diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25345 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 0,
+  "global_step": 3616,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011061946902654867,
+      "grad_norm": 0.07275390625,
+      "learning_rate": 9.997234513274337e-06,
+      "loss": 1.3771,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022123893805309734,
+      "grad_norm": 0.07080078125,
+      "learning_rate": 9.994469026548673e-06,
+      "loss": 1.3441,
+      "step": 2
+    },
+    {
+      "epoch": 0.00331858407079646,
+      "grad_norm": 0.07177734375,
+      "learning_rate": 9.99170353982301e-06,
+      "loss": 1.2924,
+      "step": 3
+    },
+    {
+      "epoch": 0.004424778761061947,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 9.988938053097345e-06,
+      "loss": 1.4005,
+      "step": 4
+    },
+    {
+      "epoch": 0.0055309734513274336,
+      "grad_norm": 0.0751953125,
+      "learning_rate": 9.986172566371683e-06,
+      "loss": 1.392,
+      "step": 5
+    },
+    {
+      "epoch": 0.00663716814159292,
+      "grad_norm": 0.07666015625,
+      "learning_rate": 9.983407079646018e-06,
+      "loss": 1.3735,
+      "step": 6
+    },
+    {
+      "epoch": 0.007743362831858407,
+      "grad_norm": 0.07568359375,
+      "learning_rate": 9.980641592920355e-06,
+      "loss": 1.2995,
+      "step": 7
+    },
+    {
+      "epoch": 0.008849557522123894,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 9.977876106194692e-06,
+      "loss": 1.3195,
+      "step": 8
+    },
+    {
+      "epoch": 0.00995575221238938,
+      "grad_norm": 0.08203125,
+      "learning_rate": 9.975110619469028e-06,
+      "loss": 1.2597,
+      "step": 9
+    },
+    {
+      "epoch": 0.011061946902654867,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 9.972345132743364e-06,
+      "loss": 1.3033,
+      "step": 10
+    },
+    {
+      "epoch": 0.012168141592920354,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 9.9695796460177e-06,
+      "loss": 1.4118,
+      "step": 11
+    },
+    {
+      "epoch": 0.01327433628318584,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.966814159292036e-06,
+      "loss": 1.3647,
+      "step": 12
+    },
+    {
+      "epoch": 0.014380530973451327,
+      "grad_norm": 0.1015625,
+      "learning_rate": 9.964048672566372e-06,
+      "loss": 1.5081,
+      "step": 13
+    },
+    {
+      "epoch": 0.015486725663716814,
+      "grad_norm": 0.111328125,
+      "learning_rate": 9.961283185840708e-06,
+      "loss": 1.488,
+      "step": 14
+    },
+    {
+      "epoch": 0.016592920353982302,
+      "grad_norm": 0.1015625,
+      "learning_rate": 9.958517699115044e-06,
+      "loss": 1.342,
+      "step": 15
+    },
+    {
+      "epoch": 0.017699115044247787,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 9.95575221238938e-06,
+      "loss": 1.325,
+      "step": 16
+    },
+    {
+      "epoch": 0.018805309734513276,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 9.952986725663718e-06,
+      "loss": 1.2776,
+      "step": 17
+    },
+    {
+      "epoch": 0.01991150442477876,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 9.950221238938054e-06,
+      "loss": 1.3764,
+      "step": 18
+    },
+    {
+      "epoch": 0.02101769911504425,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 9.94745575221239e-06,
+      "loss": 1.3939,
+      "step": 19
+    },
+    {
+      "epoch": 0.022123893805309734,
+      "grad_norm": 0.134765625,
+      "learning_rate": 9.944690265486727e-06,
+      "loss": 1.2943,
+      "step": 20
+    },
+    {
+      "epoch": 0.023230088495575223,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 9.941924778761063e-06,
+      "loss": 1.3372,
+      "step": 21
+    },
+    {
+      "epoch": 0.024336283185840708,
+      "grad_norm": 0.12109375,
+      "learning_rate": 9.939159292035399e-06,
+      "loss": 1.289,
+      "step": 22
+    },
+    {
+      "epoch": 0.025442477876106196,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 9.936393805309735e-06,
+      "loss": 1.2865,
+      "step": 23
+    },
+    {
+      "epoch": 0.02654867256637168,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.933628318584071e-06,
+      "loss": 1.4118,
+      "step": 24
+    },
+    {
+      "epoch": 0.02765486725663717,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.930862831858407e-06,
+      "loss": 1.3204,
+      "step": 25
+    },
+    {
+      "epoch": 0.028761061946902654,
+      "grad_norm": 0.146484375,
+      "learning_rate": 9.928097345132745e-06,
+      "loss": 1.4497,
+      "step": 26
+    },
+    {
+      "epoch": 0.029867256637168143,
+      "grad_norm": 0.158203125,
+      "learning_rate": 9.92533185840708e-06,
+      "loss": 1.4467,
+      "step": 27
+    },
+    {
+      "epoch": 0.030973451327433628,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.922566371681417e-06,
+      "loss": 1.3372,
+      "step": 28
+    },
+    {
+      "epoch": 0.032079646017699116,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.919800884955752e-06,
+      "loss": 1.3878,
+      "step": 29
+    },
+    {
+      "epoch": 0.033185840707964605,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.91703539823009e-06,
+      "loss": 1.3868,
+      "step": 30
+    },
+    {
+      "epoch": 0.034292035398230086,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 9.914269911504426e-06,
+      "loss": 1.2612,
+      "step": 31
+    },
+    {
+      "epoch": 0.035398230088495575,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.911504424778762e-06,
+      "loss": 1.3149,
+      "step": 32
+    },
+    {
+      "epoch": 0.03650442477876106,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.908738938053098e-06,
+      "loss": 1.4568,
+      "step": 33
+    },
+    {
+      "epoch": 0.03761061946902655,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.905973451327434e-06,
+      "loss": 1.3127,
+      "step": 34
+    },
+    {
+      "epoch": 0.03871681415929203,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 9.90320796460177e-06,
+      "loss": 1.249,
+      "step": 35
+    },
+    {
+      "epoch": 0.03982300884955752,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.900442477876106e-06,
+      "loss": 1.2219,
+      "step": 36
+    },
+    {
+      "epoch": 0.04092920353982301,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.897676991150443e-06,
+      "loss": 1.3407,
+      "step": 37
+    },
+    {
+      "epoch": 0.0420353982300885,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.89491150442478e-06,
+      "loss": 1.315,
+      "step": 38
+    },
+    {
+      "epoch": 0.04314159292035398,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 9.892146017699117e-06,
+      "loss": 1.3925,
+      "step": 39
+    },
+    {
+      "epoch": 0.04424778761061947,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.889380530973453e-06,
+      "loss": 1.2886,
+      "step": 40
+    },
+    {
+      "epoch": 0.04535398230088496,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.886615044247789e-06,
+      "loss": 1.2282,
+      "step": 41
+    },
+    {
+      "epoch": 0.046460176991150445,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.883849557522125e-06,
+      "loss": 1.3511,
+      "step": 42
+    },
+    {
+      "epoch": 0.04756637168141593,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.881084070796461e-06,
+      "loss": 1.2726,
+      "step": 43
+    },
+    {
+      "epoch": 0.048672566371681415,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 9.878318584070797e-06,
+      "loss": 1.3065,
+      "step": 44
+    },
+    {
+      "epoch": 0.049778761061946904,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.875553097345133e-06,
+      "loss": 1.313,
+      "step": 45
+    },
+    {
+      "epoch": 0.05088495575221239,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.87278761061947e-06,
+      "loss": 1.3387,
+      "step": 46
+    },
+    {
+      "epoch": 0.051991150442477874,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.870022123893807e-06,
+      "loss": 1.3307,
+      "step": 47
+    },
+    {
+      "epoch": 0.05309734513274336,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.867256637168142e-06,
+      "loss": 1.255,
+      "step": 48
+    },
+    {
+      "epoch": 0.05420353982300885,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.86449115044248e-06,
+      "loss": 1.2238,
+      "step": 49
+    },
+    {
+      "epoch": 0.05530973451327434,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.861725663716814e-06,
+      "loss": 1.2973,
+      "step": 50
+    },
+    {
+      "epoch": 0.05641592920353982,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.858960176991152e-06,
+      "loss": 1.2764,
+      "step": 51
+    },
+    {
+      "epoch": 0.05752212389380531,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.856194690265488e-06,
+      "loss": 1.1427,
+      "step": 52
+    },
+    {
+      "epoch": 0.0586283185840708,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.853429203539824e-06,
+      "loss": 1.3943,
+      "step": 53
+    },
+    {
+      "epoch": 0.059734513274336286,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.85066371681416e-06,
+      "loss": 1.228,
+      "step": 54
+    },
+    {
+      "epoch": 0.06084070796460177,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.847898230088496e-06,
+      "loss": 1.3372,
+      "step": 55
+    },
+    {
+      "epoch": 0.061946902654867256,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.845132743362832e-06,
+      "loss": 1.1932,
+      "step": 56
+    },
+    {
+      "epoch": 0.06305309734513274,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.842367256637168e-06,
+      "loss": 1.2614,
+      "step": 57
+    },
+    {
+      "epoch": 0.06415929203539823,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.839601769911505e-06,
+      "loss": 1.3055,
+      "step": 58
+    },
+    {
+      "epoch": 0.06526548672566372,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.83683628318584e-06,
+      "loss": 1.2492,
+      "step": 59
+    },
+    {
+      "epoch": 0.06637168141592921,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.834070796460179e-06,
+      "loss": 1.2859,
+      "step": 60
+    },
+    {
+      "epoch": 0.06747787610619468,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.831305309734515e-06,
+      "loss": 1.2465,
+      "step": 61
+    },
+    {
+      "epoch": 0.06858407079646017,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.82853982300885e-06,
+      "loss": 1.2437,
+      "step": 62
+    },
+    {
+      "epoch": 0.06969026548672566,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.825774336283187e-06,
+      "loss": 1.2626,
+      "step": 63
+    },
+    {
+      "epoch": 0.07079646017699115,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.823008849557523e-06,
+      "loss": 1.2529,
+      "step": 64
+    },
+    {
+      "epoch": 0.07190265486725664,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.82024336283186e-06,
+      "loss": 1.171,
+      "step": 65
+    },
+    {
+      "epoch": 0.07300884955752213,
+      "grad_norm": 0.185546875,
+      "learning_rate": 9.817477876106195e-06,
+      "loss": 1.2666,
+      "step": 66
+    },
+    {
+      "epoch": 0.07411504424778761,
+      "grad_norm": 0.1875,
+      "learning_rate": 9.814712389380531e-06,
+      "loss": 1.2449,
+      "step": 67
+    },
+    {
+      "epoch": 0.0752212389380531,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.81194690265487e-06,
+      "loss": 1.2663,
+      "step": 68
+    },
+    {
+      "epoch": 0.07632743362831858,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.809181415929204e-06,
+      "loss": 1.2034,
+      "step": 69
+    },
+    {
+      "epoch": 0.07743362831858407,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.806415929203541e-06,
+      "loss": 1.1449,
+      "step": 70
+    },
+    {
+      "epoch": 0.07853982300884955,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.803650442477876e-06,
+      "loss": 1.278,
+      "step": 71
+    },
+    {
+      "epoch": 0.07964601769911504,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.800884955752214e-06,
+      "loss": 1.1732,
+      "step": 72
+    },
+    {
+      "epoch": 0.08075221238938053,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.79811946902655e-06,
+      "loss": 1.0929,
+      "step": 73
+    },
+    {
+      "epoch": 0.08185840707964602,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.795353982300886e-06,
+      "loss": 1.1319,
+      "step": 74
+    },
+    {
+      "epoch": 0.08296460176991151,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.792588495575222e-06,
+      "loss": 1.2224,
+      "step": 75
+    },
+    {
+      "epoch": 0.084070796460177,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.789823008849558e-06,
+      "loss": 1.2126,
+      "step": 76
+    },
+    {
+      "epoch": 0.08517699115044247,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.787057522123894e-06,
+      "loss": 1.1341,
+      "step": 77
+    },
+    {
+      "epoch": 0.08628318584070796,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 9.78429203539823e-06,
+      "loss": 1.1886,
+      "step": 78
+    },
+    {
+      "epoch": 0.08738938053097345,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.781526548672567e-06,
+      "loss": 1.2075,
+      "step": 79
+    },
+    {
+      "epoch": 0.08849557522123894,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.778761061946903e-06,
+      "loss": 1.1555,
+      "step": 80
+    },
+    {
+      "epoch": 0.08960176991150443,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.77599557522124e-06,
+      "loss": 1.184,
+      "step": 81
+    },
+    {
+      "epoch": 0.09070796460176991,
+      "grad_norm": 0.146484375,
+      "learning_rate": 9.773230088495575e-06,
+      "loss": 1.1637,
+      "step": 82
+    },
+    {
+      "epoch": 0.0918141592920354,
+      "grad_norm": 0.150390625,
+      "learning_rate": 9.770464601769913e-06,
+      "loss": 1.0969,
+      "step": 83
+    },
+    {
+      "epoch": 0.09292035398230089,
+      "grad_norm": 0.158203125,
+      "learning_rate": 9.767699115044249e-06,
+      "loss": 1.1828,
+      "step": 84
+    },
+    {
+      "epoch": 0.09402654867256637,
+      "grad_norm": 0.15234375,
+      "learning_rate": 9.764933628318585e-06,
+      "loss": 1.1675,
+      "step": 85
+    },
+    {
+      "epoch": 0.09513274336283185,
+      "grad_norm": 0.142578125,
+      "learning_rate": 9.762168141592921e-06,
+      "loss": 1.0722,
+      "step": 86
+    },
+    {
+      "epoch": 0.09623893805309734,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 9.759402654867257e-06,
+      "loss": 1.1185,
+      "step": 87
+    },
+    {
+      "epoch": 0.09734513274336283,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.756637168141593e-06,
+      "loss": 1.243,
+      "step": 88
+    },
+    {
+      "epoch": 0.09845132743362832,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.75387168141593e-06,
+      "loss": 1.1313,
+      "step": 89
+    },
+    {
+      "epoch": 0.09955752212389381,
+      "grad_norm": 0.146484375,
+      "learning_rate": 9.751106194690266e-06,
+      "loss": 1.1367,
+      "step": 90
+    },
+    {
+      "epoch": 0.1006637168141593,
+      "grad_norm": 0.12890625,
+      "learning_rate": 9.748340707964604e-06,
+      "loss": 1.0047,
+      "step": 91
+    },
+    {
+      "epoch": 0.10176991150442478,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 9.745575221238938e-06,
+      "loss": 1.1659,
+      "step": 92
+    },
+    {
+      "epoch": 0.10287610619469026,
+      "grad_norm": 0.146484375,
+      "learning_rate": 9.742809734513276e-06,
+      "loss": 1.1399,
+      "step": 93
+    },
+    {
+      "epoch": 0.10398230088495575,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.740044247787612e-06,
+      "loss": 1.12,
+      "step": 94
+    },
+    {
+      "epoch": 0.10508849557522124,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.737278761061948e-06,
+      "loss": 1.1036,
+      "step": 95
+    },
+    {
+      "epoch": 0.10619469026548672,
+      "grad_norm": 0.134765625,
+      "learning_rate": 9.734513274336284e-06,
+      "loss": 1.1025,
+      "step": 96
+    },
+    {
+      "epoch": 0.10730088495575221,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.73174778761062e-06,
+      "loss": 1.1097,
+      "step": 97
+    },
+    {
+      "epoch": 0.1084070796460177,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 9.728982300884956e-06,
+      "loss": 1.0528,
+      "step": 98
+    },
+    {
+      "epoch": 0.10951327433628319,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.726216814159293e-06,
+      "loss": 1.1009,
+      "step": 99
+    },
+    {
+      "epoch": 0.11061946902654868,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 9.72345132743363e-06,
+      "loss": 1.1063,
+      "step": 100
+    },
+    {
+      "epoch": 0.11172566371681415,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 9.720685840707965e-06,
+      "loss": 1.0559,
+      "step": 101
+    },
+    {
+      "epoch": 0.11283185840707964,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.717920353982303e-06,
+      "loss": 1.2462,
+      "step": 102
+    },
+    {
+      "epoch": 0.11393805309734513,
+      "grad_norm": 0.134765625,
+      "learning_rate": 9.715154867256637e-06,
+      "loss": 1.1034,
+      "step": 103
+    },
+    {
+      "epoch": 0.11504424778761062,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.712389380530975e-06,
+      "loss": 1.1577,
+      "step": 104
+    },
+    {
+      "epoch": 0.1161504424778761,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 9.70962389380531e-06,
+      "loss": 1.0802,
+      "step": 105
+    },
+    {
+      "epoch": 0.1172566371681416,
+      "grad_norm": 0.12109375,
+      "learning_rate": 9.706858407079647e-06,
+      "loss": 1.0664,
+      "step": 106
+    },
+    {
+      "epoch": 0.11836283185840708,
+      "grad_norm": 0.12890625,
+      "learning_rate": 9.704092920353983e-06,
+      "loss": 1.0663,
+      "step": 107
+    },
+    {
+      "epoch": 0.11946902654867257,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 9.70132743362832e-06,
+      "loss": 1.0972,
+      "step": 108
+    },
+    {
+      "epoch": 0.12057522123893805,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 9.698561946902655e-06,
+      "loss": 1.1272,
+      "step": 109
+    },
+    {
+      "epoch": 0.12168141592920353,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.695796460176992e-06,
+      "loss": 1.06,
+      "step": 110
+    },
+    {
+      "epoch": 0.12278761061946902,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 9.693030973451328e-06,
+      "loss": 1.054,
+      "step": 111
+    },
+    {
+      "epoch": 0.12389380530973451,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 9.690265486725664e-06,
+      "loss": 1.1015,
+      "step": 112
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.6875e-06,
+      "loss": 1.13,
+      "step": 113
+    },
+    {
+      "epoch": 0.1261061946902655,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.684734513274338e-06,
+      "loss": 1.1012,
+      "step": 114
+    },
+    {
+      "epoch": 0.12721238938053098,
+      "grad_norm": 0.11328125,
+      "learning_rate": 9.681969026548674e-06,
+      "loss": 1.0631,
+      "step": 115
+    },
+    {
+      "epoch": 0.12831858407079647,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 9.67920353982301e-06,
+      "loss": 1.0734,
+      "step": 116
+    },
+    {
+      "epoch": 0.12942477876106195,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 9.676438053097346e-06,
+      "loss": 1.0716,
+      "step": 117
+    },
+    {
+      "epoch": 0.13053097345132744,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 9.673672566371682e-06,
+      "loss": 1.0607,
+      "step": 118
+    },
+    {
+      "epoch": 0.13163716814159293,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 9.670907079646018e-06,
+      "loss": 1.0575,
+      "step": 119
+    },
+    {
+      "epoch": 0.13274336283185842,
+      "grad_norm": 0.142578125,
+      "learning_rate": 9.668141592920355e-06,
+      "loss": 1.0845,
+      "step": 120
+    },
+    {
+      "epoch": 0.1338495575221239,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 9.665376106194692e-06,
+      "loss": 1.0203,
+      "step": 121
+    },
+    {
+      "epoch": 0.13495575221238937,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 9.662610619469027e-06,
+      "loss": 1.0951,
+      "step": 122
+    },
+    {
+      "epoch": 0.13606194690265486,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.659845132743365e-06,
+      "loss": 1.0298,
+      "step": 123
+    },
+    {
+      "epoch": 0.13716814159292035,
+      "grad_norm": 0.109375,
+      "learning_rate": 9.657079646017699e-06,
+      "loss": 1.0098,
+      "step": 124
+    },
+    {
+      "epoch": 0.13827433628318583,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.654314159292037e-06,
+      "loss": 0.9845,
+      "step": 125
+    },
+    {
+      "epoch": 0.13938053097345132,
+      "grad_norm": 0.119140625,
+      "learning_rate": 9.651548672566371e-06,
+      "loss": 1.0747,
+      "step": 126
+    },
+    {
+      "epoch": 0.1404867256637168,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 9.648783185840709e-06,
+      "loss": 1.0437,
+      "step": 127
+    },
+    {
+      "epoch": 0.1415929203539823,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 9.646017699115045e-06,
+      "loss": 1.0673,
+      "step": 128
+    },
+    {
+      "epoch": 0.1426991150442478,
+      "grad_norm": 0.126953125,
+      "learning_rate": 9.643252212389381e-06,
+      "loss": 1.0883,
+      "step": 129
+    },
+    {
+      "epoch": 0.14380530973451328,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 9.640486725663718e-06,
+      "loss": 1.0187,
+      "step": 130
+    },
+    {
+      "epoch": 0.14491150442477876,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 9.637721238938054e-06,
+      "loss": 0.9874,
+      "step": 131
+    },
+    {
+      "epoch": 0.14601769911504425,
+      "grad_norm": 0.11328125,
+      "learning_rate": 9.63495575221239e-06,
+      "loss": 1.0553,
+      "step": 132
+    },
+    {
+      "epoch": 0.14712389380530974,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 9.632190265486726e-06,
+      "loss": 0.9862,
+      "step": 133
+    },
+    {
+      "epoch": 0.14823008849557523,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 9.629424778761062e-06,
+      "loss": 1.0906,
+      "step": 134
+    },
+    {
+      "epoch": 0.14933628318584072,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 9.626659292035398e-06,
+      "loss": 0.9907,
+      "step": 135
+    },
+    {
+      "epoch": 0.1504424778761062,
+      "grad_norm": 0.123046875,
+      "learning_rate": 9.623893805309736e-06,
+      "loss": 1.0838,
+      "step": 136
+    },
+    {
+      "epoch": 0.1515486725663717,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 9.621128318584072e-06,
+      "loss": 1.0756,
+      "step": 137
+    },
+    {
+      "epoch": 0.15265486725663716,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 9.618362831858408e-06,
+      "loss": 1.0093,
+      "step": 138
+    },
+    {
+      "epoch": 0.15376106194690264,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 9.615597345132744e-06,
+      "loss": 0.9203,
+      "step": 139
+    },
+    {
+      "epoch": 0.15486725663716813,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 9.61283185840708e-06,
+      "loss": 1.02,
+      "step": 140
+    },
+    {
+      "epoch": 0.15597345132743362,
+      "grad_norm": 0.10546875,
+      "learning_rate": 9.610066371681417e-06,
+      "loss": 0.9911,
+      "step": 141
+    },
+    {
+      "epoch": 0.1570796460176991,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.607300884955753e-06,
+      "loss": 1.1058,
+      "step": 142
+    },
+    {
+      "epoch": 0.1581858407079646,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 9.604535398230089e-06,
+      "loss": 1.007,
+      "step": 143
+    },
+    {
+      "epoch": 0.1592920353982301,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 9.601769911504427e-06,
+      "loss": 1.1132,
+      "step": 144
+    },
+    {
+      "epoch": 0.16039823008849557,
+      "grad_norm": 0.107421875,
+      "learning_rate": 9.599004424778761e-06,
+      "loss": 1.0425,
+      "step": 145
+    },
+    {
+      "epoch": 0.16150442477876106,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 9.596238938053099e-06,
+      "loss": 1.045,
+      "step": 146
+    },
+    {
+      "epoch": 0.16261061946902655,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 9.593473451327433e-06,
+      "loss": 0.9718,
+      "step": 147
+    },
+    {
+      "epoch": 0.16371681415929204,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 9.590707964601771e-06,
+      "loss": 0.9833,
+      "step": 148
+    },
+    {
+      "epoch": 0.16482300884955753,
+      "grad_norm": 0.11328125,
+      "learning_rate": 9.587942477876107e-06,
+      "loss": 1.042,
+      "step": 149
+    },
+    {
+      "epoch": 0.16592920353982302,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 9.585176991150443e-06,
+      "loss": 0.9759,
+      "step": 150
+    },
+    {
+      "epoch": 0.1670353982300885,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 9.58241150442478e-06,
+      "loss": 1.0018,
+      "step": 151
+    },
+    {
+      "epoch": 0.168141592920354,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 9.579646017699116e-06,
+      "loss": 0.9685,
+      "step": 152
+    },
+    {
+      "epoch": 0.16924778761061948,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 9.576880530973452e-06,
+      "loss": 1.016,
+      "step": 153
+    },
+    {
+      "epoch": 0.17035398230088494,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 9.574115044247788e-06,
+      "loss": 1.018,
+      "step": 154
+    },
+    {
+      "epoch": 0.17146017699115043,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.571349557522124e-06,
+      "loss": 0.9577,
+      "step": 155
+    },
+    {
+      "epoch": 0.17256637168141592,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 9.56858407079646e-06,
+      "loss": 0.9991,
+      "step": 156
+    },
+    {
+      "epoch": 0.1736725663716814,
+      "grad_norm": 0.095703125,
+      "learning_rate": 9.565818584070798e-06,
+      "loss": 0.9418,
+      "step": 157
+    },
+    {
+      "epoch": 0.1747787610619469,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 9.563053097345132e-06,
+      "loss": 0.9854,
+      "step": 158
+    },
+    {
+      "epoch": 0.17588495575221239,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 9.56028761061947e-06,
+      "loss": 0.9415,
+      "step": 159
+    },
+    {
+      "epoch": 0.17699115044247787,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 9.557522123893806e-06,
+      "loss": 0.9652,
+      "step": 160
+    },
+    {
+      "epoch": 0.17809734513274336,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 9.554756637168143e-06,
+      "loss": 0.9492,
+      "step": 161
+    },
+    {
+      "epoch": 0.17920353982300885,
+      "grad_norm": 0.10546875,
+      "learning_rate": 9.551991150442479e-06,
+      "loss": 0.992,
+      "step": 162
+    },
+    {
+      "epoch": 0.18030973451327434,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.549225663716815e-06,
+      "loss": 1.0606,
+      "step": 163
+    },
+    {
+      "epoch": 0.18141592920353983,
+      "grad_norm": 0.09765625,
+      "learning_rate": 9.546460176991151e-06,
+      "loss": 0.9399,
+      "step": 164
+    },
+    {
+      "epoch": 0.18252212389380532,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 9.543694690265487e-06,
+      "loss": 0.9926,
+      "step": 165
+    },
+    {
+      "epoch": 0.1836283185840708,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 9.540929203539823e-06,
+      "loss": 0.9929,
+      "step": 166
+    },
+    {
+      "epoch": 0.1847345132743363,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 9.538163716814161e-06,
+      "loss": 0.9743,
+      "step": 167
+    },
+    {
+      "epoch": 0.18584070796460178,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 9.535398230088495e-06,
+      "loss": 0.9497,
+      "step": 168
+    },
+    {
+      "epoch": 0.18694690265486727,
+      "grad_norm": 0.12890625,
+      "learning_rate": 9.532632743362833e-06,
+      "loss": 0.9593,
+      "step": 169
+    },
+    {
+      "epoch": 0.18805309734513273,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 9.52986725663717e-06,
+      "loss": 0.9257,
+      "step": 170
+    },
+    {
+      "epoch": 0.18915929203539822,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 9.527101769911505e-06,
+      "loss": 0.9738,
+      "step": 171
+    },
+    {
+      "epoch": 0.1902654867256637,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 9.524336283185842e-06,
+      "loss": 1.0004,
+      "step": 172
+    },
+    {
+      "epoch": 0.1913716814159292,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 9.521570796460178e-06,
+      "loss": 0.987,
+      "step": 173
+    },
+    {
+      "epoch": 0.19247787610619468,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 9.518805309734514e-06,
+      "loss": 0.9239,
+      "step": 174
+    },
+    {
+      "epoch": 0.19358407079646017,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 9.51603982300885e-06,
+      "loss": 0.9693,
+      "step": 175
+    },
+    {
+      "epoch": 0.19469026548672566,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 9.513274336283188e-06,
+      "loss": 1.0253,
+      "step": 176
+    },
+    {
+      "epoch": 0.19579646017699115,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 9.510508849557522e-06,
+      "loss": 1.0259,
+      "step": 177
+    },
+    {
+      "epoch": 0.19690265486725664,
+      "grad_norm": 0.1328125,
+      "learning_rate": 9.50774336283186e-06,
+      "loss": 1.045,
+      "step": 178
+    },
+    {
+      "epoch": 0.19800884955752213,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.504977876106194e-06,
+      "loss": 0.939,
+      "step": 179
+    },
+    {
+      "epoch": 0.19911504424778761,
+      "grad_norm": 0.125,
+      "learning_rate": 9.502212389380532e-06,
+      "loss": 1.0235,
+      "step": 180
+    },
+    {
+      "epoch": 0.2002212389380531,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 9.499446902654868e-06,
+      "loss": 0.9805,
+      "step": 181
+    },
+    {
+      "epoch": 0.2013274336283186,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.496681415929205e-06,
+      "loss": 0.9111,
+      "step": 182
+    },
+    {
+      "epoch": 0.20243362831858408,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 9.49391592920354e-06,
+      "loss": 0.98,
+      "step": 183
+    },
+    {
+      "epoch": 0.20353982300884957,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.491150442477877e-06,
+      "loss": 0.97,
+      "step": 184
+    },
+    {
+      "epoch": 0.20464601769911506,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 9.488384955752213e-06,
+      "loss": 1.0103,
+      "step": 185
+    },
+    {
+      "epoch": 0.20575221238938052,
+      "grad_norm": 0.099609375,
+      "learning_rate": 9.485619469026549e-06,
+      "loss": 0.9824,
+      "step": 186
+    },
+    {
+      "epoch": 0.206858407079646,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 9.482853982300885e-06,
+      "loss": 0.9489,
+      "step": 187
+    },
+    {
+      "epoch": 0.2079646017699115,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.480088495575221e-06,
+      "loss": 0.9815,
+      "step": 188
+    },
+    {
+      "epoch": 0.20907079646017698,
+      "grad_norm": 0.1015625,
+      "learning_rate": 9.477323008849557e-06,
+      "loss": 0.9694,
+      "step": 189
+    },
+    {
+      "epoch": 0.21017699115044247,
+      "grad_norm": 0.10546875,
+      "learning_rate": 9.474557522123895e-06,
+      "loss": 0.9392,
+      "step": 190
+    },
+    {
+      "epoch": 0.21128318584070796,
+      "grad_norm": 0.08984375,
+      "learning_rate": 9.471792035398231e-06,
+      "loss": 0.9339,
+      "step": 191
+    },
+    {
+      "epoch": 0.21238938053097345,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 9.469026548672568e-06,
+      "loss": 0.9331,
+      "step": 192
+    },
+    {
+      "epoch": 0.21349557522123894,
+      "grad_norm": 0.091796875,
+      "learning_rate": 9.466261061946904e-06,
+      "loss": 0.898,
+      "step": 193
+    },
+    {
+      "epoch": 0.21460176991150443,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 9.46349557522124e-06,
+      "loss": 0.8954,
+      "step": 194
+    },
+    {
+      "epoch": 0.2157079646017699,
+      "grad_norm": 0.09765625,
+      "learning_rate": 9.460730088495576e-06,
+      "loss": 0.933,
+      "step": 195
+    },
+    {
+      "epoch": 0.2168141592920354,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 9.457964601769912e-06,
+      "loss": 0.9354,
+      "step": 196
+    },
+    {
+      "epoch": 0.2179203539823009,
+      "grad_norm": 0.109375,
+      "learning_rate": 9.45519911504425e-06,
+      "loss": 0.9548,
+      "step": 197
+    },
+    {
+      "epoch": 0.21902654867256638,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 9.452433628318584e-06,
+      "loss": 0.9656,
+      "step": 198
+    },
+    {
+      "epoch": 0.22013274336283187,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 9.449668141592922e-06,
+      "loss": 1.0013,
+      "step": 199
+    },
+    {
+      "epoch": 0.22123893805309736,
+      "grad_norm": 0.09375,
+      "learning_rate": 9.446902654867257e-06,
+      "loss": 0.9249,
+      "step": 200
+    },
+    {
+      "epoch": 0.22234513274336284,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 9.444137168141594e-06,
+      "loss": 0.887,
+      "step": 201
+    },
+    {
+      "epoch": 0.2234513274336283,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 9.441371681415929e-06,
+      "loss": 0.9147,
+      "step": 202
+    },
+    {
+      "epoch": 0.2245575221238938,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 9.438606194690267e-06,
+      "loss": 0.9193,
+      "step": 203
+    },
+    {
+      "epoch": 0.22566371681415928,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 9.435840707964603e-06,
+      "loss": 0.9591,
+      "step": 204
+    },
+    {
+      "epoch": 0.22676991150442477,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 9.433075221238939e-06,
+      "loss": 0.9349,
+      "step": 205
+    },
+    {
+      "epoch": 0.22787610619469026,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.430309734513275e-06,
+      "loss": 0.936,
+      "step": 206
+    },
+    {
+      "epoch": 0.22898230088495575,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.427544247787611e-06,
+      "loss": 0.9086,
+      "step": 207
+    },
+    {
+      "epoch": 0.23008849557522124,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 9.424778761061947e-06,
+      "loss": 0.9234,
+      "step": 208
+    },
+    {
+      "epoch": 0.23119469026548672,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.422013274336283e-06,
+      "loss": 0.9718,
+      "step": 209
+    },
+    {
+      "epoch": 0.2323008849557522,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 9.41924778761062e-06,
+      "loss": 0.9336,
+      "step": 210
+    },
+    {
+      "epoch": 0.2334070796460177,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.416482300884957e-06,
+      "loss": 0.878,
+      "step": 211
+    },
+    {
+      "epoch": 0.2345132743362832,
+      "grad_norm": 0.095703125,
+      "learning_rate": 9.413716814159293e-06,
+      "loss": 0.9051,
+      "step": 212
+    },
+    {
+      "epoch": 0.23561946902654868,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 9.41095132743363e-06,
+      "loss": 0.8874,
+      "step": 213
+    },
+    {
+      "epoch": 0.23672566371681417,
+      "grad_norm": 0.125,
+      "learning_rate": 9.408185840707966e-06,
+      "loss": 0.9931,
+      "step": 214
+    },
+    {
+      "epoch": 0.23783185840707965,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 9.405420353982302e-06,
+      "loss": 0.8991,
+      "step": 215
+    },
+    {
+      "epoch": 0.23893805309734514,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.402654867256638e-06,
+      "loss": 0.967,
+      "step": 216
+    },
+    {
+      "epoch": 0.24004424778761063,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 9.399889380530974e-06,
+      "loss": 0.9487,
+      "step": 217
+    },
+    {
+      "epoch": 0.2411504424778761,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 9.39712389380531e-06,
+      "loss": 0.8811,
+      "step": 218
+    },
+    {
+      "epoch": 0.24225663716814158,
+      "grad_norm": 0.09765625,
+      "learning_rate": 9.394358407079646e-06,
+      "loss": 0.9777,
+      "step": 219
+    },
+    {
+      "epoch": 0.24336283185840707,
+      "grad_norm": 0.126953125,
+      "learning_rate": 9.391592920353984e-06,
+      "loss": 0.9486,
+      "step": 220
+    },
+    {
+      "epoch": 0.24446902654867256,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 9.388827433628319e-06,
+      "loss": 0.9567,
+      "step": 221
+    },
+    {
+      "epoch": 0.24557522123893805,
+      "grad_norm": 0.091796875,
+      "learning_rate": 9.386061946902656e-06,
+      "loss": 0.9298,
+      "step": 222
+    },
+    {
+      "epoch": 0.24668141592920353,
+      "grad_norm": 0.080078125,
+      "learning_rate": 9.38329646017699e-06,
+      "loss": 0.8418,
+      "step": 223
+    },
+    {
+      "epoch": 0.24778761061946902,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 9.380530973451329e-06,
+      "loss": 0.9089,
+      "step": 224
+    },
+    {
+      "epoch": 0.2488938053097345,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 9.377765486725665e-06,
+      "loss": 0.9455,
+      "step": 225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.9137,
+      "step": 226
+    },
+    {
+      "epoch": 0.25110619469026546,
+      "grad_norm": 0.123046875,
+      "learning_rate": 9.372234513274337e-06,
+      "loss": 0.9315,
+      "step": 227
+    },
+    {
+      "epoch": 0.252212389380531,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 9.369469026548673e-06,
+      "loss": 0.9321,
+      "step": 228
+    },
+    {
