 1.0908,
+      "step": 10297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2559539105537088,
+      "learning_rate": 0.00011068146895972993,
+      "loss": 1.1747,
+      "step": 10298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2845273763898639,
+      "learning_rate": 0.00011066573983962642,
+      "loss": 1.0114,
+      "step": 10299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3025475977583579,
+      "learning_rate": 0.00011065001045260613,
+      "loss": 1.0435,
+      "step": 10300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2747465345583122,
+      "learning_rate": 0.00011063428079906259,
+      "loss": 1.1137,
+      "step": 10301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2990549222362699,
+      "learning_rate": 0.0001106185508793895,
+      "loss": 1.0761,
+      "step": 10302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3060379979713365,
+      "learning_rate": 0.00011060282069398052,
+      "loss": 1.0701,
+      "step": 10303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25935605894516695,
+      "learning_rate": 0.00011058709024322929,
+      "loss": 1.0854,
+      "step": 10304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.284568963561047,
+      "learning_rate": 0.00011057135952752943,
+      "loss": 0.9767,
+      "step": 10305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.268579961216263,
+      "learning_rate": 0.00011055562854727471,
+      "loss": 1.0459,
+      "step": 10306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26453950568162055,
+      "learning_rate": 0.00011053989730285869,
+      "loss": 1.0633,
+      "step": 10307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2840144675918696,
+      "learning_rate": 0.00011052416579467518,
+      "loss": 0.8992,
+      "step": 10308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2526273337620793,
+      "learning_rate": 0.00011050843402311777,
+      "loss": 1.0391,
+      "step": 10309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2907758381732847,
+      "learning_rate": 0.00011049270198858019,
+      "loss": 1.1091,
+      "step": 10310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29330977370977224,
+      "learning_rate": 0.00011047696969145618,
+      "loss": 1.1539,
+      "step": 10311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28027168248307766,
+      "learning_rate": 0.00011046123713213939,
+      "loss": 1.0466,
+      "step": 10312
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29433174543880003,
+      "learning_rate": 0.00011044550431102358,
+      "loss": 1.0868,
+      "step": 10313
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28747079837458267,
+      "learning_rate": 0.00011042977122850247,
+      "loss": 0.9917,
+      "step": 10314
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2841412064889096,
+      "learning_rate": 0.00011041403788496976,
+      "loss": 1.23,
+      "step": 10315
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26187375710340965,
+      "learning_rate": 0.0001103983042808192,
+      "loss": 1.0055,
+      "step": 10316
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32286275918136254,
+      "learning_rate": 0.00011038257041644455,
+      "loss": 0.993,
+      "step": 10317
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28197212224874324,
+      "learning_rate": 0.00011036683629223958,
+      "loss": 1.003,
+      "step": 10318
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25905993814852957,
+      "learning_rate": 0.00011035110190859796,
+      "loss": 1.0416,
+      "step": 10319
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2894375367829976,
+      "learning_rate": 0.00011033536726591356,
+      "loss": 1.0795,
+      "step": 10320
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32587938767421387,
+      "learning_rate": 0.00011031963236458008,
+      "loss": 1.0801,
+      "step": 10321
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2678020975589778,
+      "learning_rate": 0.00011030389720499132,
+      "loss": 1.1145,
+      "step": 10322
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30583316214893913,
+      "learning_rate": 0.00011028816178754104,
+      "loss": 1.0778,
+      "step": 10323
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32483933693639894,
+      "learning_rate": 0.00011027242611262306,
+      "loss": 1.0419,
+      "step": 10324
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31586042610524007,
+      "learning_rate": 0.00011025669018063116,
+      "loss": 1.0958,
+      "step": 10325
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2878776556420424,
+      "learning_rate": 0.00011024095399195913,
+      "loss": 1.0656,
+      "step": 10326
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2830198312625606,
+      "learning_rate": 0.0001102252175470008,
+      "loss": 1.0493,
+      "step": 10327
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2803658012096297,
+      "learning_rate": 0.00011020948084614995,
+      "loss": 0.9544,
+      "step": 10328
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3160224185691885,