+      "epoch": 0.25331858407079644,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.36670353982301e-06,
+      "loss": 0.9468,
+      "step": 229
+    },
+    {
+      "epoch": 0.25442477876106195,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 9.363938053097345e-06,
+      "loss": 0.9156,
+      "step": 230
+    },
+    {
+      "epoch": 0.2555309734513274,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 9.361172566371683e-06,
+      "loss": 0.8977,
+      "step": 231
+    },
+    {
+      "epoch": 0.25663716814159293,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 9.358407079646018e-06,
+      "loss": 0.9556,
+      "step": 232
+    },
+    {
+      "epoch": 0.2577433628318584,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 9.355641592920355e-06,
+      "loss": 0.9364,
+      "step": 233
+    },
+    {
+      "epoch": 0.2588495575221239,
+      "grad_norm": 0.1015625,
+      "learning_rate": 9.352876106194692e-06,
+      "loss": 0.9383,
+      "step": 234
+    },
+    {
+      "epoch": 0.25995575221238937,
+      "grad_norm": 0.08984375,
+      "learning_rate": 9.350110619469028e-06,
+      "loss": 0.8955,
+      "step": 235
+    },
+    {
+      "epoch": 0.2610619469026549,
+      "grad_norm": 0.095703125,
+      "learning_rate": 9.347345132743364e-06,
+      "loss": 0.8868,
+      "step": 236
+    },
+    {
+      "epoch": 0.26216814159292035,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 9.3445796460177e-06,
+      "loss": 0.9211,
+      "step": 237
+    },
+    {
+      "epoch": 0.26327433628318586,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 9.341814159292036e-06,
+      "loss": 0.9522,
+      "step": 238
+    },
+    {
+      "epoch": 0.2643805309734513,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 9.339048672566372e-06,
+      "loss": 0.8987,
+      "step": 239
+    },
+    {
+      "epoch": 0.26548672566371684,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.336283185840708e-06,
+      "loss": 0.8889,
+      "step": 240
+    },
+    {
+      "epoch": 0.2665929203539823,
+      "grad_norm": 0.09375,
+      "learning_rate": 9.333517699115044e-06,
+      "loss": 0.8881,
+      "step": 241
+    },
+    {
+      "epoch": 0.2676991150442478,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.33075221238938e-06,
+      "loss": 0.9134,
+      "step": 242
+    },
+    {
+      "epoch": 0.2688053097345133,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 9.327986725663718e-06,
+      "loss": 0.9334,
+      "step": 243
+    },
+    {
+      "epoch": 0.26991150442477874,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 9.325221238938053e-06,
+      "loss": 0.8891,
+      "step": 244
+    },
+    {
+      "epoch": 0.27101769911504425,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 9.32245575221239e-06,
+      "loss": 0.9231,
+      "step": 245
+    },
+    {
+      "epoch": 0.2721238938053097,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 9.319690265486727e-06,
+      "loss": 0.9043,
+      "step": 246
+    },
+    {
+      "epoch": 0.27323008849557523,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.316924778761063e-06,
+      "loss": 0.9049,
+      "step": 247
+    },
+    {
+      "epoch": 0.2743362831858407,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 9.314159292035399e-06,
+      "loss": 0.8908,
+      "step": 248
+    },
+    {
+      "epoch": 0.2754424778761062,
+      "grad_norm": 0.109375,
+      "learning_rate": 9.311393805309735e-06,
+      "loss": 0.9422,
+      "step": 249
+    },
+    {
+      "epoch": 0.27654867256637167,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 9.308628318584071e-06,
+      "loss": 0.8899,
+      "step": 250
+    },
+    {
+      "epoch": 0.2776548672566372,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 9.305862831858407e-06,
+      "loss": 0.8903,
+      "step": 251
+    },
+    {
+      "epoch": 0.27876106194690264,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 9.303097345132745e-06,
+      "loss": 0.9352,
+      "step": 252
+    },
+    {
+      "epoch": 0.27986725663716816,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 9.30033185840708e-06,
+      "loss": 0.9226,
+      "step": 253
+    },
+    {
+      "epoch": 0.2809734513274336,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 9.297566371681417e-06,
+      "loss": 0.8765,
+      "step": 254
+    },
+    {
+      "epoch": 0.28207964601769914,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 9.294800884955752e-06,
+      "loss": 0.8705,
+      "step": 255
+    },
+    {
+      "epoch": 0.2831858407079646,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 9.29203539823009e-06,
+      "loss": 0.9338,
+      "step": 256
+    },
+    {
+      "epoch": 0.2842920353982301,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 9.289269911504426e-06,
+      "loss": 0.8753,
+      "step": 257
+    },
+    {
+      "epoch": 0.2853982300884956,
+      "grad_norm": 0.08203125,
+      "learning_rate": 9.286504424778762e-06,
+      "loss": 0.8488,
+      "step": 258
+    },
+    {
+      "epoch": 0.28650442477876104,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 9.283738938053098e-06,
+      "loss": 0.9031,
+      "step": 259
+    },
+    {
+      "epoch": 0.28761061946902655,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 9.280973451327434e-06,
+      "loss": 0.8952,
+      "step": 260
+    },
+    {
+      "epoch": 0.288716814159292,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 9.27820796460177e-06,
+      "loss": 0.9301,
+      "step": 261
+    },
+    {
+      "epoch": 0.28982300884955753,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.275442477876106e-06,
+      "loss": 0.9393,
+      "step": 262
+    },
+    {
+      "epoch": 0.290929203539823,
+      "grad_norm": 0.08984375,
+      "learning_rate": 9.272676991150443e-06,
+      "loss": 0.8957,
+      "step": 263
+    },
+    {
+      "epoch": 0.2920353982300885,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.26991150442478e-06,
+      "loss": 0.981,
+      "step": 264
+    },
+    {
+      "epoch": 0.29314159292035397,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 9.267146017699115e-06,
+      "loss": 0.8882,
+      "step": 265
+    },
+    {
+      "epoch": 0.2942477876106195,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.264380530973453e-06,
+      "loss": 0.9222,
+      "step": 266
+    },
+    {
+      "epoch": 0.29535398230088494,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 9.261615044247789e-06,
+      "loss": 0.8825,
+      "step": 267
+    },
+    {
+      "epoch": 0.29646017699115046,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 9.258849557522125e-06,
+      "loss": 0.8807,
+      "step": 268
+    },
+    {
+      "epoch": 0.2975663716814159,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 9.256084070796461e-06,
+      "loss": 0.8961,
+      "step": 269
+    },
+    {
+      "epoch": 0.29867256637168144,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 9.253318584070797e-06,
+      "loss": 0.8597,
+      "step": 270
+    },
+    {
+      "epoch": 0.2997787610619469,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 9.250553097345133e-06,
+      "loss": 0.9362,
+      "step": 271
+    },
+    {
+      "epoch": 0.3008849557522124,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 9.24778761061947e-06,
+      "loss": 0.8848,
+      "step": 272
+    },
+    {
+      "epoch": 0.3019911504424779,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 9.245022123893807e-06,
+      "loss": 0.8649,
+      "step": 273
+    },
+    {
+      "epoch": 0.3030973451327434,
+      "grad_norm": 0.11328125,
+      "learning_rate": 9.242256637168142e-06,
+      "loss": 0.9254,
+      "step": 274
+    },
+    {
+      "epoch": 0.30420353982300885,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 9.23949115044248e-06,
+      "loss": 0.8977,
+      "step": 275
+    },
+    {
+      "epoch": 0.3053097345132743,
+      "grad_norm": 0.1015625,
+      "learning_rate": 9.236725663716814e-06,
+      "loss": 0.8841,
+      "step": 276
+    },
+    {
+      "epoch": 0.3064159292035398,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 9.233960176991152e-06,
+      "loss": 0.887,
+      "step": 277
+    },
+    {
+      "epoch": 0.3075221238938053,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 9.231194690265486e-06,
+      "loss": 0.8988,
+      "step": 278
+    },
+    {
+      "epoch": 0.3086283185840708,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 9.228429203539824e-06,
+      "loss": 0.8823,
+      "step": 279
+    },
+    {
+      "epoch": 0.30973451327433627,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 9.22566371681416e-06,
+      "loss": 0.9079,
+      "step": 280
+    },
+    {
+      "epoch": 0.3108407079646018,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 9.222898230088496e-06,
+      "loss": 0.8576,
+      "step": 281
+    },
+    {
+      "epoch": 0.31194690265486724,
+      "grad_norm": 0.091796875,
+      "learning_rate": 9.220132743362832e-06,
+      "loss": 0.8819,
+      "step": 282
+    },
+    {
+      "epoch": 0.31305309734513276,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 9.217367256637169e-06,
+      "loss": 0.8883,
+      "step": 283
+    },
+    {
+      "epoch": 0.3141592920353982,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 9.214601769911505e-06,
+      "loss": 0.8942,
+      "step": 284
+    },
+    {
+      "epoch": 0.31526548672566373,
+      "grad_norm": 0.095703125,
+      "learning_rate": 9.21183628318584e-06,
+      "loss": 0.8416,
+      "step": 285
+    },
+    {
+      "epoch": 0.3163716814159292,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 9.209070796460179e-06,
+      "loss": 0.834,
+      "step": 286
+    },
+    {
+      "epoch": 0.3174778761061947,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 9.206305309734515e-06,
+      "loss": 0.9043,
+      "step": 287
+    },
+    {
+      "epoch": 0.3185840707964602,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 9.203539823008851e-06,
+      "loss": 0.9304,
+      "step": 288
+    },
+    {
+      "epoch": 0.3196902654867257,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 9.200774336283187e-06,
+      "loss": 0.8676,
+      "step": 289
+    },
+    {
+      "epoch": 0.32079646017699115,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 9.198008849557523e-06,
+      "loss": 0.8873,
+      "step": 290
+    },
+    {
+      "epoch": 0.3219026548672566,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 9.19524336283186e-06,
+      "loss": 0.8884,
+      "step": 291
+    },
+    {
+      "epoch": 0.3230088495575221,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 9.192477876106195e-06,
+      "loss": 0.9317,
+      "step": 292
+    },
+    {
+      "epoch": 0.3241150442477876,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 9.189712389380531e-06,
+      "loss": 0.8607,
+      "step": 293
+    },
+    {
+      "epoch": 0.3252212389380531,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.18694690265487e-06,
+      "loss": 0.9487,
+      "step": 294
+    },
+    {
+      "epoch": 0.32632743362831856,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 9.184181415929204e-06,
+      "loss": 0.8641,
+      "step": 295
+    },
+    {
+      "epoch": 0.3274336283185841,
+      "grad_norm": 0.126953125,
+      "learning_rate": 9.181415929203542e-06,
+      "loss": 0.8736,
+      "step": 296
+    },
+    {
+      "epoch": 0.32853982300884954,
+      "grad_norm": 0.08984375,
+      "learning_rate": 9.178650442477876e-06,
+      "loss": 0.875,
+      "step": 297
+    },
+    {
+      "epoch": 0.32964601769911506,
+      "grad_norm": 0.087890625,
+      "learning_rate": 9.175884955752214e-06,
+      "loss": 0.8868,
+      "step": 298
+    },
+    {
+      "epoch": 0.3307522123893805,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 9.173119469026548e-06,
+      "loss": 0.8791,
+      "step": 299
+    },
+    {
+      "epoch": 0.33185840707964603,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 9.170353982300886e-06,
+      "loss": 0.886,
+      "step": 300
+    },
+    {
+      "epoch": 0.3329646017699115,
+      "grad_norm": 0.107421875,
+      "learning_rate": 9.167588495575222e-06,
+      "loss": 0.871,
+      "step": 301
+    },
+    {
+      "epoch": 0.334070796460177,
+      "grad_norm": 0.08984375,
+      "learning_rate": 9.164823008849558e-06,
+      "loss": 0.8676,
+      "step": 302
+    },
+    {
+      "epoch": 0.33517699115044247,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.162057522123894e-06,
+      "loss": 0.9212,
+      "step": 303
+    },
+    {
+      "epoch": 0.336283185840708,
+      "grad_norm": 0.107421875,
+      "learning_rate": 9.15929203539823e-06,
+      "loss": 0.9535,
+      "step": 304
+    },
+    {
+      "epoch": 0.33738938053097345,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 9.156526548672567e-06,
+      "loss": 0.8449,
+      "step": 305
+    },
+    {
+      "epoch": 0.33849557522123896,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 9.153761061946903e-06,
+      "loss": 0.9084,
+      "step": 306
+    },
+    {
+      "epoch": 0.3396017699115044,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 9.15099557522124e-06,
+      "loss": 0.8582,
+      "step": 307
+    },
+    {
+      "epoch": 0.3407079646017699,
+      "grad_norm": 0.109375,
+      "learning_rate": 9.148230088495575e-06,
+      "loss": 0.8501,
+      "step": 308
+    },
+    {
+      "epoch": 0.3418141592920354,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 9.145464601769913e-06,
+      "loss": 0.8705,
+      "step": 309
+    },
+    {
+      "epoch": 0.34292035398230086,
+      "grad_norm": 0.095703125,
+      "learning_rate": 9.142699115044249e-06,
+      "loss": 0.8562,
+      "step": 310
+    },
+    {
+      "epoch": 0.3440265486725664,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 9.139933628318585e-06,
+      "loss": 0.9101,
+      "step": 311
+    },
+    {
+      "epoch": 0.34513274336283184,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 9.137168141592921e-06,
+      "loss": 0.8588,
+      "step": 312
+    },
+    {
+      "epoch": 0.34623893805309736,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.134402654867257e-06,
+      "loss": 0.9012,
+      "step": 313
+    },
+    {
+      "epoch": 0.3473451327433628,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 9.131637168141594e-06,
+      "loss": 0.8713,
+      "step": 314
+    },
+    {
+      "epoch": 0.34845132743362833,
+      "grad_norm": 0.109375,
+      "learning_rate": 9.12887168141593e-06,
+      "loss": 0.8954,
+      "step": 315
+    },
+    {
+      "epoch": 0.3495575221238938,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 9.126106194690266e-06,
+      "loss": 0.8791,
+      "step": 316
+    },
+    {
+      "epoch": 0.3506637168141593,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 9.123340707964604e-06,
+      "loss": 0.8327,
+      "step": 317
+    },
+    {
+      "epoch": 0.35176991150442477,
+      "grad_norm": 0.091796875,
+      "learning_rate": 9.120575221238938e-06,
+      "loss": 0.8861,
+      "step": 318
+    },
+    {
+      "epoch": 0.3528761061946903,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 9.117809734513276e-06,
+      "loss": 0.9875,
+      "step": 319
+    },
+    {
+      "epoch": 0.35398230088495575,
+      "grad_norm": 0.1015625,
+      "learning_rate": 9.11504424778761e-06,
+      "loss": 0.9193,
+      "step": 320
+    },
+    {
+      "epoch": 0.35508849557522126,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.112278761061948e-06,
+      "loss": 0.9456,
+      "step": 321
+    },
+    {
+      "epoch": 0.3561946902654867,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.109513274336284e-06,
+      "loss": 0.8812,
+      "step": 322
+    },
+    {
+      "epoch": 0.3573008849557522,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 9.10674778761062e-06,
+      "loss": 0.9022,
+      "step": 323
+    },
+    {
+      "epoch": 0.3584070796460177,
+      "grad_norm": 0.0859375,
+      "learning_rate": 9.103982300884956e-06,
+      "loss": 0.8448,
+      "step": 324
+    },
+    {
+      "epoch": 0.35951327433628316,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.101216814159293e-06,
+      "loss": 0.8966,
+      "step": 325
+    },
+    {
+      "epoch": 0.3606194690265487,
+      "grad_norm": 0.0859375,
+      "learning_rate": 9.098451327433629e-06,
+      "loss": 0.8501,
+      "step": 326
+    },
+    {
+      "epoch": 0.36172566371681414,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 9.095685840707965e-06,
+      "loss": 0.8781,
+      "step": 327
+    },
+    {
+      "epoch": 0.36283185840707965,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 9.092920353982303e-06,
+      "loss": 0.8562,
+      "step": 328
+    },
+    {
+      "epoch": 0.3639380530973451,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 9.090154867256637e-06,
+      "loss": 0.8992,
+      "step": 329
+    },
+    {
+      "epoch": 0.36504424778761063,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 9.087389380530975e-06,
+      "loss": 0.8555,
+      "step": 330
+    },
+    {
+      "epoch": 0.3661504424778761,
+      "grad_norm": 0.0859375,
+      "learning_rate": 9.08462389380531e-06,
+      "loss": 0.8258,
+      "step": 331
+    },
+    {
+      "epoch": 0.3672566371681416,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.081858407079647e-06,
+      "loss": 0.9688,
+      "step": 332
+    },
+    {
+      "epoch": 0.36836283185840707,
+      "grad_norm": 0.099609375,
+      "learning_rate": 9.079092920353983e-06,
+      "loss": 0.8857,
+      "step": 333
+    },
+    {
+      "epoch": 0.3694690265486726,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 9.07632743362832e-06,
+      "loss": 0.9372,
+      "step": 334
+    },
+    {
+      "epoch": 0.37057522123893805,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 9.073561946902656e-06,
+      "loss": 0.9147,
+      "step": 335
+    },
+    {
+      "epoch": 0.37168141592920356,
+      "grad_norm": 0.08349609375,
+      "learning_rate": 9.070796460176992e-06,
+      "loss": 0.8448,
+      "step": 336
+    },
+    {
+      "epoch": 0.372787610619469,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 9.068030973451328e-06,
+      "loss": 0.8456,
+      "step": 337
+    },
+    {
+      "epoch": 0.37389380530973454,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 9.065265486725664e-06,
+      "loss": 0.8809,
+      "step": 338
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 9.0625e-06,
+      "loss": 0.8957,
+      "step": 339
+    },
+    {
+      "epoch": 0.37610619469026546,
+      "grad_norm": 0.08984375,
+      "learning_rate": 9.059734513274338e-06,
+      "loss": 0.8569,
+      "step": 340
+    },
+    {
+      "epoch": 0.377212389380531,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 9.056969026548674e-06,
+      "loss": 0.853,
+      "step": 341
+    },
+    {
+      "epoch": 0.37831858407079644,
+      "grad_norm": 0.111328125,
+      "learning_rate": 9.05420353982301e-06,
+      "loss": 0.8779,
+      "step": 342
+    },
+    {
+      "epoch": 0.37942477876106195,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.051438053097346e-06,
+      "loss": 0.9066,
+      "step": 343
+    },
+    {
+      "epoch": 0.3805309734513274,
+      "grad_norm": 0.115234375,
+      "learning_rate": 9.048672566371682e-06,
+      "loss": 0.8541,
+      "step": 344
+    },
+    {
+      "epoch": 0.38163716814159293,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 9.045907079646019e-06,
+      "loss": 0.9062,
+      "step": 345
+    },
+    {
+      "epoch": 0.3827433628318584,
+      "grad_norm": 0.0859375,
+      "learning_rate": 9.043141592920355e-06,
+      "loss": 0.8774,
+      "step": 346
+    },
+    {
+      "epoch": 0.3838495575221239,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 9.04037610619469e-06,
+      "loss": 0.8294,
+      "step": 347
+    },
+    {
+      "epoch": 0.38495575221238937,
+      "grad_norm": 0.099609375,
+      "learning_rate": 9.037610619469027e-06,
+      "loss": 0.8268,
+      "step": 348
+    },
+    {
+      "epoch": 0.3860619469026549,
+      "grad_norm": 0.095703125,
+      "learning_rate": 9.034845132743365e-06,
+      "loss": 0.8794,
+      "step": 349
+    },
+    {
+      "epoch": 0.38716814159292035,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 9.032079646017699e-06,
+      "loss": 0.8497,
+      "step": 350
+    },
+    {
+      "epoch": 0.38827433628318586,
+      "grad_norm": 0.09765625,
+      "learning_rate": 9.029314159292037e-06,
+      "loss": 0.8492,
+      "step": 351
+    },
+    {
+      "epoch": 0.3893805309734513,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 9.026548672566371e-06,
+      "loss": 0.8651,
+      "step": 352
+    },
+    {
+      "epoch": 0.39048672566371684,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 9.02378318584071e-06,
+      "loss": 0.8738,
+      "step": 353
+    },
+    {
+      "epoch": 0.3915929203539823,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.021017699115044e-06,
+      "loss": 0.9052,
+      "step": 354
+    },
+    {
+      "epoch": 0.3926991150442478,
+      "grad_norm": 0.07861328125,
+      "learning_rate": 9.018252212389381e-06,
+      "loss": 0.8124,
+      "step": 355
+    },
+    {
+      "epoch": 0.3938053097345133,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 9.015486725663718e-06,
+      "loss": 0.8442,
+      "step": 356
+    },
+    {
+      "epoch": 0.39491150442477874,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.012721238938054e-06,
+      "loss": 0.8475,
+      "step": 357
+    },
+    {
+      "epoch": 0.39601769911504425,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 9.00995575221239e-06,
+      "loss": 0.8577,
+      "step": 358
+    },
+    {
+      "epoch": 0.3971238938053097,
+      "grad_norm": 0.09765625,
+      "learning_rate": 9.007190265486726e-06,
+      "loss": 0.8789,
+      "step": 359
+    },
+    {
+      "epoch": 0.39823008849557523,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 9.004424778761062e-06,
+      "loss": 0.7857,
+      "step": 360
+    },
+    {
+      "epoch": 0.3993362831858407,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 9.001659292035398e-06,
+      "loss": 0.8318,
+      "step": 361
+    },
+    {
+      "epoch": 0.4004424778761062,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 8.998893805309736e-06,
+      "loss": 0.8693,
+      "step": 362
+    },
+    {
+      "epoch": 0.40154867256637167,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.996128318584072e-06,
+      "loss": 0.8363,
+      "step": 363
+    },
+    {
+      "epoch": 0.4026548672566372,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.993362831858408e-06,
+      "loss": 0.8348,
+      "step": 364
+    },
+    {
+      "epoch": 0.40376106194690264,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 8.990597345132744e-06,
+      "loss": 0.8321,
+      "step": 365
+    },
+    {
+      "epoch": 0.40486725663716816,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 8.98783185840708e-06,
+      "loss": 0.9001,
+      "step": 366
+    },
+    {
+      "epoch": 0.4059734513274336,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 8.985066371681417e-06,
+      "loss": 0.8371,
+      "step": 367
+    },
+    {
+      "epoch": 0.40707964601769914,
+      "grad_norm": 0.083984375,
+      "learning_rate": 8.982300884955753e-06,
+      "loss": 0.8114,
+      "step": 368
+    },
+    {
+      "epoch": 0.4081858407079646,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 8.979535398230089e-06,
+      "loss": 0.8845,
+      "step": 369
+    },
+    {
+      "epoch": 0.4092920353982301,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.976769911504427e-06,
+      "loss": 0.8328,
+      "step": 370
+    },
+    {
+      "epoch": 0.4103982300884956,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 8.974004424778761e-06,
+      "loss": 0.8149,
+      "step": 371
+    },
+    {
+      "epoch": 0.41150442477876104,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 8.971238938053099e-06,
+      "loss": 0.8859,
+      "step": 372
+    },
+    {
+      "epoch": 0.41261061946902655,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.968473451327433e-06,
+      "loss": 0.8228,
+      "step": 373
+    },
+    {
+      "epoch": 0.413716814159292,
+      "grad_norm": 0.083984375,
+      "learning_rate": 8.965707964601771e-06,
+      "loss": 0.8487,
+      "step": 374
+    },
+    {
+      "epoch": 0.41482300884955753,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 8.962942477876106e-06,
+      "loss": 0.7856,
+      "step": 375
+    },
+    {
+      "epoch": 0.415929203539823,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.960176991150443e-06,
+      "loss": 0.8482,
+      "step": 376
+    },
+    {
+      "epoch": 0.4170353982300885,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 8.95741150442478e-06,
+      "loss": 0.8308,
+      "step": 377
+    },
+    {
+      "epoch": 0.41814159292035397,
+      "grad_norm": 0.072265625,
+      "learning_rate": 8.954646017699116e-06,
+      "loss": 0.7822,
+      "step": 378
+    },
+    {
+      "epoch": 0.4192477876106195,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.951880530973452e-06,
+      "loss": 0.8218,
+      "step": 379
+    },
+    {
+      "epoch": 0.42035398230088494,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.949115044247788e-06,
+      "loss": 0.8155,
+      "step": 380
+    },
+    {
+      "epoch": 0.42146017699115046,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.946349557522124e-06,
+      "loss": 0.8144,
+      "step": 381
+    },
+    {
+      "epoch": 0.4225663716814159,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 8.94358407079646e-06,
+      "loss": 0.8093,
+      "step": 382
+    },
+    {
+      "epoch": 0.42367256637168144,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 8.940818584070798e-06,
+      "loss": 0.8792,
+      "step": 383
+    },
+    {
+      "epoch": 0.4247787610619469,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.938053097345133e-06,
+      "loss": 0.8572,
+      "step": 384
+    },
+    {
+      "epoch": 0.4258849557522124,
+      "grad_norm": 0.095703125,
+      "learning_rate": 8.93528761061947e-06,
+      "loss": 0.823,
+      "step": 385
+    },
+    {
+      "epoch": 0.4269911504424779,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.932522123893806e-06,
+      "loss": 0.8453,
+      "step": 386
+    },
+    {
+      "epoch": 0.4280973451327434,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 8.929756637168143e-06,
+      "loss": 0.8318,
+      "step": 387
+    },
+    {
+      "epoch": 0.42920353982300885,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 8.926991150442479e-06,
+      "loss": 0.8617,
+      "step": 388
+    },
+    {
+      "epoch": 0.4303097345132743,
+      "grad_norm": 0.0859375,
+      "learning_rate": 8.924225663716815e-06,
+      "loss": 0.8004,
+      "step": 389
+    },
+    {
+      "epoch": 0.4314159292035398,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.921460176991151e-06,
+      "loss": 0.8504,
+      "step": 390
+    },
+    {
+      "epoch": 0.4325221238938053,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.918694690265487e-06,
+      "loss": 0.8461,
+      "step": 391
+    },
+    {
+      "epoch": 0.4336283185840708,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.915929203539823e-06,
+      "loss": 0.7912,
+      "step": 392
+    },
+    {
+      "epoch": 0.43473451327433627,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 8.913163716814161e-06,
+      "loss": 0.8467,
+      "step": 393
+    },
+    {
+      "epoch": 0.4358407079646018,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.910398230088495e-06,
+      "loss": 0.8583,
+      "step": 394
+    },
+    {
+      "epoch": 0.43694690265486724,
+      "grad_norm": 0.107421875,
+      "learning_rate": 8.907632743362833e-06,
+      "loss": 0.8209,
+      "step": 395
+    },
+    {
+      "epoch": 0.43805309734513276,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 8.90486725663717e-06,
+      "loss": 0.8608,
+      "step": 396
+    },
+    {
+      "epoch": 0.4391592920353982,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 8.902101769911506e-06,
+      "loss": 0.8511,
+      "step": 397
+    },
+    {
+      "epoch": 0.44026548672566373,
+      "grad_norm": 0.095703125,
+      "learning_rate": 8.899336283185842e-06,
+      "loss": 0.8158,
+      "step": 398
+    },
+    {
+      "epoch": 0.4413716814159292,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.896570796460178e-06,
+      "loss": 0.8531,
+      "step": 399
+    },
+    {
+      "epoch": 0.4424778761061947,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 8.893805309734514e-06,
+      "loss": 0.8699,
+      "step": 400
+    },
+    {
+      "epoch": 0.4435840707964602,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 8.89103982300885e-06,
+      "loss": 0.8092,
+      "step": 401
+    },
+    {
+      "epoch": 0.4446902654867257,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.888274336283186e-06,
+      "loss": 0.821,
+      "step": 402
+    },
+    {
+      "epoch": 0.44579646017699115,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 8.885508849557522e-06,
+      "loss": 0.8509,
+      "step": 403
+    },
+    {
+      "epoch": 0.4469026548672566,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.88274336283186e-06,
+      "loss": 0.8846,
+      "step": 404
+    },
+    {
+      "epoch": 0.4480088495575221,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.879977876106195e-06,
+      "loss": 0.866,
+      "step": 405
+    },
+    {
+      "epoch": 0.4491150442477876,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.877212389380532e-06,
+      "loss": 0.8918,
+      "step": 406
+    },
+    {
+      "epoch": 0.4502212389380531,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 8.874446902654868e-06,
+      "loss": 0.8679,
+      "step": 407
+    },
+    {
+      "epoch": 0.45132743362831856,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.871681415929205e-06,
+      "loss": 0.8862,
+      "step": 408
+    },
+    {
+      "epoch": 0.4524336283185841,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 8.86891592920354e-06,
+      "loss": 0.8015,
+      "step": 409
+    },
+    {
+      "epoch": 0.45353982300884954,
+      "grad_norm": 0.099609375,
+      "learning_rate": 8.866150442477877e-06,
+      "loss": 0.8686,
+      "step": 410
+    },
+    {
+      "epoch": 0.45464601769911506,
+      "grad_norm": 0.095703125,
+      "learning_rate": 8.863384955752213e-06,
+      "loss": 0.8788,
+      "step": 411
+    },
+    {
+      "epoch": 0.4557522123893805,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 8.860619469026549e-06,
+      "loss": 0.8277,
+      "step": 412
+    },
+    {
+      "epoch": 0.45685840707964603,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 8.857853982300885e-06,
+      "loss": 0.8081,
+      "step": 413
+    },
+    {
+      "epoch": 0.4579646017699115,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.855088495575221e-06,
+      "loss": 0.8373,
+      "step": 414
+    },
+    {
+      "epoch": 0.459070796460177,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 8.852323008849557e-06,
+      "loss": 0.8426,
+      "step": 415
+    },
+    {
+      "epoch": 0.46017699115044247,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 8.849557522123895e-06,
+      "loss": 0.8903,
+      "step": 416
+    },
+    {
+      "epoch": 0.461283185840708,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.846792035398231e-06,
+      "loss": 0.8729,
+      "step": 417
+    },
+    {
+      "epoch": 0.46238938053097345,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 8.844026548672568e-06,
+      "loss": 0.8694,
+      "step": 418
+    },
+    {
+      "epoch": 0.46349557522123896,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 8.841261061946904e-06,
+      "loss": 0.8389,
+      "step": 419
+    },
+    {
+      "epoch": 0.4646017699115044,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 8.83849557522124e-06,
+      "loss": 0.8533,
+      "step": 420
+    },
+    {
+      "epoch": 0.4657079646017699,
+      "grad_norm": 0.0810546875,
+      "learning_rate": 8.835730088495576e-06,
+      "loss": 0.8262,
+      "step": 421
+    },
+    {
+      "epoch": 0.4668141592920354,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 8.832964601769912e-06,
+      "loss": 0.889,
+      "step": 422
+    },
+    {
+      "epoch": 0.46792035398230086,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 8.830199115044248e-06,
+      "loss": 0.8181,
+      "step": 423
+    },
+    {
+      "epoch": 0.4690265486725664,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 8.827433628318584e-06,
+      "loss": 0.8738,
+      "step": 424
+    },
+    {
+      "epoch": 0.47013274336283184,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.824668141592922e-06,
+      "loss": 0.8676,
+      "step": 425
+    },
+    {
+      "epoch": 0.47123893805309736,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.821902654867257e-06,
+      "loss": 0.8099,
+      "step": 426
+    },
+    {
+      "epoch": 0.4723451327433628,
+      "grad_norm": 0.0859375,
+      "learning_rate": 8.819137168141594e-06,
+      "loss": 0.8439,
+      "step": 427
+    },
+    {
+      "epoch": 0.47345132743362833,
+      "grad_norm": 0.095703125,
+      "learning_rate": 8.816371681415929e-06,
+      "loss": 0.8671,
+      "step": 428
+    },
+    {
+      "epoch": 0.4745575221238938,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 8.813606194690267e-06,
+      "loss": 0.8585,
+      "step": 429
+    },
+    {
+      "epoch": 0.4756637168141593,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 8.810840707964603e-06,
+      "loss": 0.8208,
+      "step": 430
+    },
+    {
+      "epoch": 0.47676991150442477,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 8.808075221238939e-06,
+      "loss": 0.8619,
+      "step": 431
+    },
+    {
+      "epoch": 0.4778761061946903,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 8.805309734513275e-06,
+      "loss": 0.8331,
+      "step": 432
+    },
+    {
+      "epoch": 0.47898230088495575,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 8.802544247787611e-06,
+      "loss": 0.8209,
+      "step": 433
+    },
+    {
+      "epoch": 0.48008849557522126,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.799778761061947e-06,
+      "loss": 0.886,
+      "step": 434
+    },
+    {
+      "epoch": 0.4811946902654867,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 8.797013274336283e-06,
+      "loss": 0.8304,
+      "step": 435
+    },
+    {
+      "epoch": 0.4823008849557522,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.79424778761062e-06,
+      "loss": 0.8418,
+      "step": 436
+    },
+    {
+      "epoch": 0.4834070796460177,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 8.791482300884957e-06,
+      "loss": 0.8284,
+      "step": 437
+    },
+    {
+      "epoch": 0.48451327433628316,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 8.788716814159293e-06,
+      "loss": 0.8258,
+      "step": 438
+    },
+    {
+      "epoch": 0.4856194690265487,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 8.78595132743363e-06,
+      "loss": 0.8641,
+      "step": 439
+    },
+    {
+      "epoch": 0.48672566371681414,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.783185840707966e-06,
+      "loss": 0.8718,
+      "step": 440
+    },
+    {
+      "epoch": 0.48783185840707965,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 8.780420353982302e-06,
+      "loss": 0.8229,
+      "step": 441
+    },
+    {
+      "epoch": 0.4889380530973451,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.777654867256638e-06,
+      "loss": 0.8347,
+      "step": 442
+    },
+    {
+      "epoch": 0.49004424778761063,
+      "grad_norm": 0.0791015625,
+      "learning_rate": 8.774889380530974e-06,
+      "loss": 0.7614,
+      "step": 443
+    },
+    {
+      "epoch": 0.4911504424778761,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.77212389380531e-06,
+      "loss": 0.8231,
+      "step": 444
+    },
+    {
+      "epoch": 0.4922566371681416,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 8.769358407079646e-06,
+      "loss": 0.8104,
+      "step": 445
+    },
+    {
+      "epoch": 0.49336283185840707,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.766592920353984e-06,
+      "loss": 0.8738,
+      "step": 446
+    },
+    {
+      "epoch": 0.4944690265486726,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.763827433628319e-06,
+      "loss": 0.8535,
+      "step": 447
+    },
+    {
+      "epoch": 0.49557522123893805,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.761061946902656e-06,
+      "loss": 0.8282,
+      "step": 448
+    },
+    {
+      "epoch": 0.49668141592920356,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 8.758296460176991e-06,
+      "loss": 0.8288,
+      "step": 449
+    },
+    {
+      "epoch": 0.497787610619469,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.755530973451329e-06,
+      "loss": 0.8366,
+      "step": 450
+    },
+    {
+      "epoch": 0.49889380530973454,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.752765486725663e-06,
+      "loss": 0.8576,
+      "step": 451
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.835,
+      "step": 452
+    },
+    {
+      "epoch": 0.5011061946902655,
+      "grad_norm": 0.111328125,
+      "learning_rate": 8.747234513274337e-06,
+      "loss": 0.8239,
+      "step": 453
+    },
+    {
+      "epoch": 0.5022123893805309,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.744469026548673e-06,
+      "loss": 0.8043,
+      "step": 454
+    },
+    {
+      "epoch": 0.5033185840707964,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 8.74170353982301e-06,
+      "loss": 0.8495,
+      "step": 455
+    },
+    {
+      "epoch": 0.504424778761062,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.738938053097345e-06,
+      "loss": 0.8298,
+      "step": 456
+    },
+    {
+      "epoch": 0.5055309734513275,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.736172566371682e-06,
+      "loss": 0.8213,
+      "step": 457
+    },
+    {
+      "epoch": 0.5066371681415929,
+      "grad_norm": 0.07763671875,
+      "learning_rate": 8.733407079646018e-06,
+      "loss": 0.7884,
+      "step": 458
+    },
+    {
+      "epoch": 0.5077433628318584,
+      "grad_norm": 0.33203125,
+      "learning_rate": 8.730641592920356e-06,
+      "loss": 0.8921,
+      "step": 459
+    },
+    {
+      "epoch": 0.5088495575221239,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 8.727876106194692e-06,
+      "loss": 0.8226,
+      "step": 460
+    },
+    {
+      "epoch": 0.5099557522123894,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 8.725110619469028e-06,
+      "loss": 0.8441,
+      "step": 461
+    },
+    {
+      "epoch": 0.5110619469026548,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 8.722345132743364e-06,
+      "loss": 0.8937,
+      "step": 462
+    },
+    {
+      "epoch": 0.5121681415929203,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.7195796460177e-06,
+      "loss": 0.7763,
+      "step": 463
+    },
+    {
+      "epoch": 0.5132743362831859,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.716814159292036e-06,
+      "loss": 0.8601,
+      "step": 464
+    },
+    {
+      "epoch": 0.5143805309734514,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 8.714048672566372e-06,
+      "loss": 0.8163,
+      "step": 465
+    },
+    {
+      "epoch": 0.5154867256637168,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 8.711283185840708e-06,
+      "loss": 0.8157,
+      "step": 466
+    },
+    {
+      "epoch": 0.5165929203539823,
+      "grad_norm": 0.080078125,
+      "learning_rate": 8.708517699115045e-06,
+      "loss": 0.7975,
+      "step": 467
+    },
+    {
+      "epoch": 0.5176991150442478,
+      "grad_norm": 0.08984375,
+      "learning_rate": 8.70575221238938e-06,
+      "loss": 0.7654,
+      "step": 468
+    },
+    {
+      "epoch": 0.5188053097345132,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 8.702986725663718e-06,
+      "loss": 0.8482,
+      "step": 469
+    },
+    {
+      "epoch": 0.5199115044247787,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 8.700221238938053e-06,
+      "loss": 0.833,
+      "step": 470
+    },
+    {
+      "epoch": 0.5210176991150443,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.69745575221239e-06,
+      "loss": 0.8977,
+      "step": 471
+    },
+    {
+      "epoch": 0.5221238938053098,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.694690265486727e-06,
+      "loss": 0.8902,
+      "step": 472
+    },
+    {
+      "epoch": 0.5232300884955752,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.691924778761063e-06,
+      "loss": 0.8423,
+      "step": 473
+    },
+    {
+      "epoch": 0.5243362831858407,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.689159292035399e-06,
+      "loss": 0.8301,
+      "step": 474
+    },
+    {
+      "epoch": 0.5254424778761062,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 8.686393805309735e-06,
+      "loss": 0.8679,
+      "step": 475
+    },
+    {
+      "epoch": 0.5265486725663717,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 8.683628318584071e-06,
+      "loss": 0.8334,
+      "step": 476
+    },
+    {
+      "epoch": 0.5276548672566371,
+      "grad_norm": 0.091796875,
+      "learning_rate": 8.680862831858407e-06,
+      "loss": 0.814,
+      "step": 477
+    },
+    {
+      "epoch": 0.5287610619469026,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.678097345132744e-06,
+      "loss": 0.8571,
+      "step": 478
+    },
+    {
+      "epoch": 0.5298672566371682,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.67533185840708e-06,
+      "loss": 0.7996,
+      "step": 479
+    },
+    {
+      "epoch": 0.5309734513274337,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 8.672566371681418e-06,
+      "loss": 0.8152,
+      "step": 480
+    },
+    {
+      "epoch": 0.5320796460176991,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 8.669800884955752e-06,
+      "loss": 0.8303,
+      "step": 481
+    },
+    {
+      "epoch": 0.5331858407079646,
+      "grad_norm": 0.10546875,
+      "learning_rate": 8.66703539823009e-06,
+      "loss": 0.8647,
+      "step": 482
+    },
+    {
+      "epoch": 0.5342920353982301,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.664269911504426e-06,
+      "loss": 0.8207,
+      "step": 483
+    },
+    {
+      "epoch": 0.5353982300884956,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.661504424778762e-06,
+      "loss": 0.8429,
+      "step": 484
+    },
+    {
+      "epoch": 0.536504424778761,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.658738938053098e-06,
+      "loss": 0.8134,
+      "step": 485
+    },
+    {
+      "epoch": 0.5376106194690266,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.655973451327434e-06,
+      "loss": 0.7738,
+      "step": 486
+    },
+    {
+      "epoch": 0.5387168141592921,
+      "grad_norm": 0.099609375,
+      "learning_rate": 8.65320796460177e-06,
+      "loss": 0.8431,
+      "step": 487
+    },
+    {
+      "epoch": 0.5398230088495575,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 8.650442477876107e-06,
+      "loss": 0.792,
+      "step": 488
+    },
+    {
+      "epoch": 0.540929203539823,
+      "grad_norm": 0.0830078125,
+      "learning_rate": 8.647676991150443e-06,
+      "loss": 0.7863,
+      "step": 489
+    },
+    {
+      "epoch": 0.5420353982300885,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.64491150442478e-06,
+      "loss": 0.7907,
+      "step": 490
+    },
+    {
+      "epoch": 0.543141592920354,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.642146017699115e-06,
+      "loss": 0.8102,
+      "step": 491
+    },
+    {
+      "epoch": 0.5442477876106194,
+      "grad_norm": 0.095703125,
+      "learning_rate": 8.639380530973453e-06,
+      "loss": 0.8298,
+      "step": 492
+    },
+    {
+      "epoch": 0.5453539823008849,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 8.636615044247789e-06,
+      "loss": 0.8196,
+      "step": 493
+    },
+    {
+      "epoch": 0.5464601769911505,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 8.633849557522125e-06,
+      "loss": 0.8665,
+      "step": 494
+    },
+    {
+      "epoch": 0.547566371681416,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 8.631084070796461e-06,
+      "loss": 0.8216,
+      "step": 495
+    },
+    {
+      "epoch": 0.5486725663716814,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 8.628318584070797e-06,
+      "loss": 0.7836,
+      "step": 496
+    },
+    {
+      "epoch": 0.5497787610619469,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.625553097345133e-06,
+      "loss": 0.8403,
+      "step": 497
+    },
+    {
+      "epoch": 0.5508849557522124,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.62278761061947e-06,
+      "loss": 0.7662,
+      "step": 498
+    },
+    {
+      "epoch": 0.5519911504424779,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 8.620022123893806e-06,
+      "loss": 0.7539,
+      "step": 499
+    },
+    {
+      "epoch": 0.5530973451327433,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 8.617256637168142e-06,
+      "loss": 0.8757,
+      "step": 500
+    },
+    {
+      "epoch": 0.5542035398230089,
+      "grad_norm": 0.095703125,
+      "learning_rate": 8.61449115044248e-06,
+      "loss": 0.8628,
+      "step": 501
+    },
+    {
+      "epoch": 0.5553097345132744,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.611725663716814e-06,
+      "loss": 0.8246,
+      "step": 502
+    },
+    {
+      "epoch": 0.5564159292035398,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 8.608960176991152e-06,
+      "loss": 0.8075,
+      "step": 503
+    },
+    {
+      "epoch": 0.5575221238938053,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.606194690265486e-06,
+      "loss": 0.8745,
+      "step": 504
+    },
+    {
+      "epoch": 0.5586283185840708,
+      "grad_norm": 0.08447265625,
+      "learning_rate": 8.603429203539824e-06,
+      "loss": 0.7894,
+      "step": 505
+    },
+    {
+      "epoch": 0.5597345132743363,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.60066371681416e-06,
+      "loss": 0.8169,
+      "step": 506
+    },
+    {
+      "epoch": 0.5608407079646017,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 8.597898230088496e-06,
+      "loss": 0.844,
+      "step": 507
+    },
+    {
+      "epoch": 0.5619469026548672,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.595132743362832e-06,
+      "loss": 0.8059,
+      "step": 508
+    },
+    {
+      "epoch": 0.5630530973451328,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 8.592367256637169e-06,
+      "loss": 0.8305,
+      "step": 509
+    },
+    {
+      "epoch": 0.5641592920353983,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 8.589601769911505e-06,
+      "loss": 0.8594,
+      "step": 510
+    },
+    {
+      "epoch": 0.5652654867256637,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 8.58683628318584e-06,
+      "loss": 0.8155,
+      "step": 511
+    },
+    {
+      "epoch": 0.5663716814159292,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 8.584070796460177e-06,
+      "loss": 0.8764,
+      "step": 512
+    },
+    {
+      "epoch": 0.5674778761061947,
+      "grad_norm": 0.08984375,
+      "learning_rate": 8.581305309734515e-06,
+      "loss": 0.8306,
+      "step": 513
+    },
+    {
+      "epoch": 0.5685840707964602,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.578539823008851e-06,
+      "loss": 0.8944,
+      "step": 514
+    },
+    {
+      "epoch": 0.5696902654867256,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.575774336283187e-06,
+      "loss": 0.8218,
+      "step": 515
+    },
+    {
+      "epoch": 0.5707964601769911,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.573008849557523e-06,
+      "loss": 0.8683,
+      "step": 516
+    },
+    {
+      "epoch": 0.5719026548672567,
+      "grad_norm": 0.10546875,
+      "learning_rate": 8.57024336283186e-06,
+      "loss": 0.8448,
+      "step": 517
+    },
+    {
+      "epoch": 0.5730088495575221,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.567477876106195e-06,
+      "loss": 0.7946,
+      "step": 518
+    },
+    {
+      "epoch": 0.5741150442477876,
+      "grad_norm": 0.109375,
+      "learning_rate": 8.564712389380532e-06,
+      "loss": 0.8277,
+      "step": 519
+    },
+    {
+      "epoch": 0.5752212389380531,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 8.561946902654868e-06,
+      "loss": 0.8808,
+      "step": 520
+    },
+    {
+      "epoch": 0.5763274336283186,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.559181415929204e-06,
+      "loss": 0.8109,
+      "step": 521
+    },
+    {
+      "epoch": 0.577433628318584,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 8.556415929203542e-06,
+      "loss": 0.7998,
+      "step": 522
+    },
+    {
+      "epoch": 0.5785398230088495,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.553650442477876e-06,
+      "loss": 0.8062,
+      "step": 523
+    },
+    {
+      "epoch": 0.5796460176991151,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.550884955752214e-06,
+      "loss": 0.7741,
+      "step": 524
+    },
+    {
+      "epoch": 0.5807522123893806,
+      "grad_norm": 0.10546875,
+      "learning_rate": 8.548119469026548e-06,
+      "loss": 0.8697,
+      "step": 525
+    },
+    {
+      "epoch": 0.581858407079646,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.545353982300886e-06,
+      "loss": 0.8054,
+      "step": 526
+    },
+    {
+      "epoch": 0.5829646017699115,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 8.542588495575222e-06,
+      "loss": 0.8275,
+      "step": 527
+    },
+    {
+      "epoch": 0.584070796460177,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 8.539823008849558e-06,
+      "loss": 0.8137,
+      "step": 528
+    },
+    {
+      "epoch": 0.5851769911504425,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 8.537057522123894e-06,
+      "loss": 0.7852,
+      "step": 529
+    },
+    {
+      "epoch": 0.5862831858407079,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.53429203539823e-06,
+      "loss": 0.809,
+      "step": 530
+    },
+    {
+      "epoch": 0.5873893805309734,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 8.531526548672567e-06,
+      "loss": 0.8851,
+      "step": 531
+    },
+    {
+      "epoch": 0.588495575221239,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.528761061946903e-06,
+      "loss": 0.8105,
+      "step": 532
+    },
+    {
+      "epoch": 0.5896017699115044,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 8.525995575221239e-06,
+      "loss": 0.8216,
+      "step": 533
+    },
+    {
+      "epoch": 0.5907079646017699,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 8.523230088495575e-06,
+      "loss": 0.7709,
+      "step": 534
+    },
+    {
+      "epoch": 0.5918141592920354,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.520464601769913e-06,
+      "loss": 0.8379,
+      "step": 535
+    },
+    {
+      "epoch": 0.5929203539823009,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.517699115044249e-06,
+      "loss": 0.8057,
+      "step": 536
+    },
+    {
+      "epoch": 0.5940265486725663,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 8.514933628318585e-06,
+      "loss": 0.8574,
+      "step": 537
+    },
+    {
+      "epoch": 0.5951327433628318,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 8.512168141592921e-06,
+      "loss": 0.8196,
+      "step": 538
+    },
+    {
+      "epoch": 0.5962389380530974,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.509402654867257e-06,
+      "loss": 0.7996,
+      "step": 539
+    },
+    {
+      "epoch": 0.5973451327433629,
+      "grad_norm": 0.119140625,
+      "learning_rate": 8.506637168141594e-06,
+      "loss": 0.8363,
+      "step": 540
+    },
+    {
+      "epoch": 0.5984513274336283,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.50387168141593e-06,
+      "loss": 0.8636,
+      "step": 541
+    },
+    {
+      "epoch": 0.5995575221238938,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.501106194690266e-06,
+      "loss": 0.7922,
+      "step": 542
+    },
+    {
+      "epoch": 0.6006637168141593,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.498340707964604e-06,
+      "loss": 0.8071,
+      "step": 543
+    },
+    {
+      "epoch": 0.6017699115044248,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.495575221238938e-06,
+      "loss": 0.8367,
+      "step": 544
+    },
+    {
+      "epoch": 0.6028761061946902,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.492809734513276e-06,
+      "loss": 0.8151,
+      "step": 545
+    },
+    {
+      "epoch": 0.6039823008849557,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.49004424778761e-06,
+      "loss": 0.8159,
+      "step": 546
+    },
+    {
+      "epoch": 0.6050884955752213,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 8.487278761061948e-06,
+      "loss": 0.8061,
+      "step": 547
+    },
+    {
+      "epoch": 0.6061946902654868,
+      "grad_norm": 0.091796875,
+      "learning_rate": 8.484513274336284e-06,
+      "loss": 0.8266,
+      "step": 548
+    },
+    {
+      "epoch": 0.6073008849557522,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.48174778761062e-06,
+      "loss": 0.8631,
+      "step": 549
+    },
+    {
+      "epoch": 0.6084070796460177,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 8.478982300884957e-06,
+      "loss": 0.8643,
+      "step": 550
+    },
+    {
+      "epoch": 0.6095132743362832,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 8.476216814159293e-06,
+      "loss": 0.8369,
+      "step": 551
+    },
+    {
+      "epoch": 0.6106194690265486,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 8.473451327433629e-06,
+      "loss": 0.8379,
+      "step": 552
+    },
+    {
+      "epoch": 0.6117256637168141,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 8.470685840707965e-06,
+      "loss": 0.8289,
+      "step": 553
+    },
+    {
+      "epoch": 0.6128318584070797,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.467920353982301e-06,
+      "loss": 0.8463,
+      "step": 554
+    },
+    {
+      "epoch": 0.6139380530973452,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.465154867256637e-06,
+      "loss": 0.8268,
+      "step": 555
+    },
+    {
+      "epoch": 0.6150442477876106,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.462389380530975e-06,
+      "loss": 0.8208,
+      "step": 556
+    },
+    {
+      "epoch": 0.6161504424778761,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 8.45962389380531e-06,
+      "loss": 0.8201,
+      "step": 557
+    },
+    {
+      "epoch": 0.6172566371681416,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.456858407079647e-06,
+      "loss": 0.7752,
+      "step": 558
+    },
+    {
+      "epoch": 0.6183628318584071,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.454092920353983e-06,
+      "loss": 0.883,
+      "step": 559
+    },
+    {
+      "epoch": 0.6194690265486725,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.45132743362832e-06,
+      "loss": 0.8121,
+      "step": 560
+    },
+    {
+      "epoch": 0.620575221238938,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.448561946902656e-06,
+      "loss": 0.79,
+      "step": 561
+    },
+    {
+      "epoch": 0.6216814159292036,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 8.445796460176992e-06,
+      "loss": 0.7949,
+      "step": 562
+    },
+    {
+      "epoch": 0.6227876106194691,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.443030973451328e-06,
+      "loss": 0.7928,
+      "step": 563
+    },
+    {
+      "epoch": 0.6238938053097345,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 8.440265486725664e-06,
+      "loss": 0.8075,
+      "step": 564
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 8.4375e-06,
+      "loss": 0.835,
+      "step": 565
+    },
+    {
+      "epoch": 0.6261061946902655,
+      "grad_norm": 0.095703125,
+      "learning_rate": 8.434734513274338e-06,
+      "loss": 0.8108,
+      "step": 566
+    },
+    {
+      "epoch": 0.6272123893805309,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 8.431969026548672e-06,
+      "loss": 0.8396,
+      "step": 567
+    },
+    {
+      "epoch": 0.6283185840707964,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.42920353982301e-06,
+      "loss": 0.8253,
+      "step": 568
+    },
+    {
+      "epoch": 0.629424778761062,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.426438053097346e-06,
+      "loss": 0.8263,
+      "step": 569
+    },
+    {
+      "epoch": 0.6305309734513275,
+      "grad_norm": 0.10546875,
+      "learning_rate": 8.423672566371682e-06,
+      "loss": 0.8499,
+      "step": 570
+    },
+    {
+      "epoch": 0.6316371681415929,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.420907079646019e-06,
+      "loss": 0.7957,
+      "step": 571
+    },
+    {
+      "epoch": 0.6327433628318584,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 8.418141592920355e-06,
+      "loss": 0.7729,
+      "step": 572
+    },
+    {
+      "epoch": 0.6338495575221239,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.41537610619469e-06,
+      "loss": 0.7931,
+      "step": 573
+    },
+    {
+      "epoch": 0.6349557522123894,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 8.412610619469027e-06,
+      "loss": 0.8112,
+      "step": 574
+    },
+    {
+      "epoch": 0.6360619469026548,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.409845132743363e-06,
+      "loss": 0.8417,
+      "step": 575
+    },
+    {
+      "epoch": 0.6371681415929203,
+      "grad_norm": 0.0869140625,
+      "learning_rate": 8.4070796460177e-06,
+      "loss": 0.789,
+      "step": 576
+    },
+    {
+      "epoch": 0.6382743362831859,
+      "grad_norm": 0.111328125,
+      "learning_rate": 8.404314159292037e-06,
+      "loss": 0.8299,
+      "step": 577
+    },
+    {
+      "epoch": 0.6393805309734514,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.401548672566371e-06,
+      "loss": 0.8088,
+      "step": 578
+    },
+    {
+      "epoch": 0.6404867256637168,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.39878318584071e-06,
+      "loss": 0.8277,
+      "step": 579
+    },
+    {
+      "epoch": 0.6415929203539823,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.396017699115044e-06,
+      "loss": 0.8481,
+      "step": 580
+    },
+    {
+      "epoch": 0.6426991150442478,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.393252212389382e-06,
+      "loss": 0.829,
+      "step": 581
+    },
+    {
+      "epoch": 0.6438053097345132,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 8.390486725663718e-06,
+      "loss": 0.7889,
+      "step": 582
+    },
+    {
+      "epoch": 0.6449115044247787,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.387721238938054e-06,
+      "loss": 0.8297,
+      "step": 583
+    },
+    {
+      "epoch": 0.6460176991150443,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 8.38495575221239e-06,
+      "loss": 0.8153,
+      "step": 584
+    },
+    {
+      "epoch": 0.6471238938053098,
+      "grad_norm": 0.107421875,
+      "learning_rate": 8.382190265486726e-06,
+      "loss": 0.7944,
+      "step": 585
+    },
+    {
+      "epoch": 0.6482300884955752,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 8.379424778761062e-06,
+      "loss": 0.801,
+      "step": 586
+    },
+    {
+      "epoch": 0.6493362831858407,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 8.376659292035398e-06,
+      "loss": 0.8954,
+      "step": 587
+    },
+    {
+      "epoch": 0.6504424778761062,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.373893805309734e-06,
+      "loss": 0.8318,
+      "step": 588
+    },
+    {
+      "epoch": 0.6515486725663717,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.371128318584072e-06,
+      "loss": 0.7723,
+      "step": 589
+    },
+    {
+      "epoch": 0.6526548672566371,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.368362831858408e-06,
+      "loss": 0.8451,
+      "step": 590
+    },
+    {
+      "epoch": 0.6537610619469026,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 8.365597345132744e-06,
+      "loss": 0.8182,
+      "step": 591
+    },
+    {
+      "epoch": 0.6548672566371682,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.36283185840708e-06,
+      "loss": 0.7703,
+      "step": 592
+    },
+    {
+      "epoch": 0.6559734513274337,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.360066371681417e-06,
+      "loss": 0.7927,
+      "step": 593
+    },
+    {
+      "epoch": 0.6570796460176991,
+      "grad_norm": 0.099609375,
+      "learning_rate": 8.357300884955753e-06,
+      "loss": 0.8104,
+      "step": 594
+    },
+    {
+      "epoch": 0.6581858407079646,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.354535398230089e-06,
+      "loss": 0.7873,
+      "step": 595
+    },
+    {
+      "epoch": 0.6592920353982301,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 8.351769911504427e-06,
+      "loss": 0.8329,
+      "step": 596
+    },
+    {
+      "epoch": 0.6603982300884956,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.349004424778761e-06,
+      "loss": 0.7717,
+      "step": 597
+    },
+    {
+      "epoch": 0.661504424778761,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 8.346238938053099e-06,
+      "loss": 0.7815,
+      "step": 598
+    },
+    {
+      "epoch": 0.6626106194690266,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.343473451327433e-06,
+      "loss": 0.8017,
+      "step": 599
+    },
+    {
+      "epoch": 0.6637168141592921,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.340707964601771e-06,
+      "loss": 0.7866,
+      "step": 600
+    },
+    {
+      "epoch": 0.6648230088495575,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 8.337942477876106e-06,
+      "loss": 0.8006,
+      "step": 601
+    },
+    {
+      "epoch": 0.665929203539823,
+      "grad_norm": 0.087890625,
+      "learning_rate": 8.335176991150444e-06,
+      "loss": 0.7718,
+      "step": 602
+    },
+    {
+      "epoch": 0.6670353982300885,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 8.33241150442478e-06,
+      "loss": 0.7884,
+      "step": 603
+    },
+    {
+      "epoch": 0.668141592920354,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.329646017699116e-06,
+      "loss": 0.8159,
+      "step": 604
+    },
+    {
+      "epoch": 0.6692477876106194,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.326880530973452e-06,
+      "loss": 0.7906,
+      "step": 605
+    },
+    {
+      "epoch": 0.6703539823008849,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 8.324115044247788e-06,
+      "loss": 0.7802,
+      "step": 606
+    },
+    {
+      "epoch": 0.6714601769911505,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.321349557522124e-06,
+      "loss": 0.878,
+      "step": 607
+    },
+    {
+      "epoch": 0.672566371681416,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 8.31858407079646e-06,
+      "loss": 0.8006,
+      "step": 608
+    },
+    {
+      "epoch": 0.6736725663716814,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.315818584070796e-06,
+      "loss": 0.847,
+      "step": 609
+    },
+    {
+      "epoch": 0.6747787610619469,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.313053097345133e-06,
+      "loss": 0.8437,
+      "step": 610
+    },
+    {
+      "epoch": 0.6758849557522124,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 8.31028761061947e-06,
+      "loss": 0.8285,
+      "step": 611
+    },
+    {
+      "epoch": 0.6769911504424779,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.307522123893807e-06,
+      "loss": 0.8781,
+      "step": 612
+    },
+    {
+      "epoch": 0.6780973451327433,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.304756637168143e-06,
+      "loss": 0.8166,
+      "step": 613
+    },
+    {
+      "epoch": 0.6792035398230089,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 8.301991150442479e-06,
+      "loss": 0.8639,
+      "step": 614
+    },
+    {
+      "epoch": 0.6803097345132744,
+      "grad_norm": 0.123046875,
+      "learning_rate": 8.299225663716815e-06,
+      "loss": 0.808,
+      "step": 615
+    },
+    {
+      "epoch": 0.6814159292035398,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.296460176991151e-06,
+      "loss": 0.7733,
+      "step": 616
+    },
+    {
+      "epoch": 0.6825221238938053,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 8.293694690265487e-06,
+      "loss": 0.857,
+      "step": 617
+    },
+    {
+      "epoch": 0.6836283185840708,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.290929203539823e-06,
+      "loss": 0.8342,
+      "step": 618
+    },
+    {
+      "epoch": 0.6847345132743363,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.288163716814161e-06,
+      "loss": 0.7707,
+      "step": 619
+    },
+    {
+      "epoch": 0.6858407079646017,
+      "grad_norm": 0.08984375,
+      "learning_rate": 8.285398230088496e-06,
+      "loss": 0.7488,
+      "step": 620
+    },
+    {
+      "epoch": 0.6869469026548672,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.282632743362833e-06,
+      "loss": 0.794,
+      "step": 621
+    },
+    {
+      "epoch": 0.6880530973451328,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 8.279867256637168e-06,
+      "loss": 0.831,
+      "step": 622
+    },
+    {
+      "epoch": 0.6891592920353983,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 8.277101769911506e-06,
+      "loss": 0.8474,
+      "step": 623
+    },
+    {
+      "epoch": 0.6902654867256637,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.274336283185842e-06,
+      "loss": 0.8716,
+      "step": 624
+    },
+    {
+      "epoch": 0.6913716814159292,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.271570796460178e-06,
+      "loss": 0.8067,
+      "step": 625
+    },
+    {
+      "epoch": 0.6924778761061947,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 8.268805309734514e-06,
+      "loss": 0.7858,
+      "step": 626
+    },
+    {
+      "epoch": 0.6935840707964602,
+      "grad_norm": 0.115234375,
+      "learning_rate": 8.26603982300885e-06,
+      "loss": 0.8209,
+      "step": 627
+    },
+    {
+      "epoch": 0.6946902654867256,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.263274336283186e-06,
+      "loss": 0.8252,
+      "step": 628
+    },
+    {
+      "epoch": 0.6957964601769911,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 8.260508849557522e-06,
+      "loss": 0.8108,
+      "step": 629
+    },
+    {
+      "epoch": 0.6969026548672567,
+      "grad_norm": 0.0849609375,
+      "learning_rate": 8.257743362831858e-06,
+      "loss": 0.763,
+      "step": 630
+    },
+    {
+      "epoch": 0.6980088495575221,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 8.254977876106195e-06,
+      "loss": 0.8399,
+      "step": 631
+    },
+    {
+      "epoch": 0.6991150442477876,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 8.252212389380532e-06,
+      "loss": 0.8725,
+      "step": 632
+    },
+    {
+      "epoch": 0.7002212389380531,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.249446902654869e-06,
+      "loss": 0.8419,
+      "step": 633
+    },
+    {
+      "epoch": 0.7013274336283186,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.246681415929205e-06,
+      "loss": 0.7878,
+      "step": 634
+    },
+    {
+      "epoch": 0.702433628318584,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.24391592920354e-06,
+      "loss": 0.7929,
+      "step": 635
+    },
+    {
+      "epoch": 0.7035398230088495,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.241150442477877e-06,
+      "loss": 0.787,
+      "step": 636
+    },
+    {
+      "epoch": 0.7046460176991151,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.238384955752213e-06,
+      "loss": 0.7962,
+      "step": 637
+    },
+    {
+      "epoch": 0.7057522123893806,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.23561946902655e-06,
+      "loss": 0.8427,
+      "step": 638
+    },
+    {
+      "epoch": 0.706858407079646,
+      "grad_norm": 0.0859375,
+      "learning_rate": 8.232853982300885e-06,
+      "loss": 0.7975,
+      "step": 639
+    },
+    {
+      "epoch": 0.7079646017699115,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 8.230088495575221e-06,
+      "loss": 0.8305,
+      "step": 640
+    },
+    {
+      "epoch": 0.709070796460177,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 8.227323008849558e-06,
+      "loss": 0.8375,
+      "step": 641
+    },
+    {
+      "epoch": 0.7101769911504425,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.224557522123895e-06,
+      "loss": 0.8371,
+      "step": 642
+    },
+    {
+      "epoch": 0.7112831858407079,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.22179203539823e-06,
+      "loss": 0.8304,
+      "step": 643
+    },
+    {
+      "epoch": 0.7123893805309734,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 8.219026548672568e-06,
+      "loss": 0.7761,
+      "step": 644
+    },
+    {
+      "epoch": 0.713495575221239,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.216261061946904e-06,
+      "loss": 0.7781,
+      "step": 645
+    },
+    {
+      "epoch": 0.7146017699115044,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.21349557522124e-06,
+      "loss": 0.8181,
+      "step": 646
+    },
+    {
+      "epoch": 0.7157079646017699,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.210730088495576e-06,
+      "loss": 0.8242,
+      "step": 647
+    },
+    {
+      "epoch": 0.7168141592920354,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.207964601769912e-06,
+      "loss": 0.8388,
+      "step": 648
+    },
+    {
+      "epoch": 0.7179203539823009,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 8.205199115044248e-06,
+      "loss": 0.8223,
+      "step": 649
+    },
+    {
+      "epoch": 0.7190265486725663,
+      "grad_norm": 0.123046875,
+      "learning_rate": 8.202433628318584e-06,
+      "loss": 0.845,
+      "step": 650
+    },
+    {
+      "epoch": 0.7201327433628318,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 8.199668141592922e-06,
+      "loss": 0.8311,
+      "step": 651
+    },
+    {
+      "epoch": 0.7212389380530974,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 8.196902654867257e-06,
+      "loss": 0.8016,
+      "step": 652
+    },
+    {
+      "epoch": 0.7223451327433629,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.194137168141594e-06,
+      "loss": 0.8173,
+      "step": 653
+    },
+    {
+      "epoch": 0.7234513274336283,
+      "grad_norm": 0.099609375,
+      "learning_rate": 8.191371681415929e-06,
+      "loss": 0.8066,
+      "step": 654
+    },
+    {
+      "epoch": 0.7245575221238938,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 8.188606194690267e-06,
+      "loss": 0.8023,
+      "step": 655
+    },
+    {
+      "epoch": 0.7256637168141593,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.185840707964603e-06,
+      "loss": 0.8087,
+      "step": 656
+    },
+    {
+      "epoch": 0.7267699115044248,
+      "grad_norm": 0.0859375,
+      "learning_rate": 8.183075221238939e-06,
+      "loss": 0.7654,
+      "step": 657
+    },
+    {
+      "epoch": 0.7278761061946902,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.180309734513275e-06,
+      "loss": 0.8418,
+      "step": 658
+    },
+    {
+      "epoch": 0.7289823008849557,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.177544247787611e-06,
+      "loss": 0.8419,
+      "step": 659
+    },
+    {
+      "epoch": 0.7300884955752213,
+      "grad_norm": 0.205078125,
+      "learning_rate": 8.174778761061947e-06,
+      "loss": 0.7717,
+      "step": 660
+    },
+    {
+      "epoch": 0.7311946902654868,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 8.172013274336283e-06,
+      "loss": 0.7719,
+      "step": 661
+    },
+    {
+      "epoch": 0.7323008849557522,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.16924778761062e-06,
+      "loss": 0.821,
+      "step": 662
+    },
+    {
+      "epoch": 0.7334070796460177,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 8.166482300884957e-06,
+      "loss": 0.7988,
+      "step": 663
+    },
+    {
+      "epoch": 0.7345132743362832,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 8.163716814159292e-06,
+      "loss": 0.7823,
+      "step": 664
+    },
+    {
+      "epoch": 0.7356194690265486,
+      "grad_norm": 0.091796875,
+      "learning_rate": 8.16095132743363e-06,
+      "loss": 0.7912,
+      "step": 665
+    },
+    {
+      "epoch": 0.7367256637168141,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 8.158185840707966e-06,
+      "loss": 0.8055,
+      "step": 666
+    },
+    {
+      "epoch": 0.7378318584070797,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.155420353982302e-06,
+      "loss": 0.823,
+      "step": 667
+    },
+    {
+      "epoch": 0.7389380530973452,
+      "grad_norm": 0.107421875,
+      "learning_rate": 8.152654867256638e-06,
+      "loss": 0.7857,
+      "step": 668
+    },
+    {
+      "epoch": 0.7400442477876106,
+      "grad_norm": 0.12109375,
+      "learning_rate": 8.149889380530974e-06,
+      "loss": 0.824,
+      "step": 669
+    },
+    {
+      "epoch": 0.7411504424778761,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.14712389380531e-06,
+      "loss": 0.8127,
+      "step": 670
+    },
+    {
+      "epoch": 0.7422566371681416,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.144358407079646e-06,
+      "loss": 0.7991,
+      "step": 671
+    },
+    {
+      "epoch": 0.7433628318584071,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.141592920353984e-06,
+      "loss": 0.7724,
+      "step": 672
+    },
+    {
+      "epoch": 0.7444690265486725,
+      "grad_norm": 0.10546875,
+      "learning_rate": 8.138827433628319e-06,
+      "loss": 0.7891,
+      "step": 673
+    },
+    {
+      "epoch": 0.745575221238938,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 8.136061946902656e-06,
+      "loss": 0.7939,
+      "step": 674
+    },
+    {
+      "epoch": 0.7466814159292036,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 8.133296460176991e-06,
+      "loss": 0.8113,
+      "step": 675
+    },
+    {
+      "epoch": 0.7477876106194691,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.130530973451329e-06,
+      "loss": 0.7954,
+      "step": 676
+    },
+    {
+      "epoch": 0.7488938053097345,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.127765486725663e-06,
+      "loss": 0.8043,
+      "step": 677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 0.8213,
+      "step": 678
+    },
+    {
+      "epoch": 0.7511061946902655,
+      "grad_norm": 0.091796875,
+      "learning_rate": 8.122234513274337e-06,
+      "loss": 0.7792,
+      "step": 679
+    },
+    {
+      "epoch": 0.7522123893805309,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.119469026548673e-06,
+      "loss": 0.8366,
+      "step": 680
+    },
+    {
+      "epoch": 0.7533185840707964,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.11670353982301e-06,
+      "loss": 0.8216,
+      "step": 681
+    },
+    {
+      "epoch": 0.754424778761062,
+      "grad_norm": 0.08837890625,
+      "learning_rate": 8.113938053097345e-06,
+      "loss": 0.7821,
+      "step": 682
+    },
+    {
+      "epoch": 0.7555309734513275,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.111172566371682e-06,
+      "loss": 0.7935,
+      "step": 683
+    },
+    {
+      "epoch": 0.7566371681415929,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.108407079646018e-06,
+      "loss": 0.8147,
+      "step": 684
+    },
+    {
+      "epoch": 0.7577433628318584,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 8.105641592920354e-06,
+      "loss": 0.8314,
+      "step": 685
+    },
+    {
+      "epoch": 0.7588495575221239,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.102876106194692e-06,
+      "loss": 0.8017,
+      "step": 686
+    },
+    {
+      "epoch": 0.7599557522123894,
+      "grad_norm": 0.091796875,
+      "learning_rate": 8.100110619469028e-06,
+      "loss": 0.7776,
+      "step": 687
+    },
+    {
+      "epoch": 0.7610619469026548,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 8.097345132743364e-06,
+      "loss": 0.8183,