+      "learning_rate": 0.00011019374388980046,
+      "loss": 0.9955,
+      "step": 10329
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28845814066617365,
+      "learning_rate": 0.0001101780066783461,
+      "loss": 1.1675,
+      "step": 10330
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2892793784279133,
+      "learning_rate": 0.00011016226921218074,
+      "loss": 0.9273,
+      "step": 10331
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27909437771191264,
+      "learning_rate": 0.0001101465314916982,
+      "loss": 1.1146,
+      "step": 10332
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3021556255362156,
+      "learning_rate": 0.00011013079351729232,
+      "loss": 1.0174,
+      "step": 10333
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3088606738990712,
+      "learning_rate": 0.000110115055289357,
+      "loss": 1.0696,
+      "step": 10334
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2532028338885159,
+      "learning_rate": 0.00011009931680828604,
+      "loss": 1.0509,
+      "step": 10335
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2644967377127452,
+      "learning_rate": 0.00011008357807447334,
+      "loss": 1.011,
+      "step": 10336
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25911644503390924,
+      "learning_rate": 0.00011006783908831275,
+      "loss": 0.9568,
+      "step": 10337
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3093805850346706,
+      "learning_rate": 0.00011005209985019817,
+      "loss": 1.0442,
+      "step": 10338
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26605780506457216,
+      "learning_rate": 0.00011003636036052347,
+      "loss": 1.12,
+      "step": 10339
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2738908065467255,
+      "learning_rate": 0.00011002062061968255,
+      "loss": 1.0423,
+      "step": 10340
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.38868786111630993,
+      "learning_rate": 0.00011000488062806929,
+      "loss": 1.0834,
+      "step": 10341
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3172034502101059,
+      "learning_rate": 0.00010998914038607762,
+      "loss": 1.0222,
+      "step": 10342
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2858843799078492,
+      "learning_rate": 0.0001099733998941014,
+      "loss": 1.0284,
+      "step": 10343
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2646838577150544,
+      "learning_rate": 0.00010995765915253462,
+      "loss": 1.1413,
+      "step": 10344
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27314529183764996,
+      "learning_rate": 0.00010994191816177115,
+      "loss": 1.0144,
+      "step": 10345
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25795307029613257,
+      "learning_rate": 0.0001099261769222049,
+      "loss": 1.1011,
+      "step": 10346
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.269654266329036,
+      "learning_rate": 0.00010991043543422987,
+      "loss": 1.093,
+      "step": 10347
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2873513143757092,
+      "learning_rate": 0.00010989469369823993,
+      "loss": 1.0575,
+      "step": 10348
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2966705516629044,
+      "learning_rate": 0.0001098789517146291,
+      "loss": 0.9566,
+      "step": 10349
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30235296838928694,
+      "learning_rate": 0.00010986320948379125,
+      "loss": 1.0973,
+      "step": 10350
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.22856559446563995,
+      "learning_rate": 0.00010984746700612043,
+      "loss": 1.0409,
+      "step": 10351
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29679007545691055,
+      "learning_rate": 0.00010983172428201055,
+      "loss": 1.0134,
+      "step": 10352
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3089775230936333,
+      "learning_rate": 0.00010981598131185558,
+      "loss": 1.0447,
+      "step": 10353
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29725469954817196,
+      "learning_rate": 0.00010980023809604951,
+      "loss": 1.0729,
+      "step": 10354
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2626951615117108,
+      "learning_rate": 0.00010978449463498632,
+      "loss": 0.9726,
+      "step": 10355
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27242407027742316,
+      "learning_rate": 0.00010976875092906003,
+      "loss": 1.0434,
+      "step": 10356
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2947561469672809,
+      "learning_rate": 0.00010975300697866456,
+      "loss": 1.0845,
+      "step": 10357
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29370486489779185,
+      "learning_rate": 0.00010973726278419398,
+      "loss": 0.9322,
+      "step": 10358
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2928377780175839,
+      "learning_rate": 0.00010972151834604229,
+      "loss": 1.0208,
+      "step": 10359
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.32099727413336887,
+      "learning_rate": 0.00010970577366460349,
+      "loss": 1.0144,
+      "step": 10360