+      "step": 688
+    },
+    {
+      "epoch": 0.7621681415929203,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 8.0945796460177e-06,
+      "loss": 0.7659,
+      "step": 689
+    },
+    {
+      "epoch": 0.7632743362831859,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 8.091814159292036e-06,
+      "loss": 0.82,
+      "step": 690
+    },
+    {
+      "epoch": 0.7643805309734514,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.089048672566372e-06,
+      "loss": 0.8103,
+      "step": 691
+    },
+    {
+      "epoch": 0.7654867256637168,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.086283185840708e-06,
+      "loss": 0.794,
+      "step": 692
+    },
+    {
+      "epoch": 0.7665929203539823,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 8.083517699115045e-06,
+      "loss": 0.8052,
+      "step": 693
+    },
+    {
+      "epoch": 0.7676991150442478,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.08075221238938e-06,
+      "loss": 0.8367,
+      "step": 694
+    },
+    {
+      "epoch": 0.7688053097345132,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.077986725663719e-06,
+      "loss": 0.7813,
+      "step": 695
+    },
+    {
+      "epoch": 0.7699115044247787,
+      "grad_norm": 0.091796875,
+      "learning_rate": 8.075221238938053e-06,
+      "loss": 0.7647,
+      "step": 696
+    },
+    {
+      "epoch": 0.7710176991150443,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.07245575221239e-06,
+      "loss": 0.7917,
+      "step": 697
+    },
+    {
+      "epoch": 0.7721238938053098,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 8.069690265486725e-06,
+      "loss": 0.8125,
+      "step": 698
+    },
+    {
+      "epoch": 0.7732300884955752,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 8.066924778761063e-06,
+      "loss": 0.7879,
+      "step": 699
+    },
+    {
+      "epoch": 0.7743362831858407,
+      "grad_norm": 0.109375,
+      "learning_rate": 8.064159292035399e-06,
+      "loss": 0.8156,
+      "step": 700
+    },
+    {
+      "epoch": 0.7754424778761062,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 8.061393805309735e-06,
+      "loss": 0.7847,
+      "step": 701
+    },
+    {
+      "epoch": 0.7765486725663717,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 8.058628318584071e-06,
+      "loss": 0.823,
+      "step": 702
+    },
+    {
+      "epoch": 0.7776548672566371,
+      "grad_norm": 0.08984375,
+      "learning_rate": 8.055862831858408e-06,
+      "loss": 0.7435,
+      "step": 703
+    },
+    {
+      "epoch": 0.7787610619469026,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 8.053097345132744e-06,
+      "loss": 0.7767,
+      "step": 704
+    },
+    {
+      "epoch": 0.7798672566371682,
+      "grad_norm": 0.09375,
+      "learning_rate": 8.05033185840708e-06,
+      "loss": 0.7688,
+      "step": 705
+    },
+    {
+      "epoch": 0.7809734513274337,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.047566371681416e-06,
+      "loss": 0.7321,
+      "step": 706
+    },
+    {
+      "epoch": 0.7820796460176991,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 8.044800884955752e-06,
+      "loss": 0.8054,
+      "step": 707
+    },
+    {
+      "epoch": 0.7831858407079646,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.04203539823009e-06,
+      "loss": 0.7813,
+      "step": 708
+    },
+    {
+      "epoch": 0.7842920353982301,
+      "grad_norm": 0.091796875,
+      "learning_rate": 8.039269911504426e-06,
+      "loss": 0.7788,
+      "step": 709
+    },
+    {
+      "epoch": 0.7853982300884956,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 8.036504424778762e-06,
+      "loss": 0.7836,
+      "step": 710
+    },
+    {
+      "epoch": 0.786504424778761,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.033738938053098e-06,
+      "loss": 0.8177,
+      "step": 711
+    },
+    {
+      "epoch": 0.7876106194690266,
+      "grad_norm": 0.08544921875,
+      "learning_rate": 8.030973451327434e-06,
+      "loss": 0.8077,
+      "step": 712
+    },
+    {
+      "epoch": 0.7887168141592921,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 8.02820796460177e-06,
+      "loss": 0.8156,
+      "step": 713
+    },
+    {
+      "epoch": 0.7898230088495575,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 8.025442477876107e-06,
+      "loss": 0.7796,
+      "step": 714
+    },
+    {
+      "epoch": 0.790929203539823,
+      "grad_norm": 0.099609375,
+      "learning_rate": 8.022676991150443e-06,
+      "loss": 0.7771,
+      "step": 715
+    },
+    {
+      "epoch": 0.7920353982300885,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 8.01991150442478e-06,
+      "loss": 0.7921,
+      "step": 716
+    },
+    {
+      "epoch": 0.793141592920354,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.017146017699115e-06,
+      "loss": 0.8193,
+      "step": 717
+    },
+    {
+      "epoch": 0.7942477876106194,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.014380530973453e-06,
+      "loss": 0.7846,
+      "step": 718
+    },
+    {
+      "epoch": 0.7953539823008849,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 8.011615044247787e-06,
+      "loss": 0.8789,
+      "step": 719
+    },
+    {
+      "epoch": 0.7964601769911505,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 8.008849557522125e-06,
+      "loss": 0.77,
+      "step": 720
+    },
+    {
+      "epoch": 0.797566371681416,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 8.006084070796461e-06,
+      "loss": 0.7953,
+      "step": 721
+    },
+    {
+      "epoch": 0.7986725663716814,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 8.003318584070797e-06,
+      "loss": 0.7882,
+      "step": 722
+    },
+    {
+      "epoch": 0.7997787610619469,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.000553097345133e-06,
+      "loss": 0.8293,
+      "step": 723
+    },
+    {
+      "epoch": 0.8008849557522124,
+      "grad_norm": 0.115234375,
+      "learning_rate": 7.99778761061947e-06,
+      "loss": 0.7941,
+      "step": 724
+    },
+    {
+      "epoch": 0.8019911504424779,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.995022123893806e-06,
+      "loss": 0.7618,
+      "step": 725
+    },
+    {
+      "epoch": 0.8030973451327433,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.992256637168142e-06,
+      "loss": 0.825,
+      "step": 726
+    },
+    {
+      "epoch": 0.8042035398230089,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.98949115044248e-06,
+      "loss": 0.7839,
+      "step": 727
+    },
+    {
+      "epoch": 0.8053097345132744,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.986725663716814e-06,
+      "loss": 0.7806,
+      "step": 728
+    },
+    {
+      "epoch": 0.8064159292035398,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.983960176991152e-06,
+      "loss": 0.8017,
+      "step": 729
+    },
+    {
+      "epoch": 0.8075221238938053,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.981194690265486e-06,
+      "loss": 0.8704,
+      "step": 730
+    },
+    {
+      "epoch": 0.8086283185840708,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.978429203539824e-06,
+      "loss": 0.8549,
+      "step": 731
+    },
+    {
+      "epoch": 0.8097345132743363,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 7.97566371681416e-06,
+      "loss": 0.7918,
+      "step": 732
+    },
+    {
+      "epoch": 0.8108407079646017,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.972898230088496e-06,
+      "loss": 0.7863,
+      "step": 733
+    },
+    {
+      "epoch": 0.8119469026548672,
+      "grad_norm": 0.1171875,
+      "learning_rate": 7.970132743362833e-06,
+      "loss": 0.8723,
+      "step": 734
+    },
+    {
+      "epoch": 0.8130530973451328,
+      "grad_norm": 0.107421875,
+      "learning_rate": 7.967367256637169e-06,
+      "loss": 0.8013,
+      "step": 735
+    },
+    {
+      "epoch": 0.8141592920353983,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.964601769911505e-06,
+      "loss": 0.8097,
+      "step": 736
+    },
+    {
+      "epoch": 0.8152654867256637,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.961836283185841e-06,
+      "loss": 0.8117,
+      "step": 737
+    },
+    {
+      "epoch": 0.8163716814159292,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 7.959070796460177e-06,
+      "loss": 0.7894,
+      "step": 738
+    },
+    {
+      "epoch": 0.8174778761061947,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.956305309734515e-06,
+      "loss": 0.8117,
+      "step": 739
+    },
+    {
+      "epoch": 0.8185840707964602,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.95353982300885e-06,
+      "loss": 0.814,
+      "step": 740
+    },
+    {
+      "epoch": 0.8196902654867256,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.950774336283187e-06,
+      "loss": 0.8425,
+      "step": 741
+    },
+    {
+      "epoch": 0.8207964601769911,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 7.948008849557523e-06,
+      "loss": 0.7641,
+      "step": 742
+    },
+    {
+      "epoch": 0.8219026548672567,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.94524336283186e-06,
+      "loss": 0.7799,
+      "step": 743
+    },
+    {
+      "epoch": 0.8230088495575221,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 7.942477876106195e-06,
+      "loss": 0.8097,
+      "step": 744
+    },
+    {
+      "epoch": 0.8241150442477876,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.939712389380532e-06,
+      "loss": 0.7934,
+      "step": 745
+    },
+    {
+      "epoch": 0.8252212389380531,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.936946902654868e-06,
+      "loss": 0.8074,
+      "step": 746
+    },
+    {
+      "epoch": 0.8263274336283186,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.934181415929204e-06,
+      "loss": 0.8232,
+      "step": 747
+    },
+    {
+      "epoch": 0.827433628318584,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 7.931415929203542e-06,
+      "loss": 0.7956,
+      "step": 748
+    },
+    {
+      "epoch": 0.8285398230088495,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 7.928650442477876e-06,
+      "loss": 0.7751,
+      "step": 749
+    },
+    {
+      "epoch": 0.8296460176991151,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.925884955752214e-06,
+      "loss": 0.7923,
+      "step": 750
+    },
+    {
+      "epoch": 0.8307522123893806,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 7.923119469026548e-06,
+      "loss": 0.7879,
+      "step": 751
+    },
+    {
+      "epoch": 0.831858407079646,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 7.920353982300886e-06,
+      "loss": 0.7493,
+      "step": 752
+    },
+    {
+      "epoch": 0.8329646017699115,
+      "grad_norm": 0.123046875,
+      "learning_rate": 7.91758849557522e-06,
+      "loss": 0.862,
+      "step": 753
+    },
+    {
+      "epoch": 0.834070796460177,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 7.914823008849558e-06,
+      "loss": 0.7811,
+      "step": 754
+    },
+    {
+      "epoch": 0.8351769911504425,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 7.912057522123895e-06,
+      "loss": 0.7316,
+      "step": 755
+    },
+    {
+      "epoch": 0.8362831858407079,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 7.90929203539823e-06,
+      "loss": 0.768,
+      "step": 756
+    },
+    {
+      "epoch": 0.8373893805309734,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 7.906526548672567e-06,
+      "loss": 0.7744,
+      "step": 757
+    },
+    {
+      "epoch": 0.838495575221239,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.903761061946903e-06,
+      "loss": 0.7967,
+      "step": 758
+    },
+    {
+      "epoch": 0.8396017699115044,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 7.900995575221239e-06,
+      "loss": 0.7906,
+      "step": 759
+    },
+    {
+      "epoch": 0.8407079646017699,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.898230088495575e-06,
+      "loss": 0.7468,
+      "step": 760
+    },
+    {
+      "epoch": 0.8418141592920354,
+      "grad_norm": 0.10546875,
+      "learning_rate": 7.895464601769911e-06,
+      "loss": 0.8613,
+      "step": 761
+    },
+    {
+      "epoch": 0.8429203539823009,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.892699115044249e-06,
+      "loss": 0.8098,
+      "step": 762
+    },
+    {
+      "epoch": 0.8440265486725663,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.889933628318585e-06,
+      "loss": 0.7901,
+      "step": 763
+    },
+    {
+      "epoch": 0.8451327433628318,
+      "grad_norm": 0.103515625,
+      "learning_rate": 7.887168141592921e-06,
+      "loss": 0.8162,
+      "step": 764
+    },
+    {
+      "epoch": 0.8462389380530974,
+      "grad_norm": 0.09375,
+      "learning_rate": 7.884402654867257e-06,
+      "loss": 0.7757,
+      "step": 765
+    },
+    {
+      "epoch": 0.8473451327433629,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.881637168141594e-06,
+      "loss": 0.8093,
+      "step": 766
+    },
+    {
+      "epoch": 0.8484513274336283,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.87887168141593e-06,
+      "loss": 0.7795,
+      "step": 767
+    },
+    {
+      "epoch": 0.8495575221238938,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.876106194690266e-06,
+      "loss": 0.812,
+      "step": 768
+    },
+    {
+      "epoch": 0.8506637168141593,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 7.873340707964604e-06,
+      "loss": 0.7725,
+      "step": 769
+    },
+    {
+      "epoch": 0.8517699115044248,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 7.870575221238938e-06,
+      "loss": 0.7831,
+      "step": 770
+    },
+    {
+      "epoch": 0.8528761061946902,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.867809734513276e-06,
+      "loss": 0.8032,
+      "step": 771
+    },
+    {
+      "epoch": 0.8539823008849557,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.86504424778761e-06,
+      "loss": 0.8097,
+      "step": 772
+    },
+    {
+      "epoch": 0.8550884955752213,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 7.862278761061948e-06,
+      "loss": 0.8034,
+      "step": 773
+    },
+    {
+      "epoch": 0.8561946902654868,
+      "grad_norm": 0.109375,
+      "learning_rate": 7.859513274336283e-06,
+      "loss": 0.7932,
+      "step": 774
+    },
+    {
+      "epoch": 0.8573008849557522,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.85674778761062e-06,
+      "loss": 0.787,
+      "step": 775
+    },
+    {
+      "epoch": 0.8584070796460177,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 7.853982300884957e-06,
+      "loss": 0.7583,
+      "step": 776
+    },
+    {
+      "epoch": 0.8595132743362832,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.851216814159293e-06,
+      "loss": 0.794,
+      "step": 777
+    },
+    {
+      "epoch": 0.8606194690265486,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 7.848451327433629e-06,
+      "loss": 0.7937,
+      "step": 778
+    },
+    {
+      "epoch": 0.8617256637168141,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.845685840707965e-06,
+      "loss": 0.8503,
+      "step": 779
+    },
+    {
+      "epoch": 0.8628318584070797,
+      "grad_norm": 0.11328125,
+      "learning_rate": 7.842920353982301e-06,
+      "loss": 0.8032,
+      "step": 780
+    },
+    {
+      "epoch": 0.8639380530973452,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.840154867256637e-06,
+      "loss": 0.7588,
+      "step": 781
+    },
+    {
+      "epoch": 0.8650442477876106,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.837389380530975e-06,
+      "loss": 0.8995,
+      "step": 782
+    },
+    {
+      "epoch": 0.8661504424778761,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.83462389380531e-06,
+      "loss": 0.7923,
+      "step": 783
+    },
+    {
+      "epoch": 0.8672566371681416,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 7.831858407079647e-06,
+      "loss": 0.7528,
+      "step": 784
+    },
+    {
+      "epoch": 0.8683628318584071,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.829092920353983e-06,
+      "loss": 0.7906,
+      "step": 785
+    },
+    {
+      "epoch": 0.8694690265486725,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.82632743362832e-06,
+      "loss": 0.8124,
+      "step": 786
+    },
+    {
+      "epoch": 0.870575221238938,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.823561946902656e-06,
+      "loss": 0.765,
+      "step": 787
+    },
+    {
+      "epoch": 0.8716814159292036,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.820796460176992e-06,
+      "loss": 0.8097,
+      "step": 788
+    },
+    {
+      "epoch": 0.8727876106194691,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.818030973451328e-06,
+      "loss": 0.7986,
+      "step": 789
+    },
+    {
+      "epoch": 0.8738938053097345,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.815265486725664e-06,
+      "loss": 0.7865,
+      "step": 790
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 7.8125e-06,
+      "loss": 0.7789,
+      "step": 791
+    },
+    {
+      "epoch": 0.8761061946902655,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.809734513274338e-06,
+      "loss": 0.8168,
+      "step": 792
+    },
+    {
+      "epoch": 0.8772123893805309,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 7.806969026548672e-06,
+      "loss": 0.758,
+      "step": 793
+    },
+    {
+      "epoch": 0.8783185840707964,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 7.80420353982301e-06,
+      "loss": 0.7455,
+      "step": 794
+    },
+    {
+      "epoch": 0.879424778761062,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 7.801438053097345e-06,
+      "loss": 0.7781,
+      "step": 795
+    },
+    {
+      "epoch": 0.8805309734513275,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 7.798672566371682e-06,
+      "loss": 0.806,
+      "step": 796
+    },
+    {
+      "epoch": 0.8816371681415929,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.795907079646019e-06,
+      "loss": 0.7764,
+      "step": 797
+    },
+    {
+      "epoch": 0.8827433628318584,
+      "grad_norm": 0.11328125,
+      "learning_rate": 7.793141592920355e-06,
+      "loss": 0.8144,
+      "step": 798
+    },
+    {
+      "epoch": 0.8838495575221239,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 7.790376106194691e-06,
+      "loss": 0.7941,
+      "step": 799
+    },
+    {
+      "epoch": 0.8849557522123894,
+      "grad_norm": 0.09375,
+      "learning_rate": 7.787610619469027e-06,
+      "loss": 0.792,
+      "step": 800
+    },
+    {
+      "epoch": 0.8860619469026548,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.784845132743363e-06,
+      "loss": 0.8884,
+      "step": 801
+    },
+    {
+      "epoch": 0.8871681415929203,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 7.7820796460177e-06,
+      "loss": 0.7957,
+      "step": 802
+    },
+    {
+      "epoch": 0.8882743362831859,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 7.779314159292037e-06,
+      "loss": 0.7722,
+      "step": 803
+    },
+    {
+      "epoch": 0.8893805309734514,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.776548672566371e-06,
+      "loss": 0.7912,
+      "step": 804
+    },
+    {
+      "epoch": 0.8904867256637168,
+      "grad_norm": 0.103515625,
+      "learning_rate": 7.77378318584071e-06,
+      "loss": 0.7921,
+      "step": 805
+    },
+    {
+      "epoch": 0.8915929203539823,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 7.771017699115044e-06,
+      "loss": 0.7901,
+      "step": 806
+    },
+    {
+      "epoch": 0.8926991150442478,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 7.768252212389382e-06,
+      "loss": 0.7974,
+      "step": 807
+    },
+    {
+      "epoch": 0.8938053097345132,
+      "grad_norm": 0.09375,
+      "learning_rate": 7.765486725663718e-06,
+      "loss": 0.7852,
+      "step": 808
+    },
+    {
+      "epoch": 0.8949115044247787,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.762721238938054e-06,
+      "loss": 0.8432,
+      "step": 809
+    },
+    {
+      "epoch": 0.8960176991150443,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 7.75995575221239e-06,
+      "loss": 0.7849,
+      "step": 810
+    },
+    {
+      "epoch": 0.8971238938053098,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.757190265486726e-06,
+      "loss": 0.7753,
+      "step": 811
+    },
+    {
+      "epoch": 0.8982300884955752,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.754424778761062e-06,
+      "loss": 0.7861,
+      "step": 812
+    },
+    {
+      "epoch": 0.8993362831858407,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 7.751659292035398e-06,
+      "loss": 0.785,
+      "step": 813
+    },
+    {
+      "epoch": 0.9004424778761062,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.748893805309734e-06,
+      "loss": 0.7624,
+      "step": 814
+    },
+    {
+      "epoch": 0.9015486725663717,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.746128318584072e-06,
+      "loss": 0.7716,
+      "step": 815
+    },
+    {
+      "epoch": 0.9026548672566371,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 7.743362831858407e-06,
+      "loss": 0.7948,
+      "step": 816
+    },
+    {
+      "epoch": 0.9037610619469026,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.740597345132745e-06,
+      "loss": 0.7978,
+      "step": 817
+    },
+    {
+      "epoch": 0.9048672566371682,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.73783185840708e-06,
+      "loss": 0.8226,
+      "step": 818
+    },
+    {
+      "epoch": 0.9059734513274337,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.735066371681417e-06,
+      "loss": 0.8104,
+      "step": 819
+    },
+    {
+      "epoch": 0.9070796460176991,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 7.732300884955753e-06,
+      "loss": 0.7997,
+      "step": 820
+    },
+    {
+      "epoch": 0.9081858407079646,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 7.729535398230089e-06,
+      "loss": 0.7535,
+      "step": 821
+    },
+    {
+      "epoch": 0.9092920353982301,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.726769911504425e-06,
+      "loss": 0.8201,
+      "step": 822
+    },
+    {
+      "epoch": 0.9103982300884956,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 7.724004424778761e-06,
+      "loss": 0.7601,
+      "step": 823
+    },
+    {
+      "epoch": 0.911504424778761,
+      "grad_norm": 0.103515625,
+      "learning_rate": 7.721238938053099e-06,
+      "loss": 0.7835,
+      "step": 824
+    },
+    {
+      "epoch": 0.9126106194690266,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.718473451327434e-06,
+      "loss": 0.8018,
+      "step": 825
+    },
+    {
+      "epoch": 0.9137168141592921,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.715707964601771e-06,
+      "loss": 0.8618,
+      "step": 826
+    },
+    {
+      "epoch": 0.9148230088495575,
+      "grad_norm": 0.10546875,
+      "learning_rate": 7.712942477876106e-06,
+      "loss": 0.7904,
+      "step": 827
+    },
+    {
+      "epoch": 0.915929203539823,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 7.710176991150444e-06,
+      "loss": 0.7919,
+      "step": 828
+    },
+    {
+      "epoch": 0.9170353982300885,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 7.70741150442478e-06,
+      "loss": 0.7709,
+      "step": 829
+    },
+    {
+      "epoch": 0.918141592920354,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.704646017699116e-06,
+      "loss": 0.7925,
+      "step": 830
+    },
+    {
+      "epoch": 0.9192477876106194,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 7.701880530973452e-06,
+      "loss": 0.7625,
+      "step": 831
+    },
+    {
+      "epoch": 0.9203539823008849,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 7.699115044247788e-06,
+      "loss": 0.8081,
+      "step": 832
+    },
+    {
+      "epoch": 0.9214601769911505,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 7.696349557522124e-06,
+      "loss": 0.7613,
+      "step": 833
+    },
+    {
+      "epoch": 0.922566371681416,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 7.69358407079646e-06,
+      "loss": 0.7466,
+      "step": 834
+    },
+    {
+      "epoch": 0.9236725663716814,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.690818584070796e-06,
+      "loss": 0.7835,
+      "step": 835
+    },
+    {
+      "epoch": 0.9247787610619469,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 7.688053097345133e-06,
+      "loss": 0.7788,
+      "step": 836
+    },
+    {
+      "epoch": 0.9258849557522124,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 7.68528761061947e-06,
+      "loss": 0.7883,
+      "step": 837
+    },
+    {
+      "epoch": 0.9269911504424779,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 7.682522123893807e-06,
+      "loss": 0.8219,
+      "step": 838
+    },
+    {
+      "epoch": 0.9280973451327433,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 7.679756637168143e-06,
+      "loss": 0.7471,
+      "step": 839
+    },
+    {
+      "epoch": 0.9292035398230089,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 7.676991150442479e-06,
+      "loss": 0.8081,
+      "step": 840
+    },
+    {
+      "epoch": 0.9303097345132744,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 7.674225663716815e-06,
+      "loss": 0.738,
+      "step": 841
+    },
+    {
+      "epoch": 0.9314159292035398,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.671460176991151e-06,
+      "loss": 0.7876,
+      "step": 842
+    },
+    {
+      "epoch": 0.9325221238938053,
+      "grad_norm": 0.08984375,
+      "learning_rate": 7.668694690265487e-06,
+      "loss": 0.7718,
+      "step": 843
+    },
+    {
+      "epoch": 0.9336283185840708,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.665929203539823e-06,
+      "loss": 0.8037,
+      "step": 844
+    },
+    {
+      "epoch": 0.9347345132743363,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.663163716814161e-06,
+      "loss": 0.8095,
+      "step": 845
+    },
+    {
+      "epoch": 0.9358407079646017,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.660398230088496e-06,
+      "loss": 0.7987,
+      "step": 846
+    },
+    {
+      "epoch": 0.9369469026548672,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 7.657632743362833e-06,
+      "loss": 0.8356,
+      "step": 847
+    },
+    {
+      "epoch": 0.9380530973451328,
+      "grad_norm": 0.1171875,
+      "learning_rate": 7.654867256637168e-06,
+      "loss": 0.7923,
+      "step": 848
+    },
+    {
+      "epoch": 0.9391592920353983,
+      "grad_norm": 0.091796875,
+      "learning_rate": 7.652101769911506e-06,
+      "loss": 0.7622,
+      "step": 849
+    },
+    {
+      "epoch": 0.9402654867256637,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 7.64933628318584e-06,
+      "loss": 0.7848,
+      "step": 850
+    },
+    {
+      "epoch": 0.9413716814159292,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 7.646570796460178e-06,
+      "loss": 0.7463,
+      "step": 851
+    },
+    {
+      "epoch": 0.9424778761061947,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 7.643805309734514e-06,
+      "loss": 0.7997,
+      "step": 852
+    },
+    {
+      "epoch": 0.9435840707964602,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.64103982300885e-06,
+      "loss": 0.7875,
+      "step": 853
+    },
+    {
+      "epoch": 0.9446902654867256,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.638274336283186e-06,
+      "loss": 0.7782,
+      "step": 854
+    },
+    {
+      "epoch": 0.9457964601769911,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 7.635508849557522e-06,
+      "loss": 0.8038,
+      "step": 855
+    },
+    {
+      "epoch": 0.9469026548672567,
+      "grad_norm": 0.095703125,
+      "learning_rate": 7.632743362831859e-06,
+      "loss": 0.7642,
+      "step": 856
+    },
+    {
+      "epoch": 0.9480088495575221,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.629977876106195e-06,
+      "loss": 0.7684,
+      "step": 857
+    },
+    {
+      "epoch": 0.9491150442477876,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 7.627212389380532e-06,
+      "loss": 0.7846,
+      "step": 858
+    },
+    {
+      "epoch": 0.9502212389380531,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.624446902654869e-06,
+      "loss": 0.8284,
+      "step": 859
+    },
+    {
+      "epoch": 0.9513274336283186,
+      "grad_norm": 0.08984375,
+      "learning_rate": 7.621681415929204e-06,
+      "loss": 0.7767,
+      "step": 860
+    },
+    {
+      "epoch": 0.952433628318584,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 7.618915929203541e-06,
+      "loss": 0.8293,
+      "step": 861
+    },
+    {
+      "epoch": 0.9535398230088495,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 7.616150442477876e-06,
+      "loss": 0.7763,
+      "step": 862
+    },
+    {
+      "epoch": 0.9546460176991151,
+      "grad_norm": 0.09130859375,
+      "learning_rate": 7.613384955752213e-06,
+      "loss": 0.7531,
+      "step": 863
+    },
+    {
+      "epoch": 0.9557522123893806,
+      "grad_norm": 0.10546875,
+      "learning_rate": 7.610619469026549e-06,
+      "loss": 0.7928,
+      "step": 864
+    },
+    {
+      "epoch": 0.956858407079646,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.607853982300885e-06,
+      "loss": 0.7837,
+      "step": 865
+    },
+    {
+      "epoch": 0.9579646017699115,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 7.6050884955752215e-06,
+      "loss": 0.7642,
+      "step": 866
+    },
+    {
+      "epoch": 0.959070796460177,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 7.6023230088495584e-06,
+      "loss": 0.8056,
+      "step": 867
+    },
+    {
+      "epoch": 0.9601769911504425,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.5995575221238946e-06,
+      "loss": 0.789,
+      "step": 868
+    },
+    {
+      "epoch": 0.9612831858407079,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.596792035398231e-06,
+      "loss": 0.7686,
+      "step": 869
+    },
+    {
+      "epoch": 0.9623893805309734,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.594026548672568e-06,
+      "loss": 0.7833,
+      "step": 870
+    },
+    {
+      "epoch": 0.963495575221239,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 7.591261061946903e-06,
+      "loss": 0.8023,
+      "step": 871
+    },
+    {
+      "epoch": 0.9646017699115044,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.58849557522124e-06,
+      "loss": 0.8321,
+      "step": 872
+    },
+    {
+      "epoch": 0.9657079646017699,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.585730088495575e-06,
+      "loss": 0.772,
+      "step": 873
+    },
+    {
+      "epoch": 0.9668141592920354,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 7.582964601769912e-06,
+      "loss": 0.7741,
+      "step": 874
+    },
+    {
+      "epoch": 0.9679203539823009,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 7.580199115044249e-06,
+      "loss": 0.7791,
+      "step": 875
+    },
+    {
+      "epoch": 0.9690265486725663,
+      "grad_norm": 0.09375,
+      "learning_rate": 7.5774336283185844e-06,
+      "loss": 0.7798,
+      "step": 876
+    },
+    {
+      "epoch": 0.9701327433628318,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 7.574668141592921e-06,
+      "loss": 0.7987,
+      "step": 877
+    },
+    {
+      "epoch": 0.9712389380530974,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.571902654867257e-06,
+      "loss": 0.7682,
+      "step": 878
+    },
+    {
+      "epoch": 0.9723451327433629,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.569137168141594e-06,
+      "loss": 0.8299,
+      "step": 879
+    },
+    {
+      "epoch": 0.9734513274336283,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.56637168141593e-06,
+      "loss": 0.7879,
+      "step": 880
+    },
+    {
+      "epoch": 0.9745575221238938,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.563606194690266e-06,
+      "loss": 0.7566,
+      "step": 881
+    },
+    {
+      "epoch": 0.9756637168141593,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.560840707964603e-06,
+      "loss": 0.772,
+      "step": 882
+    },
+    {
+      "epoch": 0.9767699115044248,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.558075221238938e-06,
+      "loss": 0.7672,
+      "step": 883
+    },
+    {
+      "epoch": 0.9778761061946902,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.555309734513275e-06,
+      "loss": 0.8005,
+      "step": 884
+    },
+    {
+      "epoch": 0.9789823008849557,
+      "grad_norm": 0.095703125,
+      "learning_rate": 7.552544247787611e-06,
+      "loss": 0.7374,
+      "step": 885
+    },
+    {
+      "epoch": 0.9800884955752213,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.549778761061947e-06,
+      "loss": 0.7848,
+      "step": 886
+    },
+    {
+      "epoch": 0.9811946902654868,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 7.5470132743362835e-06,
+      "loss": 0.8065,
+      "step": 887
+    },
+    {
+      "epoch": 0.9823008849557522,
+      "grad_norm": 0.12109375,
+      "learning_rate": 7.5442477876106205e-06,
+      "loss": 0.781,
+      "step": 888
+    },
+    {
+      "epoch": 0.9834070796460177,
+      "grad_norm": 0.107421875,
+      "learning_rate": 7.541482300884957e-06,
+      "loss": 0.7816,
+      "step": 889
+    },
+    {
+      "epoch": 0.9845132743362832,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.538716814159293e-06,
+      "loss": 0.8076,
+      "step": 890
+    },
+    {
+      "epoch": 0.9856194690265486,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 7.53595132743363e-06,
+      "loss": 0.7779,
+      "step": 891
+    },
+    {
+      "epoch": 0.9867256637168141,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 7.533185840707965e-06,
+      "loss": 0.7624,
+      "step": 892
+    },
+    {
+      "epoch": 0.9878318584070797,
+      "grad_norm": 0.109375,
+      "learning_rate": 7.530420353982302e-06,
+      "loss": 0.8129,
+      "step": 893
+    },
+    {
+      "epoch": 0.9889380530973452,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.527654867256637e-06,
+      "loss": 0.792,
+      "step": 894
+    },
+    {
+      "epoch": 0.9900442477876106,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 7.524889380530974e-06,
+      "loss": 0.789,
+      "step": 895
+    },
+    {
+      "epoch": 0.9911504424778761,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 7.5221238938053095e-06,
+      "loss": 0.8057,
+      "step": 896
+    },
+    {
+      "epoch": 0.9922566371681416,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.5193584070796465e-06,
+      "loss": 0.8332,
+      "step": 897
+    },
+    {
+      "epoch": 0.9933628318584071,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.5165929203539834e-06,
+      "loss": 0.803,
+      "step": 898
+    },
+    {
+      "epoch": 0.9944690265486725,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 7.513827433628319e-06,
+      "loss": 0.7568,
+      "step": 899
+    },
+    {
+      "epoch": 0.995575221238938,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.511061946902656e-06,
+      "loss": 0.7369,
+      "step": 900
+    },
+    {
+      "epoch": 0.9966814159292036,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 7.508296460176992e-06,
+      "loss": 0.8018,
+      "step": 901
+    },
+    {
+      "epoch": 0.9977876106194691,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 7.505530973451328e-06,
+      "loss": 0.7925,
+      "step": 902
+    },
+    {
+      "epoch": 0.9988938053097345,
+      "grad_norm": 0.107421875,
+      "learning_rate": 7.502765486725664e-06,
+      "loss": 0.7932,
+      "step": 903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.7496,
+      "step": 904
+    },
+    {
+      "epoch": 1.0011061946902655,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.497234513274337e-06,
+      "loss": 0.7853,
+      "step": 905
+    },
+    {
+      "epoch": 1.002212389380531,
+      "grad_norm": 0.095703125,
+      "learning_rate": 7.494469026548673e-06,
+      "loss": 0.7467,
+      "step": 906
+    },
+    {
+      "epoch": 1.0033185840707965,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.491703539823009e-06,
+      "loss": 0.7909,
+      "step": 907
+    },
+    {
+      "epoch": 1.0044247787610618,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.4889380530973455e-06,
+      "loss": 0.8124,
+      "step": 908
+    },
+    {
+      "epoch": 1.0055309734513274,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 7.4861725663716825e-06,
+      "loss": 0.7509,
+      "step": 909
+    },
+    {
+      "epoch": 1.0066371681415929,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.483407079646018e-06,
+      "loss": 0.757,
+      "step": 910
+    },
+    {
+      "epoch": 1.0077433628318584,
+      "grad_norm": 0.10546875,
+      "learning_rate": 7.480641592920355e-06,
+      "loss": 0.8034,
+      "step": 911
+    },
+    {
+      "epoch": 1.008849557522124,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.477876106194692e-06,
+      "loss": 0.7688,
+      "step": 912
+    },
+    {
+      "epoch": 1.0099557522123894,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 7.475110619469027e-06,
+      "loss": 0.8436,
+      "step": 913
+    },
+    {
+      "epoch": 1.011061946902655,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.472345132743364e-06,
+      "loss": 0.7479,
+      "step": 914
+    },
+    {
+      "epoch": 1.0121681415929205,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.469579646017699e-06,
+      "loss": 0.7936,
+      "step": 915
+    },
+    {
+      "epoch": 1.0132743362831858,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.466814159292036e-06,
+      "loss": 0.7491,
+      "step": 916
+    },
+    {
+      "epoch": 1.0143805309734513,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.4640486725663715e-06,
+      "loss": 0.7553,
+      "step": 917
+    },
+    {
+      "epoch": 1.0154867256637168,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 7.4612831858407085e-06,
+      "loss": 0.766,
+      "step": 918
+    },
+    {
+      "epoch": 1.0165929203539823,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.458517699115045e-06,
+      "loss": 0.7985,
+      "step": 919
+    },
+    {
+      "epoch": 1.0176991150442478,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.455752212389381e-06,
+      "loss": 0.7635,
+      "step": 920
+    },
+    {
+      "epoch": 1.0188053097345133,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 7.452986725663718e-06,
+      "loss": 0.7886,
+      "step": 921
+    },
+    {
+      "epoch": 1.0199115044247788,
+      "grad_norm": 0.091796875,
+      "learning_rate": 7.450221238938054e-06,
+      "loss": 0.7448,
+      "step": 922
+    },
+    {
+      "epoch": 1.0210176991150441,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.44745575221239e-06,
+      "loss": 0.7628,
+      "step": 923
+    },
+    {
+      "epoch": 1.0221238938053097,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.444690265486726e-06,
+      "loss": 0.7261,
+      "step": 924
+    },
+    {
+      "epoch": 1.0232300884955752,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.441924778761063e-06,
+      "loss": 0.7558,
+      "step": 925
+    },
+    {
+      "epoch": 1.0243362831858407,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.439159292035398e-06,
+      "loss": 0.8527,
+      "step": 926
+    },
+    {
+      "epoch": 1.0254424778761062,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.436393805309735e-06,
+      "loss": 0.8012,
+      "step": 927
+    },
+    {
+      "epoch": 1.0265486725663717,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 7.4336283185840714e-06,
+      "loss": 0.7719,
+      "step": 928
+    },
+    {
+      "epoch": 1.0276548672566372,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 7.4308628318584076e-06,
+      "loss": 0.8055,
+      "step": 929
+    },
+    {
+      "epoch": 1.0287610619469028,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 7.4280973451327445e-06,
+      "loss": 0.7986,
+      "step": 930
+    },
+    {
+      "epoch": 1.029867256637168,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 7.42533185840708e-06,
+      "loss": 0.753,
+      "step": 931
+    },
+    {
+      "epoch": 1.0309734513274336,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.422566371681417e-06,
+      "loss": 0.7751,
+      "step": 932
+    },
+    {
+      "epoch": 1.032079646017699,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.419800884955752e-06,
+      "loss": 0.7807,
+      "step": 933
+    },
+    {
+      "epoch": 1.0331858407079646,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 7.417035398230089e-06,
+      "loss": 0.7548,
+      "step": 934
+    },
+    {
+      "epoch": 1.0342920353982301,
+      "grad_norm": 0.1171875,
+      "learning_rate": 7.414269911504426e-06,
+      "loss": 0.8274,
+      "step": 935
+    },
+    {
+      "epoch": 1.0353982300884956,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.411504424778761e-06,
+      "loss": 0.8909,
+      "step": 936
+    },
+    {
+      "epoch": 1.0365044247787611,
+      "grad_norm": 0.109375,
+      "learning_rate": 7.408738938053098e-06,
+      "loss": 0.7474,
+      "step": 937
+    },
+    {
+      "epoch": 1.0376106194690267,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.4059734513274336e-06,
+      "loss": 0.813,
+      "step": 938
+    },
+    {
+      "epoch": 1.038716814159292,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.4032079646017705e-06,
+      "loss": 0.7967,
+      "step": 939
+    },
+    {
+      "epoch": 1.0398230088495575,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 7.400442477876107e-06,
+      "loss": 0.8367,
+      "step": 940
+    },
+    {
+      "epoch": 1.040929203539823,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 7.397676991150443e-06,
+      "loss": 0.7872,
+      "step": 941
+    },
+    {
+      "epoch": 1.0420353982300885,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 7.39491150442478e-06,
+      "loss": 0.7758,
+      "step": 942
+    },
+    {
+      "epoch": 1.043141592920354,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 7.392146017699116e-06,
+      "loss": 0.7607,
+      "step": 943
+    },
+    {
+      "epoch": 1.0442477876106195,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 7.389380530973452e-06,
+      "loss": 0.7459,
+      "step": 944
+    },
+    {
+      "epoch": 1.045353982300885,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.386615044247788e-06,
+      "loss": 0.7642,
+      "step": 945
+    },
+    {
+      "epoch": 1.0464601769911503,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.383849557522125e-06,
+      "loss": 0.8003,
+      "step": 946
+    },
+    {
+      "epoch": 1.0475663716814159,
+      "grad_norm": 0.119140625,
+      "learning_rate": 7.38108407079646e-06,
+      "loss": 0.8492,
+      "step": 947
+    },
+    {
+      "epoch": 1.0486725663716814,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 7.378318584070797e-06,
+      "loss": 0.7293,
+      "step": 948
+    },
+    {
+      "epoch": 1.049778761061947,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 7.375553097345133e-06,
+      "loss": 0.7819,
+      "step": 949
+    },
+    {
+      "epoch": 1.0508849557522124,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.37278761061947e-06,
+      "loss": 0.773,
+      "step": 950
+    },
+    {
+      "epoch": 1.051991150442478,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 7.370022123893807e-06,
+      "loss": 0.798,
+      "step": 951
+    },
+    {
+      "epoch": 1.0530973451327434,
+      "grad_norm": 0.095703125,
+      "learning_rate": 7.367256637168142e-06,
+      "loss": 0.7665,
+      "step": 952
+    },
+    {
+      "epoch": 1.0542035398230087,
+      "grad_norm": 0.107421875,
+      "learning_rate": 7.364491150442479e-06,
+      "loss": 0.8183,
+      "step": 953
+    },
+    {
+      "epoch": 1.0553097345132743,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 7.361725663716814e-06,
+      "loss": 0.8717,
+      "step": 954
+    },
+    {
+      "epoch": 1.0564159292035398,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 7.358960176991151e-06,
+      "loss": 0.7684,
+      "step": 955
+    },
+    {
+      "epoch": 1.0575221238938053,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 7.356194690265487e-06,
+      "loss": 0.7748,
+      "step": 956
+    },
+    {
+      "epoch": 1.0586283185840708,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.353429203539823e-06,
+      "loss": 0.7566,
+      "step": 957
+    },
+    {
+      "epoch": 1.0597345132743363,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 7.35066371681416e-06,
+      "loss": 0.7521,
+      "step": 958
+    },
+    {
+      "epoch": 1.0608407079646018,
+      "grad_norm": 0.115234375,
+      "learning_rate": 7.347898230088496e-06,
+      "loss": 0.792,
+      "step": 959
+    },
+    {
+      "epoch": 1.0619469026548674,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.3451327433628326e-06,
+      "loss": 0.833,
+      "step": 960
+    },
+    {
+      "epoch": 1.0630530973451326,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 7.342367256637169e-06,
+      "loss": 0.7339,
+      "step": 961
+    },
+    {
+      "epoch": 1.0641592920353982,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 7.339601769911505e-06,
+      "loss": 0.7248,
+      "step": 962
+    },
+    {
+      "epoch": 1.0652654867256637,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 7.336836283185841e-06,
+      "loss": 0.8279,
+      "step": 963
+    },
+    {
+      "epoch": 1.0663716814159292,
+      "grad_norm": 0.103515625,
+      "learning_rate": 7.334070796460178e-06,
+      "loss": 0.7756,
+      "step": 964
+    },
+    {
+      "epoch": 1.0674778761061947,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 7.331305309734514e-06,
+      "loss": 0.7877,
+      "step": 965
+    },
+    {
+      "epoch": 1.0685840707964602,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.32853982300885e-06,
+      "loss": 0.7643,
+      "step": 966
+    },
+    {
+      "epoch": 1.0696902654867257,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.325774336283187e-06,
+      "loss": 0.8051,
+      "step": 967
+    },
+    {
+      "epoch": 1.0707964601769913,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 7.323008849557522e-06,
+      "loss": 0.7831,
+      "step": 968
+    },
+    {
+      "epoch": 1.0719026548672566,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.320243362831859e-06,
+      "loss": 0.7633,
+      "step": 969
+    },
+    {
+      "epoch": 1.073008849557522,
+      "grad_norm": 0.09375,
+      "learning_rate": 7.317477876106195e-06,
+      "loss": 0.7842,
+      "step": 970
+    },
+    {
+      "epoch": 1.0741150442477876,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.314712389380532e-06,
+      "loss": 0.7961,
+      "step": 971
+    },
+    {
+      "epoch": 1.075221238938053,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.311946902654869e-06,
+      "loss": 0.7743,
+      "step": 972
+    },
+    {
+      "epoch": 1.0763274336283186,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.309181415929204e-06,
+      "loss": 0.7785,
+      "step": 973
+    },
+    {
+      "epoch": 1.0774336283185841,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.306415929203541e-06,
+      "loss": 0.7534,
+      "step": 974
+    },
+    {
+      "epoch": 1.0785398230088497,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.303650442477876e-06,
+      "loss": 0.756,
+      "step": 975
+    },
+    {
+      "epoch": 1.079646017699115,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 7.300884955752213e-06,
+      "loss": 0.7888,
+      "step": 976
+    },
+    {
+      "epoch": 1.0807522123893805,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 7.298119469026549e-06,
+      "loss": 0.738,
+      "step": 977
+    },
+    {
+      "epoch": 1.081858407079646,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 7.295353982300885e-06,
+      "loss": 0.8398,
+      "step": 978
+    },
+    {
+      "epoch": 1.0829646017699115,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 7.2925884955752215e-06,
+      "loss": 0.8137,
+      "step": 979
+    },
+    {
+      "epoch": 1.084070796460177,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 7.2898230088495585e-06,
+      "loss": 0.785,
+      "step": 980
+    },
+    {
+      "epoch": 1.0851769911504425,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.287057522123895e-06,
+      "loss": 0.755,
+      "step": 981
+    },
+    {
+      "epoch": 1.086283185840708,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 7.284292035398231e-06,
+      "loss": 0.8098,
+      "step": 982
+    },
+    {
+      "epoch": 1.0873893805309733,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.281526548672567e-06,
+      "loss": 0.796,
+      "step": 983
+    },
+    {
+      "epoch": 1.0884955752212389,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 7.278761061946903e-06,
+      "loss": 0.7861,
+      "step": 984
+    },
+    {
+      "epoch": 1.0896017699115044,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.27599557522124e-06,
+      "loss": 0.7847,
+      "step": 985
+    },
+    {
+      "epoch": 1.0907079646017699,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 7.273230088495575e-06,
+      "loss": 0.7623,
+      "step": 986
+    },
+    {
+      "epoch": 1.0918141592920354,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.270464601769912e-06,
+      "loss": 0.7581,
+      "step": 987
+    },
+    {
+      "epoch": 1.092920353982301,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 7.267699115044249e-06,
+      "loss": 0.7309,
+      "step": 988
+    },
+    {
+      "epoch": 1.0940265486725664,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.2649336283185845e-06,
+      "loss": 0.7691,
+      "step": 989
+    },
+    {
+      "epoch": 1.095132743362832,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.2621681415929214e-06,
+      "loss": 0.7982,
+      "step": 990
+    },
+    {
+      "epoch": 1.0962389380530972,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.259402654867257e-06,
+      "loss": 0.7879,
+      "step": 991
+    },
+    {
+      "epoch": 1.0973451327433628,
+      "grad_norm": 0.08935546875,
+      "learning_rate": 7.256637168141594e-06,
+      "loss": 0.7759,
+      "step": 992
+    },
+    {
+      "epoch": 1.0984513274336283,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.253871681415929e-06,
+      "loss": 0.7768,
+      "step": 993
+    },
+    {
+      "epoch": 1.0995575221238938,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.251106194690266e-06,
+      "loss": 0.7541,
+      "step": 994
+    },
+    {
+      "epoch": 1.1006637168141593,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 7.248340707964603e-06,
+      "loss": 0.7765,
+      "step": 995
+    },
+    {
+      "epoch": 1.1017699115044248,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.245575221238938e-06,
+      "loss": 0.7628,
+      "step": 996
+    },
+    {
+      "epoch": 1.1028761061946903,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 7.242809734513275e-06,
+      "loss": 0.7752,
+      "step": 997
+    },
+    {
+      "epoch": 1.1039823008849559,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.240044247787611e-06,
+      "loss": 0.7609,
+      "step": 998
+    },
+    {
+      "epoch": 1.1050884955752212,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 7.237278761061947e-06,
+      "loss": 0.8013,
+      "step": 999
+    },
+    {
+      "epoch": 1.1061946902654867,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.2345132743362835e-06,
+      "loss": 0.7972,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1073008849557522,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 7.2317477876106205e-06,
+      "loss": 0.8247,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1084070796460177,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.228982300884957e-06,
+      "loss": 0.7551,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1095132743362832,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 7.226216814159293e-06,
+      "loss": 0.8104,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1106194690265487,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 7.22345132743363e-06,
+      "loss": 0.7784,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1117256637168142,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.220685840707965e-06,
+      "loss": 0.8033,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1128318584070795,
+      "grad_norm": 0.109375,
+      "learning_rate": 7.217920353982302e-06,
+      "loss": 0.7409,
+      "step": 1006
+    },
+    {
+      "epoch": 1.113938053097345,
+      "grad_norm": 0.099609375,
+      "learning_rate": 7.215154867256637e-06,
+      "loss": 0.7638,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1150442477876106,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.212389380530974e-06,
+      "loss": 0.762,
+      "step": 1008
+    },
+    {
+      "epoch": 1.116150442477876,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 7.2096238938053095e-06,
+      "loss": 0.7857,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1172566371681416,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.2068584070796465e-06,
+      "loss": 0.7387,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1183628318584071,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.2040929203539835e-06,
+      "loss": 0.7699,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1194690265486726,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 7.201327433628319e-06,
+      "loss": 0.7591,
+      "step": 1012
+    },
+    {
+      "epoch": 1.120575221238938,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 7.198561946902656e-06,
+      "loss": 0.8649,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1216814159292035,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 7.195796460176991e-06,
+      "loss": 0.7655,
+      "step": 1014
+    },
+    {
+      "epoch": 1.122787610619469,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 7.193030973451328e-06,
+      "loss": 0.7609,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1238938053097345,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 7.190265486725664e-06,
+      "loss": 0.7901,
+      "step": 1016
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.1875e-06,
+      "loss": 0.8171,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1261061946902655,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 7.184734513274337e-06,
+      "loss": 0.742,
+      "step": 1018
+    },
+    {
+      "epoch": 1.127212389380531,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.181969026548673e-06,
+      "loss": 0.7832,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1283185840707965,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 7.1792035398230094e-06,
+      "loss": 0.7978,
+      "step": 1020
+    },
+    {
+      "epoch": 1.129424778761062,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 7.1764380530973456e-06,
+      "loss": 0.7898,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1305309734513274,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.1736725663716825e-06,
+      "loss": 0.7851,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1316371681415929,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 7.170907079646018e-06,
+      "loss": 0.8118,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1327433628318584,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.168141592920355e-06,
+      "loss": 0.7745,
+      "step": 1024
+    },
+    {
+      "epoch": 1.133849557522124,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.165376106194692e-06,
+      "loss": 0.7858,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1349557522123894,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 7.162610619469027e-06,
+      "loss": 0.7637,
+      "step": 1026
+    },
+    {
+      "epoch": 1.136061946902655,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.159845132743364e-06,
+      "loss": 0.7615,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1371681415929205,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 7.157079646017699e-06,
+      "loss": 0.803,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1382743362831858,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.154314159292036e-06,
+      "loss": 0.7915,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1393805309734513,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 7.1515486725663715e-06,
+      "loss": 0.7648,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1404867256637168,
+      "grad_norm": 0.08642578125,
+      "learning_rate": 7.1487831858407085e-06,
+      "loss": 0.7349,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1415929203539823,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.146017699115045e-06,
+      "loss": 0.7658,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1426991150442478,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 7.143252212389381e-06,
+      "loss": 0.8175,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1438053097345133,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 7.140486725663718e-06,
+      "loss": 0.7667,
+      "step": 1034
+    },
+    {
+      "epoch": 1.1449115044247788,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 7.137721238938054e-06,
+      "loss": 0.7798,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1460176991150441,
+      "grad_norm": 0.103515625,
+      "learning_rate": 7.13495575221239e-06,
+      "loss": 0.7808,
+      "step": 1036
+    },
+    {
+      "epoch": 1.1471238938053097,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 7.132190265486726e-06,
+      "loss": 0.7687,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1482300884955752,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.129424778761062e-06,
+      "loss": 0.7826,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1493362831858407,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.126659292035398e-06,
+      "loss": 0.7974,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1504424778761062,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.123893805309735e-06,
+      "loss": 0.7749,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1515486725663717,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.1211283185840715e-06,
+      "loss": 0.842,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1526548672566372,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 7.118362831858408e-06,
+      "loss": 0.7612,
+      "step": 1042
+    },
+    {
+      "epoch": 1.1537610619469025,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.1155973451327446e-06,
+      "loss": 0.7619,
+      "step": 1043
+    },
+    {
+      "epoch": 1.154867256637168,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 7.11283185840708e-06,
+      "loss": 0.8538,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1559734513274336,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.110066371681417e-06,
+      "loss": 0.8343,
+      "step": 1045
+    },
+    {
+      "epoch": 1.157079646017699,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.107300884955752e-06,
+      "loss": 0.8183,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1581858407079646,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.104535398230089e-06,
+      "loss": 0.78,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1592920353982301,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 7.101769911504426e-06,
+      "loss": 0.801,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1603982300884956,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 7.099004424778761e-06,
+      "loss": 0.8024,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1615044247787611,
+      "grad_norm": 0.12109375,
+      "learning_rate": 7.096238938053098e-06,
+      "loss": 0.8094,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1626106194690267,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 7.093473451327434e-06,
+      "loss": 0.8084,
+      "step": 1051
+    },
+    {
+      "epoch": 1.163716814159292,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 7.0907079646017706e-06,
+      "loss": 0.7784,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1648230088495575,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 7.087942477876107e-06,
+      "loss": 0.764,
+      "step": 1053
+    },
+    {
+      "epoch": 1.165929203539823,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 7.085176991150443e-06,
+      "loss": 0.7925,
+      "step": 1054
+    },
+    {
+      "epoch": 1.1670353982300885,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.08241150442478e-06,
+      "loss": 0.8845,
+      "step": 1055
+    },
+    {
+      "epoch": 1.168141592920354,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.079646017699116e-06,
+      "loss": 0.8728,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1692477876106195,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.076880530973452e-06,
+      "loss": 0.8038,
+      "step": 1057
+    },
+    {
+      "epoch": 1.170353982300885,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 7.074115044247788e-06,
+      "loss": 0.806,
+      "step": 1058
+    },
+    {
+      "epoch": 1.1714601769911503,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 7.071349557522124e-06,
+      "loss": 0.7484,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1725663716814159,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 7.06858407079646e-06,
+      "loss": 0.7393,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1736725663716814,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 7.065818584070797e-06,
+      "loss": 0.7423,
+      "step": 1061
+    },
+    {
+      "epoch": 1.174778761061947,
+      "grad_norm": 0.10546875,
+      "learning_rate": 7.063053097345133e-06,
+      "loss": 0.786,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1758849557522124,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 7.06028761061947e-06,
+      "loss": 0.7567,
+      "step": 1063
+    },
+    {
+      "epoch": 1.176991150442478,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.057522123893807e-06,
+      "loss": 0.7849,
+      "step": 1064
+    },
+    {
+      "epoch": 1.1780973451327434,
+      "grad_norm": 0.12109375,
+      "learning_rate": 7.054756637168142e-06,
+      "loss": 0.7922,
+      "step": 1065
+    },
+    {
+      "epoch": 1.1792035398230087,
+      "grad_norm": 0.103515625,
+      "learning_rate": 7.051991150442479e-06,
+      "loss": 0.7846,
+      "step": 1066
+    },
+    {
+      "epoch": 1.1803097345132743,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.049225663716814e-06,
+      "loss": 0.7832,
+      "step": 1067
+    },
+    {
+      "epoch": 1.1814159292035398,
+      "grad_norm": 0.115234375,
+      "learning_rate": 7.046460176991151e-06,
+      "loss": 0.765,
+      "step": 1068
+    },
+    {
+      "epoch": 1.1825221238938053,
+      "grad_norm": 0.09765625,
+      "learning_rate": 7.043694690265486e-06,
+      "loss": 0.7346,
+      "step": 1069
+    },
+    {
+      "epoch": 1.1836283185840708,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 7.040929203539823e-06,
+      "loss": 0.7891,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1847345132743363,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 7.03816371681416e-06,
+      "loss": 0.8092,
+      "step": 1071
+    },
+    {
+      "epoch": 1.1858407079646018,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 7.035398230088496e-06,
+      "loss": 0.763,
+      "step": 1072
+    },
+    {
+      "epoch": 1.1869469026548674,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.032632743362833e-06,
+      "loss": 0.7644,
+      "step": 1073
+    },
+    {
+      "epoch": 1.1880530973451326,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 7.029867256637169e-06,
+      "loss": 0.7774,
+      "step": 1074
+    },
+    {
+      "epoch": 1.1891592920353982,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.027101769911505e-06,
+      "loss": 0.7798,
+      "step": 1075
+    },
+    {
+      "epoch": 1.1902654867256637,
+      "grad_norm": 0.103515625,
+      "learning_rate": 7.024336283185841e-06,
+      "loss": 0.8048,
+      "step": 1076
+    },
+    {
+      "epoch": 1.1913716814159292,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.021570796460178e-06,
+      "loss": 0.7817,
+      "step": 1077
+    },
+    {
+      "epoch": 1.1924778761061947,
+      "grad_norm": 0.109375,
+      "learning_rate": 7.018805309734514e-06,
+      "loss": 0.7504,
+      "step": 1078
+    },
+    {
+      "epoch": 1.1935840707964602,
+      "grad_norm": 0.08984375,
+      "learning_rate": 7.01603982300885e-06,
+      "loss": 0.7687,
+      "step": 1079
+    },
+    {
+      "epoch": 1.1946902654867257,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.013274336283187e-06,
+      "loss": 0.8765,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1957964601769913,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 7.0105088495575224e-06,
+      "loss": 0.7717,
+      "step": 1081
+    },
+    {
+      "epoch": 1.1969026548672566,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.007743362831859e-06,
+      "loss": 0.7628,
+      "step": 1082
+    },
+    {
+      "epoch": 1.198008849557522,
+      "grad_norm": 0.107421875,
+      "learning_rate": 7.004977876106195e-06,
+      "loss": 0.8009,
+      "step": 1083
+    },
+    {
+      "epoch": 1.1991150442477876,
+      "grad_norm": 0.111328125,
+      "learning_rate": 7.002212389380532e-06,
+      "loss": 0.7531,
+      "step": 1084
+    },
+    {
+      "epoch": 1.200221238938053,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 6.999446902654869e-06,
+      "loss": 0.7927,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2013274336283186,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.996681415929204e-06,
+      "loss": 0.7699,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2024336283185841,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.993915929203541e-06,
+      "loss": 0.8373,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2035398230088497,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.991150442477876e-06,
+      "loss": 0.8341,
+      "step": 1088
+    },
+    {
+      "epoch": 1.204646017699115,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.988384955752213e-06,
+      "loss": 0.7701,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2057522123893805,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 6.985619469026549e-06,
+      "loss": 0.7496,
+      "step": 1090
+    },
+    {
+      "epoch": 1.206858407079646,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.982853982300885e-06,
+      "loss": 0.7633,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2079646017699115,
+      "grad_norm": 0.095703125,
+      "learning_rate": 6.9800884955752215e-06,
+      "loss": 0.7704,
+      "step": 1092
+    },
+    {
+      "epoch": 1.209070796460177,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.977323008849558e-06,
+      "loss": 0.7652,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2101769911504425,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.974557522123895e-06,
+      "loss": 0.7599,
+      "step": 1094
+    },
+    {
+      "epoch": 1.211283185840708,
+      "grad_norm": 0.125,
+      "learning_rate": 6.971792035398231e-06,
+      "loss": 0.7608,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2123893805309733,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.969026548672567e-06,
+      "loss": 0.8105,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2134955752212389,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 6.966261061946903e-06,
+      "loss": 0.78,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2146017699115044,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 6.96349557522124e-06,
+      "loss": 0.7612,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2157079646017699,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.960730088495575e-06,
+      "loss": 0.8058,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2168141592920354,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.957964601769912e-06,
+      "loss": 0.7551,
+      "step": 1100
+    },
+    {
+      "epoch": 1.217920353982301,
+      "grad_norm": 0.109375,
+      "learning_rate": 6.955199115044249e-06,
+      "loss": 0.757,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2190265486725664,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 6.9524336283185845e-06,
+      "loss": 0.788,
+      "step": 1102
+    },
+    {
+      "epoch": 1.220132743362832,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.9496681415929215e-06,
+      "loss": 0.7949,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2212389380530975,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.946902654867257e-06,
+      "loss": 0.8009,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2223451327433628,
+      "grad_norm": 0.123046875,
+      "learning_rate": 6.944137168141594e-06,
+      "loss": 0.8713,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2234513274336283,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.941371681415929e-06,
+      "loss": 0.7359,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2245575221238938,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.938606194690266e-06,
+      "loss": 0.8229,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2256637168141593,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.935840707964603e-06,
+      "loss": 0.7369,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2267699115044248,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.933075221238938e-06,
+      "loss": 0.793,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2278761061946903,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.930309734513275e-06,
+      "loss": 0.7647,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2289823008849559,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.927544247787611e-06,
+      "loss": 0.7669,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2300884955752212,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 6.9247787610619474e-06,
+      "loss": 0.7999,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2311946902654867,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.9220132743362836e-06,
+      "loss": 0.7727,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2323008849557522,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 6.91924778761062e-06,
+      "loss": 0.7928,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2334070796460177,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.916482300884957e-06,
+      "loss": 0.7367,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2345132743362832,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 6.913716814159293e-06,
+      "loss": 0.8032,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2356194690265487,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 6.910951327433629e-06,
+      "loss": 0.7747,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2367256637168142,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 6.908185840707965e-06,
+      "loss": 0.7694,
+      "step": 1118
+    },
+    {
+      "epoch": 1.2378318584070795,
+      "grad_norm": 0.25390625,
+      "learning_rate": 6.905420353982302e-06,
+      "loss": 0.7554,
+      "step": 1119
+    },
+    {
+      "epoch": 1.238938053097345,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.902654867256637e-06,
+      "loss": 0.7731,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2400442477876106,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 6.899889380530974e-06,
+      "loss": 0.7791,
+      "step": 1121
+    },
+    {
+      "epoch": 1.241150442477876,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.8971238938053095e-06,
+      "loss": 0.7727,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2422566371681416,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 6.8943584070796465e-06,
+      "loss": 0.7843,
+      "step": 1123
+    },
+    {
+      "epoch": 1.2433628318584071,
+      "grad_norm": 0.0927734375,
+      "learning_rate": 6.8915929203539835e-06,
+      "loss": 0.7487,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2444690265486726,
+      "grad_norm": 0.115234375,
+      "learning_rate": 6.888827433628319e-06,
+      "loss": 0.8183,
+      "step": 1125
+    },
+    {
+      "epoch": 1.245575221238938,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.886061946902656e-06,
+      "loss": 0.7943,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2466814159292035,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 6.883296460176991e-06,
+      "loss": 0.7965,
+      "step": 1127
+    },
+    {
+      "epoch": 1.247787610619469,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.880530973451328e-06,
+      "loss": 0.7322,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2488938053097345,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.877765486725664e-06,
+      "loss": 0.8306,
+      "step": 1129
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.875e-06,
+      "loss": 0.7835,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2511061946902655,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.872234513274337e-06,
+      "loss": 0.7938,
+      "step": 1131
+    },
+    {
+      "epoch": 1.252212389380531,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.869469026548673e-06,
+      "loss": 0.7784,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2533185840707963,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.8667035398230095e-06,
+      "loss": 0.7628,
+      "step": 1133
+    },
+    {
+      "epoch": 1.254424778761062,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.863938053097346e-06,
+      "loss": 0.7955,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2555309734513274,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.8611725663716826e-06,
+      "loss": 0.7898,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2566371681415929,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.858407079646018e-06,
+      "loss": 0.8178,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2577433628318584,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 6.855641592920355e-06,
+      "loss": 0.785,
+      "step": 1137
+    },
+    {
+      "epoch": 1.258849557522124,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 6.852876106194691e-06,
+      "loss": 0.7692,
+      "step": 1138
+    },
+    {
+      "epoch": 1.2599557522123894,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.850110619469027e-06,
+      "loss": 0.7774,
+      "step": 1139
+    },
+    {
+      "epoch": 1.261061946902655,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.847345132743364e-06,
+      "loss": 0.7896,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2621681415929205,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 6.844579646017699e-06,
+      "loss": 0.8065,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2632743362831858,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.841814159292036e-06,
+      "loss": 0.7661,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2643805309734513,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.839048672566372e-06,
+      "loss": 0.7448,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2654867256637168,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.8362831858407086e-06,
+      "loss": 0.7713,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2665929203539823,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 6.833517699115045e-06,
+      "loss": 0.7594,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2676991150442478,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.830752212389381e-06,
+      "loss": 0.7739,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2688053097345133,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 6.827986725663718e-06,
+      "loss": 0.7515,
+      "step": 1147
+    },
+    {
+      "epoch": 1.2699115044247788,
+      "grad_norm": 0.123046875,
+      "learning_rate": 6.825221238938053e-06,
+      "loss": 0.7894,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2710176991150441,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.82245575221239e-06,