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2763946202818681,
+      "learning_rate": 0.00010969002874027161,
+      "loss": 1.206,
+      "step": 10361
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.29031856568736375,
+      "learning_rate": 0.00010967428357344067,
+      "loss": 1.0013,
+      "step": 10362
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2600423194599826,
+      "learning_rate": 0.00010965853816450469,
+      "loss": 1.1104,
+      "step": 10363
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.262125842100789,
+      "learning_rate": 0.00010964279251385774,
+      "loss": 1.0854,
+      "step": 10364
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2983375163299008,
+      "learning_rate": 0.00010962704662189383,
+      "loss": 1.0486,
+      "step": 10365
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27574617002830204,
+      "learning_rate": 0.00010961130048900705,
+      "loss": 1.0465,
+      "step": 10366
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2474828198501618,
+      "learning_rate": 0.0001095955541155914,
+      "loss": 1.0645,
+      "step": 10367
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.33229191383287654,
+      "learning_rate": 0.000109579807502041,
+      "loss": 0.9869,
+      "step": 10368
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2982566291542284,
+      "learning_rate": 0.0001095640606487499,
+      "loss": 1.0616,
+      "step": 10369
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3011080575254373,
+      "learning_rate": 0.00010954831355611215,
+      "loss": 1.0475,
+      "step": 10370
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2954042572701708,
+      "learning_rate": 0.00010953256622452185,
+      "loss": 1.0212,
+      "step": 10371
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3015044088263099,
+      "learning_rate": 0.0001095168186543731,
+      "loss": 1.0839,
+      "step": 10372
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.36161978315568827,
+      "learning_rate": 0.00010950107084605998,
+      "loss": 0.9966,
+      "step": 10373
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2947245497967197,
+      "learning_rate": 0.00010948532279997664,
+      "loss": 1.0052,
+      "step": 10374
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.278629231797063,
+      "learning_rate": 0.00010946957451651709,
+      "loss": 0.974,
+      "step": 10375
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3081240007766056,
+      "learning_rate": 0.0001094538259960755,
+      "loss": 0.9513,
+      "step": 10376
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.37939549753517904,
+      "learning_rate": 0.00010943807723904593,
+      "loss": 1.1191,
+      "step": 10377
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24633288069760065,
+      "learning_rate": 0.0001094223282458226,
+      "loss": 1.0712,
+      "step": 10378
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3096687909717489,
+      "learning_rate": 0.00010940657901679956,
+      "loss": 0.9914,
+      "step": 10379
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3063298337056405,
+      "learning_rate": 0.00010939082955237096,
+      "loss": 1.1432,
+      "step": 10380
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2745100518027808,
+      "learning_rate": 0.00010937507985293098,
+      "loss": 1.0335,
+      "step": 10381
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.25605757624283976,
+      "learning_rate": 0.00010935932991887372,
+      "loss": 0.9888,
+      "step": 10382
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27108913673944884,
+      "learning_rate": 0.00010934357975059334,
+      "loss": 1.1554,
+      "step": 10383
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3465731349344641,
+      "learning_rate": 0.000109327829348484,
+      "loss": 1.0209,
+      "step": 10384
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2661351715791654,
+      "learning_rate": 0.00010931207871293987,
+      "loss": 1.1225,
+      "step": 10385
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2992239551094853,
+      "learning_rate": 0.00010929632784435513,
+      "loss": 1.1448,
+      "step": 10386
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2714660858795361,
+      "learning_rate": 0.00010928057674312393,
+      "loss": 1.1303,
+      "step": 10387
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3560152461135453,
+      "learning_rate": 0.00010926482540964047,
+      "loss": 1.0939,
+      "step": 10388
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.33230188942732725,
+      "learning_rate": 0.00010924907384429892,
+      "loss": 1.0654,
+      "step": 10389
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2648343149387062,
+      "learning_rate": 0.0001092333220474935,
+      "loss": 1.0357,
+      "step": 10390
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2959171922667274,
+      "learning_rate": 0.00010921757001961839,
+      "loss": 1.0695,
+      "step": 10391
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.26506456061345024,