+      "loss": 0.7775,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2721238938053097,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.819690265486726e-06,
+      "loss": 0.7895,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2732300884955752,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 6.816924778761062e-06,
+      "loss": 0.836,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2743362831858407,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.814159292035398e-06,
+      "loss": 0.8022,
+      "step": 1152
+    },
+    {
+      "epoch": 1.2754424778761062,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.811393805309735e-06,
+      "loss": 0.7869,
+      "step": 1153
+    },
+    {
+      "epoch": 1.2765486725663717,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.8086283185840715e-06,
+      "loss": 0.7703,
+      "step": 1154
+    },
+    {
+      "epoch": 1.2776548672566372,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 6.805862831858408e-06,
+      "loss": 0.7715,
+      "step": 1155
+    },
+    {
+      "epoch": 1.2787610619469025,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.803097345132745e-06,
+      "loss": 0.7691,
+      "step": 1156
+    },
+    {
+      "epoch": 1.2798672566371683,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.80033185840708e-06,
+      "loss": 0.7627,
+      "step": 1157
+    },
+    {
+      "epoch": 1.2809734513274336,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.797566371681417e-06,
+      "loss": 0.7931,
+      "step": 1158
+    },
+    {
+      "epoch": 1.282079646017699,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.794800884955752e-06,
+      "loss": 0.7756,
+      "step": 1159
+    },
+    {
+      "epoch": 1.2831858407079646,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.792035398230089e-06,
+      "loss": 0.7569,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2842920353982301,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.789269911504426e-06,
+      "loss": 0.7585,
+      "step": 1161
+    },
+    {
+      "epoch": 1.2853982300884956,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.786504424778761e-06,
+      "loss": 0.8527,
+      "step": 1162
+    },
+    {
+      "epoch": 1.286504424778761,
+      "grad_norm": 0.109375,
+      "learning_rate": 6.783738938053098e-06,
+      "loss": 0.7866,
+      "step": 1163
+    },
+    {
+      "epoch": 1.2876106194690267,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.780973451327434e-06,
+      "loss": 0.7519,
+      "step": 1164
+    },
+    {
+      "epoch": 1.288716814159292,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.778207964601771e-06,
+      "loss": 0.7773,
+      "step": 1165
+    },
+    {
+      "epoch": 1.2898230088495575,
+      "grad_norm": 0.12890625,
+      "learning_rate": 6.775442477876107e-06,
+      "loss": 0.8457,
+      "step": 1166
+    },
+    {
+      "epoch": 1.290929203539823,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 6.772676991150443e-06,
+      "loss": 0.8189,
+      "step": 1167
+    },
+    {
+      "epoch": 1.2920353982300885,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.76991150442478e-06,
+      "loss": 0.7468,
+      "step": 1168
+    },
+    {
+      "epoch": 1.293141592920354,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.767146017699115e-06,
+      "loss": 0.7546,
+      "step": 1169
+    },
+    {
+      "epoch": 1.2942477876106195,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 6.764380530973452e-06,
+      "loss": 0.7815,
+      "step": 1170
+    },
+    {
+      "epoch": 1.295353982300885,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.761615044247788e-06,
+      "loss": 0.8779,
+      "step": 1171
+    },
+    {
+      "epoch": 1.2964601769911503,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.758849557522124e-06,
+      "loss": 0.72,
+      "step": 1172
+    },
+    {
+      "epoch": 1.2975663716814159,
+      "grad_norm": 0.09375,
+      "learning_rate": 6.7560840707964604e-06,
+      "loss": 0.7391,
+      "step": 1173
+    },
+    {
+      "epoch": 1.2986725663716814,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.753318584070797e-06,
+      "loss": 0.7184,
+      "step": 1174
+    },
+    {
+      "epoch": 1.299778761061947,
+      "grad_norm": 0.1171875,
+      "learning_rate": 6.750553097345133e-06,
+      "loss": 0.7913,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3008849557522124,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.74778761061947e-06,
+      "loss": 0.7679,
+      "step": 1176
+    },
+    {
+      "epoch": 1.301991150442478,
+      "grad_norm": 0.107421875,
+      "learning_rate": 6.745022123893807e-06,
+      "loss": 0.792,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3030973451327434,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.742256637168142e-06,
+      "loss": 0.7582,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3042035398230087,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.739491150442479e-06,
+      "loss": 0.7699,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3053097345132743,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 6.736725663716814e-06,
+      "loss": 0.7616,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3064159292035398,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 6.733960176991151e-06,
+      "loss": 0.8144,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3075221238938053,
+      "grad_norm": 0.111328125,
+      "learning_rate": 6.731194690265486e-06,
+      "loss": 0.7795,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3086283185840708,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.728429203539823e-06,
+      "loss": 0.812,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3097345132743363,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 6.72566371681416e-06,
+      "loss": 0.7692,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3108407079646018,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.722898230088496e-06,
+      "loss": 0.7729,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3119469026548671,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 6.720132743362833e-06,
+      "loss": 0.7648,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3130530973451329,
+      "grad_norm": 0.111328125,
+      "learning_rate": 6.717367256637169e-06,
+      "loss": 0.7385,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3141592920353982,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.714601769911505e-06,
+      "loss": 0.7913,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3152654867256637,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.711836283185841e-06,
+      "loss": 0.7711,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3163716814159292,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.709070796460178e-06,
+      "loss": 0.7872,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3174778761061947,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.706305309734514e-06,
+      "loss": 0.8227,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3185840707964602,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.70353982300885e-06,
+      "loss": 0.7868,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3196902654867257,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.700774336283186e-06,
+      "loss": 0.7612,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3207964601769913,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.6980088495575225e-06,
+      "loss": 0.7857,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3219026548672566,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.6952433628318594e-06,
+      "loss": 0.7508,
+      "step": 1195
+    },
+    {
+      "epoch": 1.323008849557522,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.692477876106195e-06,
+      "loss": 0.7526,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3241150442477876,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.689712389380532e-06,
+      "loss": 0.8166,
+      "step": 1197
+    },
+    {
+      "epoch": 1.325221238938053,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 6.686946902654869e-06,
+      "loss": 0.7383,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3263274336283186,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 6.684181415929204e-06,
+      "loss": 0.741,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3274336283185841,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.681415929203541e-06,
+      "loss": 0.7676,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3285398230088497,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.678650442477876e-06,
+      "loss": 0.7933,
+      "step": 1201
+    },
+    {
+      "epoch": 1.329646017699115,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.675884955752213e-06,
+      "loss": 0.7664,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3307522123893805,
+      "grad_norm": 0.09375,
+      "learning_rate": 6.6731194690265485e-06,
+      "loss": 0.7462,
+      "step": 1203
+    },
+    {
+      "epoch": 1.331858407079646,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.6703539823008854e-06,
+      "loss": 0.7726,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3329646017699115,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.6675884955752216e-06,
+      "loss": 0.7598,
+      "step": 1205
+    },
+    {
+      "epoch": 1.334070796460177,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.664823008849558e-06,
+      "loss": 0.7645,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3351769911504425,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.662057522123895e-06,
+      "loss": 0.7764,
+      "step": 1207
+    },
+    {
+      "epoch": 1.336283185840708,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.659292035398231e-06,
+      "loss": 0.7874,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3373893805309733,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.656526548672567e-06,
+      "loss": 0.7558,
+      "step": 1209
+    },
+    {
+      "epoch": 1.338495575221239,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 6.653761061946903e-06,
+      "loss": 0.7827,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3396017699115044,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.65099557522124e-06,
+      "loss": 0.7479,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3407079646017699,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.648230088495575e-06,
+      "loss": 0.832,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3418141592920354,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.645464601769912e-06,
+      "loss": 0.7491,
+      "step": 1213
+    },
+    {
+      "epoch": 1.342920353982301,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.642699115044248e-06,
+      "loss": 0.7928,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3440265486725664,
+      "grad_norm": 0.109375,
+      "learning_rate": 6.6399336283185845e-06,
+      "loss": 0.7811,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3451327433628317,
+      "grad_norm": 0.11328125,
+      "learning_rate": 6.6371681415929215e-06,
+      "loss": 0.7844,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3462389380530975,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 6.634402654867257e-06,
+      "loss": 0.7757,
+      "step": 1217
+    },
+    {
+      "epoch": 1.3473451327433628,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.631637168141594e-06,
+      "loss": 0.7721,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3484513274336283,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.628871681415929e-06,
+      "loss": 0.8306,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3495575221238938,
+      "grad_norm": 0.107421875,
+      "learning_rate": 6.626106194690266e-06,
+      "loss": 0.7593,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3506637168141593,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.623340707964603e-06,
+      "loss": 0.7918,
+      "step": 1221
+    },
+    {
+      "epoch": 1.3517699115044248,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.620575221238938e-06,
+      "loss": 0.8604,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3528761061946903,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.617809734513275e-06,
+      "loss": 0.7875,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3539823008849559,
+      "grad_norm": 0.095703125,
+      "learning_rate": 6.6150442477876105e-06,
+      "loss": 0.7454,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3550884955752212,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.6122787610619475e-06,
+      "loss": 0.8187,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3561946902654867,
+      "grad_norm": 0.12109375,
+      "learning_rate": 6.609513274336284e-06,
+      "loss": 0.8003,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3573008849557522,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 6.60674778761062e-06,
+      "loss": 0.8318,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3584070796460177,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 6.603982300884957e-06,
+      "loss": 0.7791,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3595132743362832,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.601216814159293e-06,
+      "loss": 0.7794,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3606194690265487,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 6.598451327433629e-06,
+      "loss": 0.7748,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3617256637168142,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.595685840707965e-06,
+      "loss": 0.7927,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3628318584070795,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.592920353982302e-06,
+      "loss": 0.7595,
+      "step": 1232
+    },
+    {
+      "epoch": 1.363938053097345,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.590154867256637e-06,
+      "loss": 0.7952,
+      "step": 1233
+    },
+    {
+      "epoch": 1.3650442477876106,
+      "grad_norm": 0.107421875,
+      "learning_rate": 6.587389380530974e-06,
+      "loss": 0.7888,
+      "step": 1234
+    },
+    {
+      "epoch": 1.366150442477876,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.5846238938053096e-06,
+      "loss": 0.7755,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3672566371681416,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.5818584070796465e-06,
+      "loss": 0.7885,
+      "step": 1236
+    },
+    {
+      "epoch": 1.3683628318584071,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.5790929203539835e-06,
+      "loss": 0.7607,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3694690265486726,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.576327433628319e-06,
+      "loss": 0.771,
+      "step": 1238
+    },
+    {
+      "epoch": 1.370575221238938,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.573561946902656e-06,
+      "loss": 0.7422,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3716814159292037,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 6.570796460176991e-06,
+      "loss": 0.7917,
+      "step": 1240
+    },
+    {
+      "epoch": 1.372787610619469,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.568030973451328e-06,
+      "loss": 0.7519,
+      "step": 1241
+    },
+    {
+      "epoch": 1.3738938053097345,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.565265486725664e-06,
+      "loss": 0.7201,
+      "step": 1242
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.5625e-06,
+      "loss": 0.7777,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3761061946902655,
+      "grad_norm": 0.123046875,
+      "learning_rate": 6.559734513274337e-06,
+      "loss": 0.7796,
+      "step": 1244
+    },
+    {
+      "epoch": 1.377212389380531,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.556969026548673e-06,
+      "loss": 0.7346,
+      "step": 1245
+    },
+    {
+      "epoch": 1.3783185840707963,
+      "grad_norm": 0.123046875,
+      "learning_rate": 6.5542035398230095e-06,
+      "loss": 0.7469,
+      "step": 1246
+    },
+    {
+      "epoch": 1.379424778761062,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 6.551438053097346e-06,
+      "loss": 0.7363,
+      "step": 1247
+    },
+    {
+      "epoch": 1.3805309734513274,
+      "grad_norm": 0.111328125,
+      "learning_rate": 6.548672566371682e-06,
+      "loss": 0.7895,
+      "step": 1248
+    },
+    {
+      "epoch": 1.3816371681415929,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 6.545907079646018e-06,
+      "loss": 0.742,
+      "step": 1249
+    },
+    {
+      "epoch": 1.3827433628318584,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.543141592920355e-06,
+      "loss": 0.7676,
+      "step": 1250
+    },
+    {
+      "epoch": 1.383849557522124,
+      "grad_norm": 0.09765625,
+      "learning_rate": 6.540376106194691e-06,
+      "loss": 0.7516,
+      "step": 1251
+    },
+    {
+      "epoch": 1.3849557522123894,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.537610619469027e-06,
+      "loss": 0.8043,
+      "step": 1252
+    },
+    {
+      "epoch": 1.386061946902655,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.534845132743364e-06,
+      "loss": 0.8382,
+      "step": 1253
+    },
+    {
+      "epoch": 1.3871681415929205,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.532079646017699e-06,
+      "loss": 0.7522,
+      "step": 1254
+    },
+    {
+      "epoch": 1.3882743362831858,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 6.529314159292036e-06,
+      "loss": 0.8026,
+      "step": 1255
+    },
+    {
+      "epoch": 1.3893805309734513,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.526548672566372e-06,
+      "loss": 0.7833,
+      "step": 1256
+    },
+    {
+      "epoch": 1.3904867256637168,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.523783185840709e-06,
+      "loss": 0.799,
+      "step": 1257
+    },
+    {
+      "epoch": 1.3915929203539823,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.521017699115044e-06,
+      "loss": 0.7723,
+      "step": 1258
+    },
+    {
+      "epoch": 1.3926991150442478,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.518252212389381e-06,
+      "loss": 0.7997,
+      "step": 1259
+    },
+    {
+      "epoch": 1.3938053097345133,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 6.515486725663718e-06,
+      "loss": 0.7701,
+      "step": 1260
+    },
+    {
+      "epoch": 1.3949115044247788,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.512721238938053e-06,
+      "loss": 0.7566,
+      "step": 1261
+    },
+    {
+      "epoch": 1.3960176991150441,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.50995575221239e-06,
+      "loss": 0.7776,
+      "step": 1262
+    },
+    {
+      "epoch": 1.3971238938053097,
+      "grad_norm": 0.09375,
+      "learning_rate": 6.507190265486726e-06,
+      "loss": 0.7708,
+      "step": 1263
+    },
+    {
+      "epoch": 1.3982300884955752,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.504424778761062e-06,
+      "loss": 0.7764,
+      "step": 1264
+    },
+    {
+      "epoch": 1.3993362831858407,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.5016592920353984e-06,
+      "loss": 0.8041,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4004424778761062,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.498893805309735e-06,
+      "loss": 0.7704,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4015486725663717,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 6.4961283185840715e-06,
+      "loss": 0.8234,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4026548672566372,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.493362831858408e-06,
+      "loss": 0.7865,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4037610619469025,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.490597345132744e-06,
+      "loss": 0.7973,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4048672566371683,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.48783185840708e-06,
+      "loss": 0.734,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4059734513274336,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.485066371681417e-06,
+      "loss": 0.7328,
+      "step": 1271
+    },
+    {
+      "epoch": 1.407079646017699,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.482300884955752e-06,
+      "loss": 0.7525,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4081858407079646,
+      "grad_norm": 0.115234375,
+      "learning_rate": 6.479535398230089e-06,
+      "loss": 0.8237,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4092920353982301,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 6.476769911504426e-06,
+      "loss": 0.7651,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4103982300884956,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.474004424778761e-06,
+      "loss": 0.7534,
+      "step": 1275
+    },
+    {
+      "epoch": 1.411504424778761,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.471238938053098e-06,
+      "loss": 0.7881,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4126106194690267,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.468473451327434e-06,
+      "loss": 0.7812,
+      "step": 1277
+    },
+    {
+      "epoch": 1.413716814159292,
+      "grad_norm": 0.107421875,
+      "learning_rate": 6.465707964601771e-06,
+      "loss": 0.742,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4148230088495575,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.462942477876106e-06,
+      "loss": 0.7834,
+      "step": 1279
+    },
+    {
+      "epoch": 1.415929203539823,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.460176991150443e-06,
+      "loss": 0.7579,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4170353982300885,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 6.45741150442478e-06,
+      "loss": 0.7651,
+      "step": 1281
+    },
+    {
+      "epoch": 1.418141592920354,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.454646017699115e-06,
+      "loss": 0.7925,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4192477876106195,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 6.451880530973452e-06,
+      "loss": 0.862,
+      "step": 1283
+    },
+    {
+      "epoch": 1.420353982300885,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 6.449115044247788e-06,
+      "loss": 0.8362,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4214601769911503,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.446349557522124e-06,
+      "loss": 0.7326,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4225663716814159,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.4435840707964605e-06,
+      "loss": 0.7883,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4236725663716814,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 6.4408185840707974e-06,
+      "loss": 0.8145,
+      "step": 1287
+    },
+    {
+      "epoch": 1.424778761061947,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 6.438053097345133e-06,
+      "loss": 0.8031,
+      "step": 1288
+    },
+    {
+      "epoch": 1.4258849557522124,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.43528761061947e-06,
+      "loss": 0.79,
+      "step": 1289
+    },
+    {
+      "epoch": 1.426991150442478,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 6.432522123893807e-06,
+      "loss": 0.7141,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4280973451327434,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.429756637168142e-06,
+      "loss": 0.8021,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4292035398230087,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.426991150442479e-06,
+      "loss": 0.7981,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4303097345132743,
+      "grad_norm": 0.09765625,
+      "learning_rate": 6.424225663716814e-06,
+      "loss": 0.7301,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4314159292035398,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.421460176991151e-06,
+      "loss": 0.7971,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4325221238938053,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.4186946902654864e-06,
+      "loss": 0.7934,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4336283185840708,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.415929203539823e-06,
+      "loss": 0.7381,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4347345132743363,
+      "grad_norm": 0.111328125,
+      "learning_rate": 6.41316371681416e-06,
+      "loss": 0.7928,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4358407079646018,
+      "grad_norm": 0.107421875,
+      "learning_rate": 6.410398230088496e-06,
+      "loss": 0.7714,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4369469026548671,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 6.407632743362833e-06,
+      "loss": 0.7851,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4380530973451329,
+      "grad_norm": 0.09765625,
+      "learning_rate": 6.404867256637169e-06,
+      "loss": 0.7426,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4391592920353982,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.402101769911505e-06,
+      "loss": 0.7766,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4402654867256637,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 6.399336283185841e-06,
+      "loss": 0.7509,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4413716814159292,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.396570796460177e-06,
+      "loss": 0.7581,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4424778761061947,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.393805309734514e-06,
+      "loss": 0.7491,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4435840707964602,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.39103982300885e-06,
+      "loss": 0.7274,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4446902654867257,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 6.388274336283186e-06,
+      "loss": 0.7639,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4457964601769913,
+      "grad_norm": 0.11328125,
+      "learning_rate": 6.3855088495575225e-06,
+      "loss": 0.8163,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4469026548672566,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 6.3827433628318595e-06,
+      "loss": 0.7313,
+      "step": 1308
+    },
+    {
+      "epoch": 1.448008849557522,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.379977876106195e-06,
+      "loss": 0.771,
+      "step": 1309
+    },
+    {
+      "epoch": 1.4491150442477876,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 6.377212389380532e-06,
+      "loss": 0.7736,
+      "step": 1310
+    },
+    {
+      "epoch": 1.450221238938053,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.374446902654869e-06,
+      "loss": 0.7662,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4513274336283186,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.371681415929204e-06,
+      "loss": 0.784,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4524336283185841,
+      "grad_norm": 0.12109375,
+      "learning_rate": 6.368915929203541e-06,
+      "loss": 0.8322,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4535398230088497,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.366150442477876e-06,
+      "loss": 0.7751,
+      "step": 1314
+    },
+    {
+      "epoch": 1.454646017699115,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 6.363384955752213e-06,
+      "loss": 0.7712,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4557522123893805,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 6.3606194690265485e-06,
+      "loss": 0.7461,
+      "step": 1316
+    },
+    {
+      "epoch": 1.456858407079646,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.3578539823008855e-06,
+      "loss": 0.7939,
+      "step": 1317
+    },
+    {
+      "epoch": 1.4579646017699115,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 6.355088495575222e-06,
+      "loss": 0.7745,
+      "step": 1318
+    },
+    {
+      "epoch": 1.459070796460177,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 6.352323008849558e-06,
+      "loss": 0.8152,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4601769911504425,
+      "grad_norm": 0.125,
+      "learning_rate": 6.349557522123895e-06,
+      "loss": 0.8529,
+      "step": 1320
+    },
+    {
+      "epoch": 1.461283185840708,
+      "grad_norm": 0.1171875,
+      "learning_rate": 6.346792035398231e-06,
+      "loss": 0.8006,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4623893805309733,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.344026548672567e-06,
+      "loss": 0.789,
+      "step": 1322
+    },
+    {
+      "epoch": 1.463495575221239,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.341261061946903e-06,
+      "loss": 0.806,
+      "step": 1323
+    },
+    {
+      "epoch": 1.4646017699115044,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 6.338495575221239e-06,
+      "loss": 0.7686,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4657079646017699,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.335730088495575e-06,
+      "loss": 0.7296,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4668141592920354,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 6.332964601769912e-06,
+      "loss": 0.831,
+      "step": 1326
+    },
+    {
+      "epoch": 1.467920353982301,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.330199115044248e-06,
+      "loss": 0.7542,
+      "step": 1327
+    },
+    {
+      "epoch": 1.4690265486725664,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 6.3274336283185845e-06,
+      "loss": 0.7748,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4701327433628317,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 6.3246681415929215e-06,
+      "loss": 0.7787,
+      "step": 1329
+    },
+    {
+      "epoch": 1.4712389380530975,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.321902654867257e-06,
+      "loss": 0.7621,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4723451327433628,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.319137168141594e-06,
+      "loss": 0.7829,
+      "step": 1331
+    },
+    {
+      "epoch": 1.4734513274336283,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.316371681415929e-06,
+      "loss": 0.8228,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4745575221238938,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.313606194690266e-06,
+      "loss": 0.8048,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4756637168141593,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 6.310840707964603e-06,
+      "loss": 0.8119,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4767699115044248,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 6.308075221238938e-06,
+      "loss": 0.7564,
+      "step": 1335
+    },
+    {
+      "epoch": 1.4778761061946903,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.305309734513275e-06,
+      "loss": 0.796,
+      "step": 1336
+    },
+    {
+      "epoch": 1.4789823008849559,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.3025442477876105e-06,
+      "loss": 0.7367,
+      "step": 1337
+    },
+    {
+      "epoch": 1.4800884955752212,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.2997787610619475e-06,
+      "loss": 0.7695,
+      "step": 1338
+    },
+    {
+      "epoch": 1.4811946902654867,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.297013274336284e-06,
+      "loss": 0.761,
+      "step": 1339
+    },
+    {
+      "epoch": 1.4823008849557522,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.29424778761062e-06,
+      "loss": 0.7709,
+      "step": 1340
+    },
+    {
+      "epoch": 1.4834070796460177,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 6.291482300884957e-06,
+      "loss": 0.8421,
+      "step": 1341
+    },
+    {
+      "epoch": 1.4845132743362832,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 6.288716814159293e-06,
+      "loss": 0.7971,
+      "step": 1342
+    },
+    {
+      "epoch": 1.4856194690265487,
+      "grad_norm": 0.115234375,
+      "learning_rate": 6.285951327433629e-06,
+      "loss": 0.8255,
+      "step": 1343
+    },
+    {
+      "epoch": 1.4867256637168142,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.283185840707965e-06,
+      "loss": 0.7843,
+      "step": 1344
+    },
+    {
+      "epoch": 1.4878318584070795,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.280420353982302e-06,
+      "loss": 0.767,
+      "step": 1345
+    },
+    {
+      "epoch": 1.488938053097345,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.277654867256637e-06,
+      "loss": 0.77,
+      "step": 1346
+    },
+    {
+      "epoch": 1.4900442477876106,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 6.274889380530974e-06,
+      "loss": 0.7652,
+      "step": 1347
+    },
+    {
+      "epoch": 1.491150442477876,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 6.27212389380531e-06,
+      "loss": 0.7472,
+      "step": 1348
+    },
+    {
+      "epoch": 1.4922566371681416,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.2693584070796466e-06,
+      "loss": 0.8007,
+      "step": 1349
+    },
+    {
+      "epoch": 1.4933628318584071,
+      "grad_norm": 0.107421875,
+      "learning_rate": 6.2665929203539835e-06,
+      "loss": 0.7702,
+      "step": 1350
+    },
+    {
+      "epoch": 1.4944690265486726,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.263827433628319e-06,
+      "loss": 0.7775,
+      "step": 1351
+    },
+    {
+      "epoch": 1.495575221238938,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.261061946902656e-06,
+      "loss": 0.7493,
+      "step": 1352
+    },
+    {
+      "epoch": 1.4966814159292037,
+      "grad_norm": 0.115234375,
+      "learning_rate": 6.258296460176991e-06,
+      "loss": 0.8142,
+      "step": 1353
+    },
+    {
+      "epoch": 1.497787610619469,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.255530973451328e-06,
+      "loss": 0.7946,
+      "step": 1354
+    },
+    {
+      "epoch": 1.4988938053097345,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 6.252765486725663e-06,
+      "loss": 0.8088,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.115234375,
+      "learning_rate": 6.25e-06,
+      "loss": 0.7917,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5011061946902655,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.247234513274337e-06,
+      "loss": 0.7767,
+      "step": 1357
+    },
+    {
+      "epoch": 1.502212389380531,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.2444690265486726e-06,
+      "loss": 0.7521,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5033185840707963,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.2417035398230095e-06,
+      "loss": 0.7443,
+      "step": 1359
+    },
+    {
+      "epoch": 1.504424778761062,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 6.238938053097346e-06,
+      "loss": 0.8354,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5055309734513274,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.236172566371682e-06,
+      "loss": 0.769,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5066371681415929,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 6.233407079646018e-06,
+      "loss": 0.8175,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5077433628318584,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.230641592920355e-06,
+      "loss": 0.8051,
+      "step": 1363
+    },
+    {
+      "epoch": 1.508849557522124,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.227876106194691e-06,
+      "loss": 0.7656,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5099557522123894,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.225110619469027e-06,
+      "loss": 0.7638,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5110619469026547,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 6.222345132743364e-06,
+      "loss": 0.7911,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5121681415929205,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.219579646017699e-06,
+      "loss": 0.8184,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5132743362831858,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.216814159292036e-06,
+      "loss": 0.7467,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5143805309734515,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 6.214048672566372e-06,
+      "loss": 0.7946,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5154867256637168,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.211283185840709e-06,
+      "loss": 0.7438,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5165929203539823,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 6.208517699115044e-06,
+      "loss": 0.7792,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5176991150442478,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.205752212389381e-06,
+      "loss": 0.7757,
+      "step": 1372
+    },
+    {
+      "epoch": 1.518805309734513,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 6.202986725663718e-06,
+      "loss": 0.7736,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5199115044247788,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.200221238938053e-06,
+      "loss": 0.7904,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5210176991150441,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.19745575221239e-06,
+      "loss": 0.7462,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5221238938053099,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.194690265486726e-06,
+      "loss": 0.7541,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5232300884955752,
+      "grad_norm": 0.125,
+      "learning_rate": 6.191924778761062e-06,
+      "loss": 0.8348,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5243362831858407,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 6.1891592920353985e-06,
+      "loss": 0.7517,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5254424778761062,
+      "grad_norm": 0.1171875,
+      "learning_rate": 6.186393805309735e-06,
+      "loss": 0.7829,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5265486725663717,
+      "grad_norm": 0.109375,
+      "learning_rate": 6.1836283185840716e-06,
+      "loss": 0.7696,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5276548672566372,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.180862831858408e-06,
+      "loss": 0.7842,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5287610619469025,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 6.178097345132744e-06,
+      "loss": 0.8161,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5298672566371683,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.17533185840708e-06,
+      "loss": 0.7831,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5309734513274336,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 6.172566371681417e-06,
+      "loss": 0.7904,
+      "step": 1384
+    },
+    {
+      "epoch": 1.532079646017699,
+      "grad_norm": 0.099609375,
+      "learning_rate": 6.169800884955752e-06,
+      "loss": 0.783,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5331858407079646,
+      "grad_norm": 0.111328125,
+      "learning_rate": 6.167035398230089e-06,
+      "loss": 0.7514,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5342920353982301,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.164269911504426e-06,
+      "loss": 0.7837,
+      "step": 1387
+    },
+    {
+      "epoch": 1.5353982300884956,
+      "grad_norm": 0.103515625,
+      "learning_rate": 6.161504424778761e-06,
+      "loss": 0.7558,
+      "step": 1388
+    },
+    {
+      "epoch": 1.536504424778761,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 6.158738938053098e-06,
+      "loss": 0.8057,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5376106194690267,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.155973451327434e-06,
+      "loss": 0.7409,
+      "step": 1390
+    },
+    {
+      "epoch": 1.538716814159292,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.153207964601771e-06,
+      "loss": 0.7506,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5398230088495575,
+      "grad_norm": 0.1171875,
+      "learning_rate": 6.150442477876106e-06,
+      "loss": 0.7677,
+      "step": 1392
+    },
+    {
+      "epoch": 1.540929203539823,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 6.147676991150443e-06,
+      "loss": 0.785,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5420353982300885,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 6.14491150442478e-06,
+      "loss": 0.7919,
+      "step": 1394
+    },
+    {
+      "epoch": 1.543141592920354,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.142146017699115e-06,
+      "loss": 0.7928,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5442477876106193,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.139380530973452e-06,
+      "loss": 0.7403,
+      "step": 1396
+    },
+    {
+      "epoch": 1.545353982300885,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.136615044247788e-06,
+      "loss": 0.7423,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5464601769911503,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.133849557522124e-06,
+      "loss": 0.7626,
+      "step": 1398
+    },
+    {
+      "epoch": 1.547566371681416,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.1310840707964605e-06,
+      "loss": 0.7716,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5486725663716814,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.1283185840707975e-06,
+      "loss": 0.7424,
+      "step": 1400
+    },
+    {
+      "epoch": 1.549778761061947,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 6.125553097345133e-06,
+      "loss": 0.7253,
+      "step": 1401
+    },
+    {
+      "epoch": 1.5508849557522124,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 6.12278761061947e-06,
+      "loss": 0.7797,
+      "step": 1402
+    },
+    {
+      "epoch": 1.551991150442478,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 6.120022123893806e-06,
+      "loss": 0.7897,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5530973451327434,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.117256637168142e-06,
+      "loss": 0.7841,
+      "step": 1404
+    },
+    {
+      "epoch": 1.5542035398230087,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 6.114491150442479e-06,
+      "loss": 0.7442,
+      "step": 1405
+    },
+    {
+      "epoch": 1.5553097345132745,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 6.111725663716814e-06,
+      "loss": 0.7515,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5564159292035398,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 6.108960176991151e-06,
+      "loss": 0.7788,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5575221238938053,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 6.1061946902654865e-06,
+      "loss": 0.8172,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5586283185840708,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.1034292035398234e-06,
+      "loss": 0.8012,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5597345132743363,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.10066371681416e-06,
+      "loss": 0.7796,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5608407079646018,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.097898230088496e-06,
+      "loss": 0.7448,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5619469026548671,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.095132743362833e-06,
+      "loss": 0.7877,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5630530973451329,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 6.092367256637168e-06,
+      "loss": 0.7464,
+      "step": 1413
+    },
+    {
+      "epoch": 1.5641592920353982,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.089601769911505e-06,
+      "loss": 0.7515,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5652654867256637,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.086836283185841e-06,
+      "loss": 0.7774,
+      "step": 1415
+    },
+    {
+      "epoch": 1.5663716814159292,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.084070796460177e-06,
+      "loss": 0.9205,
+      "step": 1416
+    },
+    {
+      "epoch": 1.5674778761061947,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 6.081305309734514e-06,
+      "loss": 0.7856,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5685840707964602,
+      "grad_norm": 0.1328125,
+      "learning_rate": 6.07853982300885e-06,
+      "loss": 0.7921,
+      "step": 1418
+    },
+    {
+      "epoch": 1.5696902654867255,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.075774336283186e-06,
+      "loss": 0.7577,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5707964601769913,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.0730088495575225e-06,
+      "loss": 0.7669,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5719026548672566,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.0702433628318595e-06,
+      "loss": 0.7907,
+      "step": 1421
+    },
+    {
+      "epoch": 1.573008849557522,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.067477876106195e-06,
+      "loss": 0.7543,
+      "step": 1422
+    },
+    {
+      "epoch": 1.5741150442477876,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.064712389380532e-06,
+      "loss": 0.8084,
+      "step": 1423
+    },
+    {
+      "epoch": 1.575221238938053,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.061946902654868e-06,
+      "loss": 0.7494,
+      "step": 1424
+    },
+    {
+      "epoch": 1.5763274336283186,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.059181415929204e-06,
+      "loss": 0.7702,
+      "step": 1425
+    },
+    {
+      "epoch": 1.577433628318584,
+      "grad_norm": 0.10546875,
+      "learning_rate": 6.056415929203541e-06,
+      "loss": 0.7991,
+      "step": 1426
+    },
+    {
+      "epoch": 1.5785398230088497,
+      "grad_norm": 0.09765625,
+      "learning_rate": 6.053650442477876e-06,
+      "loss": 0.7315,
+      "step": 1427
+    },
+    {
+      "epoch": 1.579646017699115,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 6.050884955752213e-06,
+      "loss": 0.7738,
+      "step": 1428
+    },
+    {
+      "epoch": 1.5807522123893807,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.0481194690265485e-06,
+      "loss": 0.7326,
+      "step": 1429
+    },
+    {
+      "epoch": 1.581858407079646,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 6.0453539823008855e-06,
+      "loss": 0.7502,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5829646017699115,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 6.042588495575222e-06,
+      "loss": 0.8144,
+      "step": 1431
+    },
+    {
+      "epoch": 1.584070796460177,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 6.039823008849558e-06,
+      "loss": 0.8158,
+      "step": 1432
+    },
+    {
+      "epoch": 1.5851769911504425,
+      "grad_norm": 0.123046875,
+      "learning_rate": 6.037057522123895e-06,
+      "loss": 0.7925,
+      "step": 1433
+    },
+    {
+      "epoch": 1.586283185840708,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 6.03429203539823e-06,
+      "loss": 0.785,
+      "step": 1434
+    },
+    {
+      "epoch": 1.5873893805309733,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.031526548672567e-06,
+      "loss": 0.8454,
+      "step": 1435
+    },
+    {
+      "epoch": 1.588495575221239,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.028761061946903e-06,
+      "loss": 0.7438,
+      "step": 1436
+    },
+    {
+      "epoch": 1.5896017699115044,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 6.025995575221239e-06,
+      "loss": 0.7923,
+      "step": 1437
+    },
+    {
+      "epoch": 1.5907079646017699,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 6.023230088495575e-06,
+      "loss": 0.7397,
+      "step": 1438
+    },
+    {
+      "epoch": 1.5918141592920354,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.020464601769912e-06,
+      "loss": 0.8359,
+      "step": 1439
+    },
+    {
+      "epoch": 1.592920353982301,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 6.0176991150442484e-06,
+      "loss": 0.7321,
+      "step": 1440
+    },
+    {
+      "epoch": 1.5940265486725664,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.0149336283185846e-06,
+      "loss": 0.7625,
+      "step": 1441
+    },
+    {
+      "epoch": 1.5951327433628317,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 6.0121681415929215e-06,
+      "loss": 0.7821,
+      "step": 1442
+    },
+    {
+      "epoch": 1.5962389380530975,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.009402654867257e-06,
+      "loss": 0.7619,
+      "step": 1443
+    },
+    {
+      "epoch": 1.5973451327433628,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 6.006637168141594e-06,
+      "loss": 0.7333,
+      "step": 1444
+    },
+    {
+      "epoch": 1.5984513274336283,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.003871681415929e-06,
+      "loss": 0.7691,
+      "step": 1445
+    },
+    {
+      "epoch": 1.5995575221238938,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.001106194690266e-06,
+      "loss": 0.7728,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6006637168141593,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.998340707964603e-06,
+      "loss": 0.7761,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6017699115044248,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.995575221238938e-06,
+      "loss": 0.7417,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6028761061946901,
+      "grad_norm": 0.123046875,
+      "learning_rate": 5.992809734513275e-06,
+      "loss": 0.7742,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6039823008849559,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 5.9900442477876105e-06,
+      "loss": 0.7954,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6050884955752212,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.9872787610619475e-06,
+      "loss": 0.8015,
+      "step": 1451
+    },
+    {
+      "epoch": 1.606194690265487,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.984513274336284e-06,
+      "loss": 0.7438,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6073008849557522,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 5.98174778761062e-06,
+      "loss": 0.7816,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6084070796460177,
+      "grad_norm": 0.11328125,
+      "learning_rate": 5.978982300884957e-06,
+      "loss": 0.7403,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6095132743362832,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 5.976216814159293e-06,
+      "loss": 0.7814,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6106194690265485,
+      "grad_norm": 0.126953125,
+      "learning_rate": 5.973451327433629e-06,
+      "loss": 0.7485,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6117256637168142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.970685840707965e-06,
+      "loss": 0.8125,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6128318584070795,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.967920353982301e-06,
+      "loss": 0.7774,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6139380530973453,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 5.965154867256637e-06,
+      "loss": 0.7866,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6150442477876106,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 5.962389380530974e-06,
+      "loss": 0.8027,
+      "step": 1460
+    },
+    {
+      "epoch": 1.616150442477876,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.95962389380531e-06,
+      "loss": 0.7799,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6172566371681416,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.956858407079647e-06,
+      "loss": 0.8049,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6183628318584071,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.9540929203539836e-06,
+      "loss": 0.7752,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6194690265486726,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 5.951327433628319e-06,
+      "loss": 0.7555,
+      "step": 1464
+    },
+    {
+      "epoch": 1.620575221238938,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 5.948561946902656e-06,
+      "loss": 0.7794,
+      "step": 1465
+    },
+    {
+      "epoch": 1.6216814159292037,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.945796460176991e-06,
+      "loss": 0.8217,
+      "step": 1466
+    },
+    {
+      "epoch": 1.622787610619469,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.943030973451328e-06,
+      "loss": 0.748,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6238938053097345,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.940265486725663e-06,
+      "loss": 0.7414,
+      "step": 1468
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 5.9375e-06,
+      "loss": 0.7372,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6261061946902655,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.934734513274337e-06,
+      "loss": 0.7652,
+      "step": 1470
+    },
+    {
+      "epoch": 1.627212389380531,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.931969026548673e-06,
+      "loss": 0.7592,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6283185840707963,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.9292035398230096e-06,
+      "loss": 0.7615,
+      "step": 1472
+    },
+    {
+      "epoch": 1.629424778761062,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.926438053097346e-06,
+      "loss": 0.7545,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6305309734513274,
+      "grad_norm": 0.12109375,
+      "learning_rate": 5.923672566371682e-06,
+      "loss": 0.793,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6316371681415929,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.920907079646018e-06,
+      "loss": 0.754,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6327433628318584,
+      "grad_norm": 0.125,
+      "learning_rate": 5.918141592920355e-06,
+      "loss": 0.8105,
+      "step": 1476
+    },
+    {
+      "epoch": 1.633849557522124,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.915376106194691e-06,
+      "loss": 0.7868,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6349557522123894,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.912610619469027e-06,
+      "loss": 0.7598,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6360619469026547,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.909845132743363e-06,
+      "loss": 0.7796,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6371681415929205,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 5.907079646017699e-06,
+      "loss": 0.7882,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6382743362831858,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 5.904314159292036e-06,
+      "loss": 0.7824,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6393805309734515,
+      "grad_norm": 0.126953125,
+      "learning_rate": 5.901548672566372e-06,
+      "loss": 0.7533,
+      "step": 1482
+    },
+    {
+      "epoch": 1.6404867256637168,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.898783185840709e-06,
+      "loss": 0.7913,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6415929203539823,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.896017699115044e-06,
+      "loss": 0.8106,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6426991150442478,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.893252212389381e-06,
+      "loss": 0.7783,
+      "step": 1485
+    },
+    {
+      "epoch": 1.643805309734513,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.890486725663718e-06,
+      "loss": 0.7747,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6449115044247788,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.887721238938053e-06,
+      "loss": 0.748,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6460176991150441,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 5.88495575221239e-06,
+      "loss": 0.795,
+      "step": 1488
+    },
+    {
+      "epoch": 1.6471238938053099,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.882190265486725e-06,
+      "loss": 0.7913,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6482300884955752,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 5.879424778761062e-06,
+      "loss": 0.8064,
+      "step": 1490
+    },
+    {
+      "epoch": 1.6493362831858407,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 5.8766592920353985e-06,
+      "loss": 0.745,
+      "step": 1491
+    },
+    {
+      "epoch": 1.6504424778761062,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.873893805309735e-06,
+      "loss": 0.8044,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6515486725663717,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.871128318584072e-06,
+      "loss": 0.7558,
+      "step": 1493
+    },
+    {
+      "epoch": 1.6526548672566372,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.868362831858408e-06,
+      "loss": 0.7956,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6537610619469025,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.865597345132744e-06,
+      "loss": 0.7605,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6548672566371683,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 5.86283185840708e-06,
+      "loss": 0.7831,
+      "step": 1496
+    },
+    {
+      "epoch": 1.6559734513274336,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.860066371681417e-06,
+      "loss": 0.7522,
+      "step": 1497
+    },
+    {
+      "epoch": 1.657079646017699,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.857300884955752e-06,
+      "loss": 0.7712,
+      "step": 1498
+    },
+    {
+      "epoch": 1.6581858407079646,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.854535398230089e-06,
+      "loss": 0.792,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6592920353982301,
+      "grad_norm": 0.109375,
+      "learning_rate": 5.851769911504426e-06,
+      "loss": 0.792,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6603982300884956,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.8490044247787614e-06,
+      "loss": 0.7801,
+      "step": 1501
+    },
+    {
+      "epoch": 1.661504424778761,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.846238938053098e-06,
+      "loss": 0.7714,
+      "step": 1502
+    },
+    {
+      "epoch": 1.6626106194690267,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.843473451327434e-06,
+      "loss": 0.7745,
+      "step": 1503
+    },
+    {
+      "epoch": 1.663716814159292,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 5.840707964601771e-06,
+      "loss": 0.7309,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6648230088495575,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 5.837942477876106e-06,
+      "loss": 0.7794,
+      "step": 1505
+    },
+    {
+      "epoch": 1.665929203539823,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 5.835176991150443e-06,
+      "loss": 0.8119,
+      "step": 1506
+    },
+    {
+      "epoch": 1.6670353982300885,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.83241150442478e-06,
+      "loss": 0.749,
+      "step": 1507
+    },
+    {
+      "epoch": 1.668141592920354,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.829646017699115e-06,
+      "loss": 0.7454,
+      "step": 1508
+    },
+    {
+      "epoch": 1.6692477876106193,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.826880530973452e-06,
+      "loss": 0.7814,
+      "step": 1509
+    },
+    {
+      "epoch": 1.670353982300885,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.824115044247787e-06,
+      "loss": 0.8073,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6714601769911503,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 5.821349557522124e-06,
+      "loss": 0.8017,
+      "step": 1511
+    },
+    {
+      "epoch": 1.672566371681416,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.8185840707964605e-06,
+      "loss": 0.8499,
+      "step": 1512
+    },
+    {
+      "epoch": 1.6736725663716814,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.815818584070797e-06,
+      "loss": 0.7858,
+      "step": 1513
+    },
+    {
+      "epoch": 1.674778761061947,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.813053097345133e-06,
+      "loss": 0.7649,
+      "step": 1514
+    },
+    {
+      "epoch": 1.6758849557522124,
+      "grad_norm": 0.1015625,
+      "learning_rate": 5.81028761061947e-06,
+      "loss": 0.7746,
+      "step": 1515
+    },
+    {
+      "epoch": 1.676991150442478,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 5.807522123893806e-06,
+      "loss": 0.7676,
+      "step": 1516
+    },
+    {
+      "epoch": 1.6780973451327434,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 5.804756637168142e-06,
+      "loss": 0.7542,
+      "step": 1517
+    },
+    {
+      "epoch": 1.6792035398230087,
+      "grad_norm": 0.099609375,
+      "learning_rate": 5.801991150442479e-06,
+      "loss": 0.7562,
+      "step": 1518
+    },
+    {
+      "epoch": 1.6803097345132745,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.799225663716814e-06,
+      "loss": 0.8017,
+      "step": 1519
+    },
+    {
+      "epoch": 1.6814159292035398,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.796460176991151e-06,
+      "loss": 0.7577,
+      "step": 1520
+    },
+    {
+      "epoch": 1.6825221238938053,
+      "grad_norm": 0.12109375,
+      "learning_rate": 5.7936946902654865e-06,
+      "loss": 0.8063,
+      "step": 1521
+    },
+    {
+      "epoch": 1.6836283185840708,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.7909292035398235e-06,
+      "loss": 0.7456,
+      "step": 1522
+    },
+    {
+      "epoch": 1.6847345132743363,
+      "grad_norm": 0.119140625,
+      "learning_rate": 5.7881637168141604e-06,
+      "loss": 0.823,
+      "step": 1523
+    },
+    {
+      "epoch": 1.6858407079646018,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.785398230088496e-06,
+      "loss": 0.774,
+      "step": 1524
+    },
+    {
+      "epoch": 1.6869469026548671,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.782632743362833e-06,
+      "loss": 0.7451,
+      "step": 1525
+    },
+    {
+      "epoch": 1.6880530973451329,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 5.779867256637168e-06,
+      "loss": 0.7807,
+      "step": 1526
+    },
+    {
+      "epoch": 1.6891592920353982,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.777101769911505e-06,
+      "loss": 0.7594,
+      "step": 1527
+    },
+    {
+      "epoch": 1.6902654867256637,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.774336283185841e-06,
+      "loss": 0.7831,
+      "step": 1528
+    },
+    {
+      "epoch": 1.6913716814159292,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 5.771570796460177e-06,
+      "loss": 0.7465,
+      "step": 1529
+    },
+    {
+      "epoch": 1.6924778761061947,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.768805309734514e-06,
+      "loss": 0.7637,
+      "step": 1530
+    },
+    {
+      "epoch": 1.6935840707964602,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.76603982300885e-06,
+      "loss": 0.7798,
+      "step": 1531
+    },
+    {
+      "epoch": 1.6946902654867255,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.7632743362831864e-06,
+      "loss": 0.7606,
+      "step": 1532
+    },
+    {
+      "epoch": 1.6957964601769913,
+      "grad_norm": 0.11328125,
+      "learning_rate": 5.7605088495575226e-06,
+      "loss": 0.7665,
+      "step": 1533
+    },
+    {
+      "epoch": 1.6969026548672566,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 5.757743362831859e-06,
+      "loss": 0.7594,
+      "step": 1534
+    },
+    {
+      "epoch": 1.698008849557522,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.754977876106195e-06,
+      "loss": 0.7501,
+      "step": 1535
+    },
+    {
+      "epoch": 1.6991150442477876,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.752212389380532e-06,
+      "loss": 0.7838,
+      "step": 1536
+    },
+    {
+      "epoch": 1.700221238938053,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 5.749446902654868e-06,
+      "loss": 0.801,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7013274336283186,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 5.746681415929204e-06,
+      "loss": 0.7628,
+      "step": 1538
+    },
+    {
+      "epoch": 1.702433628318584,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.743915929203541e-06,
+      "loss": 0.7814,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7035398230088497,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 5.741150442477876e-06,
+      "loss": 0.7673,
+      "step": 1540
+    },
+    {
+      "epoch": 1.704646017699115,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 5.738384955752213e-06,
+      "loss": 0.8113,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7057522123893807,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.7356194690265485e-06,
+      "loss": 0.791,
+      "step": 1542
+    },
+    {
+      "epoch": 1.706858407079646,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 5.7328539823008855e-06,
+      "loss": 0.7841,
+      "step": 1543
+    },
+    {
+      "epoch": 1.7079646017699115,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.730088495575221e-06,
+      "loss": 0.7863,
+      "step": 1544
+    },
+    {
+      "epoch": 1.709070796460177,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.727323008849558e-06,
+      "loss": 0.7586,
+      "step": 1545
+    },
+    {
+      "epoch": 1.7101769911504425,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.724557522123895e-06,
+      "loss": 0.7462,
+      "step": 1546
+    },
+    {
+      "epoch": 1.711283185840708,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.72179203539823e-06,
+      "loss": 0.7541,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7123893805309733,
+      "grad_norm": 0.12109375,
+      "learning_rate": 5.719026548672567e-06,
+      "loss": 0.7956,
+      "step": 1548
+    },
+    {
+      "epoch": 1.713495575221239,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 5.716261061946903e-06,
+      "loss": 0.7797,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7146017699115044,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.713495575221239e-06,
+      "loss": 0.7608,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7157079646017699,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.710730088495575e-06,
+      "loss": 0.7717,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7168141592920354,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.707964601769912e-06,
+      "loss": 0.8648,
+      "step": 1552
+    },
+    {
+      "epoch": 1.717920353982301,
+      "grad_norm": 0.1328125,
+      "learning_rate": 5.7051991150442485e-06,
+      "loss": 0.8262,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7190265486725664,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.702433628318585e-06,
+      "loss": 0.7534,
+      "step": 1554
+    },
+    {
+      "epoch": 1.7201327433628317,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 5.6996681415929216e-06,
+      "loss": 0.7555,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7212389380530975,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.696902654867257e-06,
+      "loss": 0.8664,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7223451327433628,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 5.694137168141594e-06,
+      "loss": 0.8388,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7234513274336283,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.691371681415929e-06,
+      "loss": 0.7471,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7245575221238938,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.688606194690266e-06,
+      "loss": 0.8793,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7256637168141593,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.685840707964603e-06,
+      "loss": 0.7368,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7267699115044248,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.683075221238938e-06,
+      "loss": 0.7666,
+      "step": 1561
+    },
+    {
+      "epoch": 1.7278761061946901,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.680309734513275e-06,
+      "loss": 0.7793,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7289823008849559,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.6775442477876106e-06,
+      "loss": 0.766,
+      "step": 1563
+    },
+    {
+      "epoch": 1.7300884955752212,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.6747787610619475e-06,
+      "loss": 0.7562,
+      "step": 1564
+    },
+    {
+      "epoch": 1.731194690265487,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 5.672013274336283e-06,
+      "loss": 0.7627,
+      "step": 1565
+    },
+    {
+      "epoch": 1.7323008849557522,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.66924778761062e-06,
+      "loss": 0.7863,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7334070796460177,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.666482300884957e-06,
+      "loss": 0.7467,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7345132743362832,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 5.663716814159292e-06,
+      "loss": 0.7464,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7356194690265485,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 5.660951327433629e-06,
+      "loss": 0.7199,
+      "step": 1569
+    },
+    {
+      "epoch": 1.7367256637168142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 5.658185840707965e-06,
+      "loss": 0.784,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7378318584070795,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.655420353982301e-06,
+      "loss": 0.7578,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7389380530973453,
+      "grad_norm": 0.123046875,
+      "learning_rate": 5.652654867256637e-06,
+      "loss": 0.8062,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7400442477876106,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.649889380530974e-06,
+      "loss": 0.7389,
+      "step": 1573
+    },
+    {
+      "epoch": 1.741150442477876,
+      "grad_norm": 0.115234375,
+      "learning_rate": 5.64712389380531e-06,
+      "loss": 0.7942,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7422566371681416,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.644358407079647e-06,
+      "loss": 0.7869,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7433628318584071,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.641592920353984e-06,
+      "loss": 0.7519,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7444690265486726,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.638827433628319e-06,
+      "loss": 0.7825,
+      "step": 1577
+    },
+    {
+      "epoch": 1.745575221238938,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 5.636061946902656e-06,
+      "loss": 0.8034,
+      "step": 1578
+    },
+    {
+      "epoch": 1.7466814159292037,
+      "grad_norm": 0.119140625,
+      "learning_rate": 5.633296460176991e-06,
+      "loss": 0.7751,
+      "step": 1579
+    },
+    {
+      "epoch": 1.747787610619469,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 5.630530973451328e-06,
+      "loss": 0.7219,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7488938053097345,
+      "grad_norm": 0.1015625,
+      "learning_rate": 5.627765486725663e-06,
+      "loss": 0.7529,
+      "step": 1581
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 5.625e-06,
+      "loss": 0.7434,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7511061946902655,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.622234513274337e-06,
+      "loss": 0.7684,
+      "step": 1583
+    },
+    {
+      "epoch": 1.752212389380531,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 5.619469026548673e-06,
+      "loss": 0.7453,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7533185840707963,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.61670353982301e-06,
+      "loss": 0.8349,
+      "step": 1585
+    },
+    {
+      "epoch": 1.754424778761062,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.613938053097346e-06,
+      "loss": 0.7575,
+      "step": 1586
+    },
+    {
+      "epoch": 1.7555309734513274,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.611172566371682e-06,
+      "loss": 0.8107,
+      "step": 1587
+    },
+    {
+      "epoch": 1.7566371681415929,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 5.608407079646018e-06,
+      "loss": 0.7631,
+      "step": 1588
+    },
+    {
+      "epoch": 1.7577433628318584,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 5.605641592920354e-06,
+      "loss": 0.7943,
+      "step": 1589
+    },
+    {
+      "epoch": 1.758849557522124,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.602876106194691e-06,
+      "loss": 0.7595,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7599557522123894,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 5.600110619469027e-06,
+      "loss": 0.8044,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7610619469026547,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.597345132743363e-06,
+      "loss": 0.7731,
+      "step": 1592
+    },
+    {
+      "epoch": 1.7621681415929205,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 5.5945796460176994e-06,
+      "loss": 0.8202,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7632743362831858,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.591814159292036e-06,
+      "loss": 0.7363,
+      "step": 1594
+    },
+    {
+      "epoch": 1.7643805309734515,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.589048672566372e-06,
+      "loss": 0.7608,
+      "step": 1595
+    },
+    {
+      "epoch": 1.7654867256637168,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.586283185840709e-06,
+      "loss": 0.7677,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7665929203539823,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 5.583517699115044e-06,
+      "loss": 0.7597,
+      "step": 1597
+    },
+    {
+      "epoch": 1.7676991150442478,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 5.580752212389381e-06,
+      "loss": 0.703,
+      "step": 1598
+    },
+    {
+      "epoch": 1.768805309734513,
+      "grad_norm": 0.1171875,
+      "learning_rate": 5.577986725663718e-06,
+      "loss": 0.7977,
+      "step": 1599
+    },
+    {
+      "epoch": 1.7699115044247788,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 5.575221238938053e-06,
+      "loss": 0.79,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7710176991150441,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.57245575221239e-06,
+      "loss": 0.7494,
+      "step": 1601
+    },
+    {
+      "epoch": 1.7721238938053099,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.569690265486725e-06,
+      "loss": 0.766,
+      "step": 1602
+    },
+    {
+      "epoch": 1.7732300884955752,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.566924778761062e-06,
+      "loss": 0.7384,
+      "step": 1603
+    },
+    {
+      "epoch": 1.7743362831858407,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.5641592920353985e-06,
+      "loss": 0.768,
+      "step": 1604
+    },
+    {
+      "epoch": 1.7754424778761062,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.561393805309735e-06,
+      "loss": 0.7971,
+      "step": 1605
+    },
+    {
+      "epoch": 1.7765486725663717,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 5.558628318584072e-06,
+      "loss": 0.7716,
+      "step": 1606
+    },
+    {
+      "epoch": 1.7776548672566372,
+      "grad_norm": 0.123046875,
+      "learning_rate": 5.555862831858408e-06,
+      "loss": 0.8003,
+      "step": 1607
+    },
+    {
+      "epoch": 1.7787610619469025,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 5.553097345132744e-06,
+      "loss": 0.7576,
+      "step": 1608
+    },
+    {
+      "epoch": 1.7798672566371683,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.55033185840708e-06,
+      "loss": 0.7489,
+      "step": 1609