+      "learning_rate": 0.00010920181776106779,
+      "loss": 1.1593,
+      "step": 10392
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3204886263895637,
+      "learning_rate": 0.00010918606527223593,
+      "loss": 1.0391,
+      "step": 10393
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24450093898680367,
+      "learning_rate": 0.00010917031255351699,
+      "loss": 0.9267,
+      "step": 10394
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28085674386349635,
+      "learning_rate": 0.00010915455960530526,
+      "loss": 1.0893,
+      "step": 10395
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28257173644304456,
+      "learning_rate": 0.00010913880642799488,
+      "loss": 1.0546,
+      "step": 10396
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3084129119694214,
+      "learning_rate": 0.00010912305302198014,
+      "loss": 1.0221,
+      "step": 10397
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2778700782623621,
+      "learning_rate": 0.00010910729938765528,
+      "loss": 1.0197,
+      "step": 10398
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.24322367397269376,
+      "learning_rate": 0.00010909154552541449,
+      "loss": 0.9353,
+      "step": 10399
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.27278544222892476,
+      "learning_rate": 0.0001090757914356521,
+      "loss": 1.076,
+      "step": 10400
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2869093663218051,
+      "learning_rate": 0.00010906003711876229,
+      "loss": 0.9683,
+      "step": 10401
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2941519770990777,
+      "learning_rate": 0.00010904428257513939,
+      "loss": 1.0158,
+      "step": 10402
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27338008253801777,
+      "learning_rate": 0.00010902852780517763,
+      "loss": 1.0386,
+      "step": 10403
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28389255664388,
+      "learning_rate": 0.00010901277280927124,
+      "loss": 1.1183,
+      "step": 10404
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3443690879581628,
+      "learning_rate": 0.0001089970175878146,
+      "loss": 1.098,
+      "step": 10405
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2551799172999775,
+      "learning_rate": 0.00010898126214120194,
+      "loss": 1.0581,
+      "step": 10406
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.282716317579218,
+      "learning_rate": 0.0001089655064698275,
+      "loss": 0.9715,
+      "step": 10407
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31282993584728686,
+      "learning_rate": 0.00010894975057408568,
+      "loss": 1.047,
+      "step": 10408
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30036322798930976,
+      "learning_rate": 0.00010893399445437071,
+      "loss": 1.188,
+      "step": 10409
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3325931928877523,
+      "learning_rate": 0.0001089182381110769,
+      "loss": 0.9747,
+      "step": 10410
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3117101250717752,
+      "learning_rate": 0.00010890248154459858,
+      "loss": 1.0603,
+      "step": 10411
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28934935878323786,
+      "learning_rate": 0.00010888672475533006,
+      "loss": 1.093,
+      "step": 10412
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2716220894787029,
+      "learning_rate": 0.0001088709677436657,
+      "loss": 1.1221,
+      "step": 10413
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30194803612986443,
+      "learning_rate": 0.00010885521050999976,
+      "loss": 1.0063,
+      "step": 10414
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28931680238920227,
+      "learning_rate": 0.00010883945305472662,
+      "loss": 0.9928,
+      "step": 10415
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3229463397228268,
+      "learning_rate": 0.00010882369537824064,
+      "loss": 1.0712,
+      "step": 10416
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2954911003697874,
+      "learning_rate": 0.00010880793748093615,
+      "loss": 1.0046,
+      "step": 10417
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2872966086169036,
+      "learning_rate": 0.00010879217936320743,
+      "loss": 1.0814,
+      "step": 10418
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3310289766979793,
+      "learning_rate": 0.00010877642102544891,
+      "loss": 0.9988,
+      "step": 10419
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2661925275942424,
+      "learning_rate": 0.00010876066246805496,
+      "loss": 1.0533,
+      "step": 10420
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2868056408701368,
+      "learning_rate": 0.00010874490369141991,
+      "loss": 1.0454,
+      "step": 10421
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25382291752633346,
+      "learning_rate": 0.00010872914469593816,
+      "loss": 1.0787,
+      "step": 10422
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32950064922007866,
+      "learning_rate": 0.00010871338548200409,
+      "loss": 1.0575,
+      "step": 10423
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28542084992905475,
+      "learning_rate": 0.00010869762605001204,
+      "loss": 1.0273,
+      "step": 10424
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.274404919205005,
+      "learning_rate": 0.00010868186640035645,
+      "loss": 1.1522,
+      "step": 10425
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27910073576406763,
+      "learning_rate": 0.0001086661065334317,
+      "loss": 0.9747,
+      "step": 10426
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2754135797678086,
+      "learning_rate": 0.00010865034644963219,
+      "loss": 1.0593,
+      "step": 10427
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3036570067584713,
+      "learning_rate": 0.00010863458614935228,
+      "loss": 0.9586,
+      "step": 10428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25912219351365356,
+      "learning_rate": 0.00010861882563298648,
+      "loss": 1.1329,
+      "step": 10429
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30796423013527807,
+      "learning_rate": 0.00010860306490092916,
+      "loss": 0.9919,
+      "step": 10430
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2939383687791828,
+      "learning_rate": 0.00010858730395357468,
+      "loss": 0.9745,
+      "step": 10431
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27337663742366286,
+      "learning_rate": 0.00010857154279131754,
+      "loss": 0.9214,
+      "step": 10432
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26204823094059426,
+      "learning_rate": 0.00010855578141455216,
+      "loss": 1.0411,
+      "step": 10433
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26527110818824695,
+      "learning_rate": 0.00010854001982367296,
+      "loss": 1.0424,
+      "step": 10434
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25265488574920947,
+      "learning_rate": 0.00010852425801907442,
+      "loss": 0.9679,
+      "step": 10435
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26557773202249946,
+      "learning_rate": 0.00010850849600115096,
+      "loss": 1.0498,
+      "step": 10436
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.3095199638697367,
+      "learning_rate": 0.00010849273377029705,
+      "loss": 0.9924,
+      "step": 10437
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25086882514656306,
+      "learning_rate": 0.00010847697132690713,
+      "loss": 1.0061,
+      "step": 10438
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26227634701423524,
+      "learning_rate": 0.00010846120867137567,
+      "loss": 1.022,
+      "step": 10439
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.272747535197043,
+      "learning_rate": 0.00010844544580409717,
+      "loss": 1.049,
+      "step": 10440
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.30486635134117596,
+      "learning_rate": 0.00010842968272546603,
+      "loss": 0.9928,
+      "step": 10441
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2911807248936414,
+      "learning_rate": 0.00010841391943587682,
+      "loss": 1.0824,
+      "step": 10442
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24115347100619006,
+      "learning_rate": 0.00010839815593572398,
+      "loss": 1.0671,
+      "step": 10443
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2749883873988357,
+      "learning_rate": 0.00010838239222540203,
+      "loss": 1.1308,
+      "step": 10444
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25592035706779787,
+      "learning_rate": 0.00010836662830530539,
+      "loss": 1.0291,
+      "step": 10445
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.27534119534699625,
+      "learning_rate": 0.00010835086417582867,
+      "loss": 1.075,
+      "step": 10446
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.29613831949975483,
+      "learning_rate": 0.00010833509983736632,
+      "loss": 1.0167,
+      "step": 10447
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2971382111384855,
+      "learning_rate": 0.00010831933529031284,
+      "loss": 1.0102,
+      "step": 10448
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2833249451857255,
+      "learning_rate": 0.00010830357053506277,
+      "loss": 1.1205,
+      "step": 10449
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28978501988105787,
+      "learning_rate": 0.00010828780557201066,
+      "loss": 1.0154,
+      "step": 10450
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.32646478206285645,
+      "learning_rate": 0.00010827204040155094,
+      "loss": 0.9872,
+      "step": 10451
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.28028521065676737,
+      "learning_rate": 0.00010825627502407827,
+      "loss": 0.9816,
+      "step": 10452
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.123805046081543,
+      "eval_runtime": 4227.9159,
+      "eval_samples_per_second": 19.778,
+      "eval_steps_per_second": 2.472,
+      "step": 10452
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 20904,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 10452,
+  "total_flos": 2.304536495731507e+16,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}