+    },
+    {
+      "epoch": 1.7809734513274336,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 5.547566371681416e-06,
+      "loss": 0.772,
+      "step": 1610
+    },
+    {
+      "epoch": 1.782079646017699,
+      "grad_norm": 0.125,
+      "learning_rate": 5.544800884955752e-06,
+      "loss": 0.7939,
+      "step": 1611
+    },
+    {
+      "epoch": 1.7831858407079646,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.542035398230089e-06,
+      "loss": 0.8179,
+      "step": 1612
+    },
+    {
+      "epoch": 1.7842920353982301,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.539269911504425e-06,
+      "loss": 0.7844,
+      "step": 1613
+    },
+    {
+      "epoch": 1.7853982300884956,
+      "grad_norm": 0.109375,
+      "learning_rate": 5.5365044247787615e-06,
+      "loss": 0.7914,
+      "step": 1614
+    },
+    {
+      "epoch": 1.786504424778761,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.5337389380530984e-06,
+      "loss": 0.7814,
+      "step": 1615
+    },
+    {
+      "epoch": 1.7876106194690267,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 5.530973451327434e-06,
+      "loss": 0.7672,
+      "step": 1616
+    },
+    {
+      "epoch": 1.788716814159292,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.528207964601771e-06,
+      "loss": 0.778,
+      "step": 1617
+    },
+    {
+      "epoch": 1.7898230088495575,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.525442477876106e-06,
+      "loss": 0.7963,
+      "step": 1618
+    },
+    {
+      "epoch": 1.790929203539823,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 5.522676991150443e-06,
+      "loss": 0.7816,
+      "step": 1619
+    },
+    {
+      "epoch": 1.7920353982300885,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.51991150442478e-06,
+      "loss": 0.7291,
+      "step": 1620
+    },
+    {
+      "epoch": 1.793141592920354,
+      "grad_norm": 0.119140625,
+      "learning_rate": 5.517146017699115e-06,
+      "loss": 0.793,
+      "step": 1621
+    },
+    {
+      "epoch": 1.7942477876106193,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 5.514380530973452e-06,
+      "loss": 0.7313,
+      "step": 1622
+    },
+    {
+      "epoch": 1.795353982300885,
+      "grad_norm": 0.09765625,
+      "learning_rate": 5.5116150442477875e-06,
+      "loss": 0.7659,
+      "step": 1623
+    },
+    {
+      "epoch": 1.7964601769911503,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.508849557522124e-06,
+      "loss": 0.8055,
+      "step": 1624
+    },
+    {
+      "epoch": 1.797566371681416,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.5060840707964605e-06,
+      "loss": 0.7291,
+      "step": 1625
+    },
+    {
+      "epoch": 1.7986725663716814,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.503318584070797e-06,
+      "loss": 0.7242,
+      "step": 1626
+    },
+    {
+      "epoch": 1.799778761061947,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.500553097345133e-06,
+      "loss": 0.7585,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8008849557522124,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.49778761061947e-06,
+      "loss": 0.797,
+      "step": 1628
+    },
+    {
+      "epoch": 1.801991150442478,
+      "grad_norm": 0.09765625,
+      "learning_rate": 5.495022123893806e-06,
+      "loss": 0.7372,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8030973451327434,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 5.492256637168142e-06,
+      "loss": 0.7616,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8042035398230087,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.489491150442479e-06,
+      "loss": 0.7653,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8053097345132745,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 5.486725663716814e-06,
+      "loss": 0.8818,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8064159292035398,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.483960176991151e-06,
+      "loss": 0.7661,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8075221238938053,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 5.4811946902654865e-06,
+      "loss": 0.7653,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8086283185840708,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.4784292035398235e-06,
+      "loss": 0.7461,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8097345132743363,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.4756637168141605e-06,
+      "loss": 0.7205,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8108407079646018,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.472898230088496e-06,
+      "loss": 0.7747,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8119469026548671,
+      "grad_norm": 0.125,
+      "learning_rate": 5.470132743362833e-06,
+      "loss": 0.8055,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8130530973451329,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.467367256637168e-06,
+      "loss": 0.7803,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8141592920353982,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 5.464601769911505e-06,
+      "loss": 0.6922,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8152654867256637,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.461836283185841e-06,
+      "loss": 0.8039,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8163716814159292,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.459070796460177e-06,
+      "loss": 0.816,
+      "step": 1642
+    },
+    {
+      "epoch": 1.8174778761061947,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.456305309734514e-06,
+      "loss": 0.7618,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8185840707964602,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 5.4535398230088495e-06,
+      "loss": 0.7482,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8196902654867255,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.4507743362831865e-06,
+      "loss": 0.779,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8207964601769913,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 5.448008849557523e-06,
+      "loss": 0.7587,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8219026548672566,
+      "grad_norm": 0.09765625,
+      "learning_rate": 5.445243362831859e-06,
+      "loss": 0.7882,
+      "step": 1647
+    },
+    {
+      "epoch": 1.823008849557522,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.442477876106195e-06,
+      "loss": 0.7451,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8241150442477876,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.439712389380532e-06,
+      "loss": 0.7763,
+      "step": 1649
+    },
+    {
+      "epoch": 1.825221238938053,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.436946902654868e-06,
+      "loss": 0.7994,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8263274336283186,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 5.434181415929204e-06,
+      "loss": 0.7784,
+      "step": 1651
+    },
+    {
+      "epoch": 1.827433628318584,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.431415929203541e-06,
+      "loss": 0.7398,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8285398230088497,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.428650442477876e-06,
+      "loss": 0.7499,
+      "step": 1653
+    },
+    {
+      "epoch": 1.829646017699115,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.425884955752213e-06,
+      "loss": 0.7481,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8307522123893807,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 5.4231194690265486e-06,
+      "loss": 0.7382,
+      "step": 1655
+    },
+    {
+      "epoch": 1.831858407079646,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.4203539823008855e-06,
+      "loss": 0.7804,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8329646017699115,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 5.417588495575221e-06,
+      "loss": 0.7561,
+      "step": 1657
+    },
+    {
+      "epoch": 1.834070796460177,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 5.414823008849558e-06,
+      "loss": 0.7346,
+      "step": 1658
+    },
+    {
+      "epoch": 1.8351769911504425,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 5.412057522123895e-06,
+      "loss": 0.7562,
+      "step": 1659
+    },
+    {
+      "epoch": 1.836283185840708,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.40929203539823e-06,
+      "loss": 0.7989,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8373893805309733,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 5.406526548672567e-06,
+      "loss": 0.7063,
+      "step": 1661
+    },
+    {
+      "epoch": 1.838495575221239,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.403761061946903e-06,
+      "loss": 0.7859,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8396017699115044,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.400995575221239e-06,
+      "loss": 0.8041,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8407079646017699,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.398230088495575e-06,
+      "loss": 0.725,
+      "step": 1664
+    },
+    {
+      "epoch": 1.8418141592920354,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.3954646017699115e-06,
+      "loss": 0.7809,
+      "step": 1665
+    },
+    {
+      "epoch": 1.842920353982301,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 5.3926991150442485e-06,
+      "loss": 0.7783,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8440265486725664,
+      "grad_norm": 0.115234375,
+      "learning_rate": 5.389933628318585e-06,
+      "loss": 0.8061,
+      "step": 1667
+    },
+    {
+      "epoch": 1.8451327433628317,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 5.387168141592921e-06,
+      "loss": 0.7714,
+      "step": 1668
+    },
+    {
+      "epoch": 1.8462389380530975,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.384402654867257e-06,
+      "loss": 0.7685,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8473451327433628,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.381637168141594e-06,
+      "loss": 0.7653,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8484513274336283,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.378871681415929e-06,
+      "loss": 0.8039,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8495575221238938,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.376106194690266e-06,
+      "loss": 0.7866,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8506637168141593,
+      "grad_norm": 0.11328125,
+      "learning_rate": 5.373340707964603e-06,
+      "loss": 0.7821,
+      "step": 1673
+    },
+    {
+      "epoch": 1.8517699115044248,
+      "grad_norm": 0.109375,
+      "learning_rate": 5.370575221238938e-06,
+      "loss": 0.7915,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8528761061946901,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.367809734513275e-06,
+      "loss": 0.7781,
+      "step": 1675
+    },
+    {
+      "epoch": 1.8539823008849559,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.365044247787611e-06,
+      "loss": 0.7975,
+      "step": 1676
+    },
+    {
+      "epoch": 1.8550884955752212,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.3622787610619476e-06,
+      "loss": 0.7934,
+      "step": 1677
+    },
+    {
+      "epoch": 1.856194690265487,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.359513274336283e-06,
+      "loss": 0.7418,
+      "step": 1678
+    },
+    {
+      "epoch": 1.8573008849557522,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 5.35674778761062e-06,
+      "loss": 0.7796,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8584070796460177,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.353982300884957e-06,
+      "loss": 0.7647,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8595132743362832,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 5.351216814159292e-06,
+      "loss": 0.75,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8606194690265485,
+      "grad_norm": 0.1015625,
+      "learning_rate": 5.348451327433629e-06,
+      "loss": 0.8062,
+      "step": 1682
+    },
+    {
+      "epoch": 1.8617256637168142,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.345685840707965e-06,
+      "loss": 0.7335,
+      "step": 1683
+    },
+    {
+      "epoch": 1.8628318584070795,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 5.342920353982301e-06,
+      "loss": 0.7592,
+      "step": 1684
+    },
+    {
+      "epoch": 1.8639380530973453,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.3401548672566374e-06,
+      "loss": 0.8063,
+      "step": 1685
+    },
+    {
+      "epoch": 1.8650442477876106,
+      "grad_norm": 0.109375,
+      "learning_rate": 5.337389380530974e-06,
+      "loss": 0.7972,
+      "step": 1686
+    },
+    {
+      "epoch": 1.866150442477876,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.33462389380531e-06,
+      "loss": 0.7724,
+      "step": 1687
+    },
+    {
+      "epoch": 1.8672566371681416,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 5.331858407079647e-06,
+      "loss": 0.7316,
+      "step": 1688
+    },
+    {
+      "epoch": 1.8683628318584071,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 5.329092920353983e-06,
+      "loss": 0.762,
+      "step": 1689
+    },
+    {
+      "epoch": 1.8694690265486726,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.326327433628319e-06,
+      "loss": 0.7526,
+      "step": 1690
+    },
+    {
+      "epoch": 1.870575221238938,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 5.323561946902656e-06,
+      "loss": 0.7266,
+      "step": 1691
+    },
+    {
+      "epoch": 1.8716814159292037,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.320796460176991e-06,
+      "loss": 0.7636,
+      "step": 1692
+    },
+    {
+      "epoch": 1.872787610619469,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 5.318030973451328e-06,
+      "loss": 0.7351,
+      "step": 1693
+    },
+    {
+      "epoch": 1.8738938053097345,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.315265486725663e-06,
+      "loss": 0.7794,
+      "step": 1694
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.3125e-06,
+      "loss": 0.7087,
+      "step": 1695
+    },
+    {
+      "epoch": 1.8761061946902655,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.309734513274337e-06,
+      "loss": 0.7563,
+      "step": 1696
+    },
+    {
+      "epoch": 1.877212389380531,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.306969026548673e-06,
+      "loss": 0.8121,
+      "step": 1697
+    },
+    {
+      "epoch": 1.8783185840707963,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 5.30420353982301e-06,
+      "loss": 0.7873,
+      "step": 1698
+    },
+    {
+      "epoch": 1.879424778761062,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.301438053097345e-06,
+      "loss": 0.7744,
+      "step": 1699
+    },
+    {
+      "epoch": 1.8805309734513274,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 5.298672566371682e-06,
+      "loss": 0.7798,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8816371681415929,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.295907079646018e-06,
+      "loss": 0.7847,
+      "step": 1701
+    },
+    {
+      "epoch": 1.8827433628318584,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.293141592920354e-06,
+      "loss": 0.7935,
+      "step": 1702
+    },
+    {
+      "epoch": 1.883849557522124,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.290376106194691e-06,
+      "loss": 0.7791,
+      "step": 1703
+    },
+    {
+      "epoch": 1.8849557522123894,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 5.287610619469027e-06,
+      "loss": 0.7798,
+      "step": 1704
+    },
+    {
+      "epoch": 1.8860619469026547,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.284845132743363e-06,
+      "loss": 0.7475,
+      "step": 1705
+    },
+    {
+      "epoch": 1.8871681415929205,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.2820796460176995e-06,
+      "loss": 0.7799,
+      "step": 1706
+    },
+    {
+      "epoch": 1.8882743362831858,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.2793141592920364e-06,
+      "loss": 0.7435,
+      "step": 1707
+    },
+    {
+      "epoch": 1.8893805309734515,
+      "grad_norm": 0.1171875,
+      "learning_rate": 5.276548672566372e-06,
+      "loss": 0.753,
+      "step": 1708
+    },
+    {
+      "epoch": 1.8904867256637168,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.273783185840709e-06,
+      "loss": 0.8119,
+      "step": 1709
+    },
+    {
+      "epoch": 1.8915929203539823,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.271017699115044e-06,
+      "loss": 0.7779,
+      "step": 1710
+    },
+    {
+      "epoch": 1.8926991150442478,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 5.268252212389381e-06,
+      "loss": 0.7643,
+      "step": 1711
+    },
+    {
+      "epoch": 1.893805309734513,
+      "grad_norm": 0.109375,
+      "learning_rate": 5.265486725663718e-06,
+      "loss": 0.782,
+      "step": 1712
+    },
+    {
+      "epoch": 1.8949115044247788,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.262721238938053e-06,
+      "loss": 0.7611,
+      "step": 1713
+    },
+    {
+      "epoch": 1.8960176991150441,
+      "grad_norm": 0.115234375,
+      "learning_rate": 5.25995575221239e-06,
+      "loss": 0.8026,
+      "step": 1714
+    },
+    {
+      "epoch": 1.8971238938053099,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.2571902654867254e-06,
+      "loss": 0.7635,
+      "step": 1715
+    },
+    {
+      "epoch": 1.8982300884955752,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.254424778761062e-06,
+      "loss": 0.7746,
+      "step": 1716
+    },
+    {
+      "epoch": 1.8993362831858407,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 5.2516592920353985e-06,
+      "loss": 0.7742,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9004424778761062,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 5.248893805309735e-06,
+      "loss": 0.7733,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9015486725663717,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 5.246128318584072e-06,
+      "loss": 0.7698,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9026548672566372,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.243362831858407e-06,
+      "loss": 0.7838,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9037610619469025,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.240597345132744e-06,
+      "loss": 0.7286,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9048672566371683,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.23783185840708e-06,
+      "loss": 0.7633,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9059734513274336,
+      "grad_norm": 0.125,
+      "learning_rate": 5.235066371681416e-06,
+      "loss": 0.8256,
+      "step": 1723
+    },
+    {
+      "epoch": 1.907079646017699,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.232300884955752e-06,
+      "loss": 0.7846,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9081858407079646,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.229535398230089e-06,
+      "loss": 0.7704,
+      "step": 1725
+    },
+    {
+      "epoch": 1.9092920353982301,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 5.226769911504425e-06,
+      "loss": 0.796,
+      "step": 1726
+    },
+    {
+      "epoch": 1.9103982300884956,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.2240044247787615e-06,
+      "loss": 0.7357,
+      "step": 1727
+    },
+    {
+      "epoch": 1.911504424778761,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.2212389380530985e-06,
+      "loss": 0.8218,
+      "step": 1728
+    },
+    {
+      "epoch": 1.9126106194690267,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.218473451327434e-06,
+      "loss": 0.8054,
+      "step": 1729
+    },
+    {
+      "epoch": 1.913716814159292,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 5.215707964601771e-06,
+      "loss": 0.7697,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9148230088495575,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 5.212942477876106e-06,
+      "loss": 0.7932,
+      "step": 1731
+    },
+    {
+      "epoch": 1.915929203539823,
+      "grad_norm": 0.1015625,
+      "learning_rate": 5.210176991150443e-06,
+      "loss": 0.7608,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9170353982300885,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 5.20741150442478e-06,
+      "loss": 0.7856,
+      "step": 1733
+    },
+    {
+      "epoch": 1.918141592920354,
+      "grad_norm": 0.1015625,
+      "learning_rate": 5.204646017699115e-06,
+      "loss": 0.7109,
+      "step": 1734
+    },
+    {
+      "epoch": 1.9192477876106193,
+      "grad_norm": 0.09765625,
+      "learning_rate": 5.201880530973452e-06,
+      "loss": 0.7529,
+      "step": 1735
+    },
+    {
+      "epoch": 1.920353982300885,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.1991150442477875e-06,
+      "loss": 0.801,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9214601769911503,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.1963495575221245e-06,
+      "loss": 0.7589,
+      "step": 1737
+    },
+    {
+      "epoch": 1.922566371681416,
+      "grad_norm": 0.115234375,
+      "learning_rate": 5.193584070796461e-06,
+      "loss": 0.8137,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9236725663716814,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 5.190818584070797e-06,
+      "loss": 0.7658,
+      "step": 1739
+    },
+    {
+      "epoch": 1.924778761061947,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 5.188053097345133e-06,
+      "loss": 0.759,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9258849557522124,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.18528761061947e-06,
+      "loss": 0.7571,
+      "step": 1741
+    },
+    {
+      "epoch": 1.926991150442478,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.182522123893806e-06,
+      "loss": 0.7533,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9280973451327434,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.179756637168142e-06,
+      "loss": 0.7209,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9292035398230087,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 5.176991150442478e-06,
+      "loss": 0.7581,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9303097345132745,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.174225663716814e-06,
+      "loss": 0.7405,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9314159292035398,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 5.171460176991151e-06,
+      "loss": 0.7855,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9325221238938053,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.1686946902654866e-06,
+      "loss": 0.7399,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9336283185840708,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.1659292035398235e-06,
+      "loss": 0.7756,
+      "step": 1748
+    },
+    {
+      "epoch": 1.9347345132743363,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 5.1631637168141605e-06,
+      "loss": 0.7713,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9358407079646018,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.160398230088496e-06,
+      "loss": 0.778,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9369469026548671,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 5.157632743362833e-06,
+      "loss": 0.8,
+      "step": 1751
+    },
+    {
+      "epoch": 1.9380530973451329,
+      "grad_norm": 0.099609375,
+      "learning_rate": 5.154867256637168e-06,
+      "loss": 0.7415,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9391592920353982,
+      "grad_norm": 0.1171875,
+      "learning_rate": 5.152101769911505e-06,
+      "loss": 0.8,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9402654867256637,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.14933628318584e-06,
+      "loss": 0.7532,
+      "step": 1754
+    },
+    {
+      "epoch": 1.9413716814159292,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.146570796460177e-06,
+      "loss": 0.7722,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9424778761061947,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.143805309734514e-06,
+      "loss": 0.7343,
+      "step": 1756
+    },
+    {
+      "epoch": 1.9435840707964602,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 5.1410398230088495e-06,
+      "loss": 0.7747,
+      "step": 1757
+    },
+    {
+      "epoch": 1.9446902654867255,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.1382743362831865e-06,
+      "loss": 0.7826,
+      "step": 1758
+    },
+    {
+      "epoch": 1.9457964601769913,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.135508849557523e-06,
+      "loss": 0.7741,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9469026548672566,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 5.132743362831859e-06,
+      "loss": 0.7476,
+      "step": 1760
+    },
+    {
+      "epoch": 1.948008849557522,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 5.129977876106195e-06,
+      "loss": 0.8713,
+      "step": 1761
+    },
+    {
+      "epoch": 1.9491150442477876,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 5.127212389380532e-06,
+      "loss": 0.7594,
+      "step": 1762
+    },
+    {
+      "epoch": 1.950221238938053,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.124446902654868e-06,
+      "loss": 0.7727,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9513274336283186,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 5.121681415929204e-06,
+      "loss": 0.7307,
+      "step": 1764
+    },
+    {
+      "epoch": 1.952433628318584,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 5.11891592920354e-06,
+      "loss": 0.7514,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9535398230088497,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 5.116150442477876e-06,
+      "loss": 0.7722,
+      "step": 1766
+    },
+    {
+      "epoch": 1.954646017699115,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.113384955752213e-06,
+      "loss": 0.7512,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9557522123893807,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.110619469026549e-06,
+      "loss": 0.7894,
+      "step": 1768
+    },
+    {
+      "epoch": 1.956858407079646,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.1078539823008856e-06,
+      "loss": 0.7474,
+      "step": 1769
+    },
+    {
+      "epoch": 1.9579646017699115,
+      "grad_norm": 0.1171875,
+      "learning_rate": 5.105088495575221e-06,
+      "loss": 0.7485,
+      "step": 1770
+    },
+    {
+      "epoch": 1.959070796460177,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 5.102323008849558e-06,
+      "loss": 0.832,
+      "step": 1771
+    },
+    {
+      "epoch": 1.9601769911504425,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.099557522123895e-06,
+      "loss": 0.7507,
+      "step": 1772
+    },
+    {
+      "epoch": 1.961283185840708,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.09679203539823e-06,
+      "loss": 0.7685,
+      "step": 1773
+    },
+    {
+      "epoch": 1.9623893805309733,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 5.094026548672567e-06,
+      "loss": 0.7377,
+      "step": 1774
+    },
+    {
+      "epoch": 1.963495575221239,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.091261061946902e-06,
+      "loss": 0.7624,
+      "step": 1775
+    },
+    {
+      "epoch": 1.9646017699115044,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 5.088495575221239e-06,
+      "loss": 0.748,
+      "step": 1776
+    },
+    {
+      "epoch": 1.9657079646017699,
+      "grad_norm": 0.1171875,
+      "learning_rate": 5.085730088495575e-06,
+      "loss": 0.7834,
+      "step": 1777
+    },
+    {
+      "epoch": 1.9668141592920354,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.0829646017699115e-06,
+      "loss": 0.757,
+      "step": 1778
+    },
+    {
+      "epoch": 1.967920353982301,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 5.0801991150442485e-06,
+      "loss": 0.7936,
+      "step": 1779
+    },
+    {
+      "epoch": 1.9690265486725664,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 5.077433628318585e-06,
+      "loss": 0.751,
+      "step": 1780
+    },
+    {
+      "epoch": 1.9701327433628317,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 5.074668141592921e-06,
+      "loss": 0.7482,
+      "step": 1781
+    },
+    {
+      "epoch": 1.9712389380530975,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 5.071902654867257e-06,
+      "loss": 0.7505,
+      "step": 1782
+    },
+    {
+      "epoch": 1.9723451327433628,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.069137168141594e-06,
+      "loss": 0.7741,
+      "step": 1783
+    },
+    {
+      "epoch": 1.9734513274336283,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.066371681415929e-06,
+      "loss": 0.773,
+      "step": 1784
+    },
+    {
+      "epoch": 1.9745575221238938,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 5.063606194690266e-06,
+      "loss": 0.8268,
+      "step": 1785
+    },
+    {
+      "epoch": 1.9756637168141593,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.060840707964603e-06,
+      "loss": 0.7862,
+      "step": 1786
+    },
+    {
+      "epoch": 1.9767699115044248,
+      "grad_norm": 0.1015625,
+      "learning_rate": 5.058075221238938e-06,
+      "loss": 0.7046,
+      "step": 1787
+    },
+    {
+      "epoch": 1.9778761061946901,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.055309734513275e-06,
+      "loss": 0.7902,
+      "step": 1788
+    },
+    {
+      "epoch": 1.9789823008849559,
+      "grad_norm": 0.119140625,
+      "learning_rate": 5.052544247787611e-06,
+      "loss": 0.745,
+      "step": 1789
+    },
+    {
+      "epoch": 1.9800884955752212,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 5.049778761061948e-06,
+      "loss": 0.8201,
+      "step": 1790
+    },
+    {
+      "epoch": 1.981194690265487,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.047013274336283e-06,
+      "loss": 0.7484,
+      "step": 1791
+    },
+    {
+      "epoch": 1.9823008849557522,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.04424778761062e-06,
+      "loss": 0.7657,
+      "step": 1792
+    },
+    {
+      "epoch": 1.9834070796460177,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 5.041482300884957e-06,
+      "loss": 0.8077,
+      "step": 1793
+    },
+    {
+      "epoch": 1.9845132743362832,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 5.038716814159292e-06,
+      "loss": 0.7629,
+      "step": 1794
+    },
+    {
+      "epoch": 1.9856194690265485,
+      "grad_norm": 0.103515625,
+      "learning_rate": 5.035951327433629e-06,
+      "loss": 0.8018,
+      "step": 1795
+    },
+    {
+      "epoch": 1.9867256637168142,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.033185840707965e-06,
+      "loss": 0.7589,
+      "step": 1796
+    },
+    {
+      "epoch": 1.9878318584070795,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 5.030420353982301e-06,
+      "loss": 0.7971,
+      "step": 1797
+    },
+    {
+      "epoch": 1.9889380530973453,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.0276548672566375e-06,
+      "loss": 0.8354,
+      "step": 1798
+    },
+    {
+      "epoch": 1.9900442477876106,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 5.024889380530974e-06,
+      "loss": 0.7938,
+      "step": 1799
+    },
+    {
+      "epoch": 1.991150442477876,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 5.02212389380531e-06,
+      "loss": 0.8132,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9922566371681416,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 5.019358407079647e-06,
+      "loss": 0.8051,
+      "step": 1801
+    },
+    {
+      "epoch": 1.9933628318584071,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.016592920353983e-06,
+      "loss": 0.7749,
+      "step": 1802
+    },
+    {
+      "epoch": 1.9944690265486726,
+      "grad_norm": 0.099609375,
+      "learning_rate": 5.013827433628319e-06,
+      "loss": 0.7691,
+      "step": 1803
+    },
+    {
+      "epoch": 1.995575221238938,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.011061946902656e-06,
+      "loss": 0.7471,
+      "step": 1804
+    },
+    {
+      "epoch": 1.9966814159292037,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 5.008296460176991e-06,
+      "loss": 0.7659,
+      "step": 1805
+    },
+    {
+      "epoch": 1.997787610619469,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.005530973451328e-06,
+      "loss": 0.7529,
+      "step": 1806
+    },
+    {
+      "epoch": 1.9988938053097345,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.0027654867256634e-06,
+      "loss": 0.7637,
+      "step": 1807
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 5e-06,
+      "loss": 0.799,
+      "step": 1808
+    },
+    {
+      "epoch": 2.0011061946902653,
+      "grad_norm": 0.130859375,
+      "learning_rate": 4.9972345132743365e-06,
+      "loss": 0.7357,
+      "step": 1809
+    },
+    {
+      "epoch": 2.002212389380531,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.994469026548673e-06,
+      "loss": 0.7745,
+      "step": 1810
+    },
+    {
+      "epoch": 2.0033185840707963,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.991703539823009e-06,
+      "loss": 0.7839,
+      "step": 1811
+    },
+    {
+      "epoch": 2.004424778761062,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 4.988938053097346e-06,
+      "loss": 0.7809,
+      "step": 1812
+    },
+    {
+      "epoch": 2.0055309734513274,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.986172566371682e-06,
+      "loss": 0.7667,
+      "step": 1813
+    },
+    {
+      "epoch": 2.006637168141593,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.983407079646018e-06,
+      "loss": 0.7629,
+      "step": 1814
+    },
+    {
+      "epoch": 2.0077433628318584,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 4.980641592920354e-06,
+      "loss": 0.816,
+      "step": 1815
+    },
+    {
+      "epoch": 2.0088495575221237,
+      "grad_norm": 0.119140625,
+      "learning_rate": 4.97787610619469e-06,
+      "loss": 0.7724,
+      "step": 1816
+    },
+    {
+      "epoch": 2.0099557522123894,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.975110619469027e-06,
+      "loss": 0.7728,
+      "step": 1817
+    },
+    {
+      "epoch": 2.0110619469026547,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 4.972345132743363e-06,
+      "loss": 0.7711,
+      "step": 1818
+    },
+    {
+      "epoch": 2.0121681415929205,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.9695796460176995e-06,
+      "loss": 0.7921,
+      "step": 1819
+    },
+    {
+      "epoch": 2.0132743362831858,
+      "grad_norm": 0.125,
+      "learning_rate": 4.966814159292036e-06,
+      "loss": 0.7754,
+      "step": 1820
+    },
+    {
+      "epoch": 2.0143805309734515,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 4.964048672566373e-06,
+      "loss": 0.7914,
+      "step": 1821
+    },
+    {
+      "epoch": 2.015486725663717,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 4.961283185840709e-06,
+      "loss": 0.7801,
+      "step": 1822
+    },
+    {
+      "epoch": 2.0165929203539825,
+      "grad_norm": 0.099609375,
+      "learning_rate": 4.958517699115045e-06,
+      "loss": 0.7433,
+      "step": 1823
+    },
+    {
+      "epoch": 2.017699115044248,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.955752212389381e-06,
+      "loss": 0.7993,
+      "step": 1824
+    },
+    {
+      "epoch": 2.018805309734513,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 4.952986725663717e-06,
+      "loss": 0.8008,
+      "step": 1825
+    },
+    {
+      "epoch": 2.019911504424779,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.950221238938053e-06,
+      "loss": 0.7929,
+      "step": 1826
+    },
+    {
+      "epoch": 2.021017699115044,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.94745575221239e-06,
+      "loss": 0.774,
+      "step": 1827
+    },
+    {
+      "epoch": 2.02212389380531,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 4.944690265486726e-06,
+      "loss": 0.7876,
+      "step": 1828
+    },
+    {
+      "epoch": 2.023230088495575,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.9419247787610624e-06,
+      "loss": 0.818,
+      "step": 1829
+    },
+    {
+      "epoch": 2.024336283185841,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.9391592920353986e-06,
+      "loss": 0.7626,
+      "step": 1830
+    },
+    {
+      "epoch": 2.025442477876106,
+      "grad_norm": 0.162109375,
+      "learning_rate": 4.936393805309735e-06,
+      "loss": 0.7437,
+      "step": 1831
+    },
+    {
+      "epoch": 2.0265486725663715,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.933628318584071e-06,
+      "loss": 0.7631,
+      "step": 1832
+    },
+    {
+      "epoch": 2.0276548672566372,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 4.930862831858407e-06,
+      "loss": 0.7907,
+      "step": 1833
+    },
+    {
+      "epoch": 2.0287610619469025,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.928097345132744e-06,
+      "loss": 0.8103,
+      "step": 1834
+    },
+    {
+      "epoch": 2.0298672566371683,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 4.92533185840708e-06,
+      "loss": 0.7289,
+      "step": 1835
+    },
+    {
+      "epoch": 2.0309734513274336,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.922566371681416e-06,
+      "loss": 0.8503,
+      "step": 1836
+    },
+    {
+      "epoch": 2.0320796460176993,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 4.919800884955752e-06,
+      "loss": 0.7663,
+      "step": 1837
+    },
+    {
+      "epoch": 2.0331858407079646,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.917035398230089e-06,
+      "loss": 0.7767,
+      "step": 1838
+    },
+    {
+      "epoch": 2.03429203539823,
+      "grad_norm": 0.09765625,
+      "learning_rate": 4.914269911504425e-06,
+      "loss": 0.7541,
+      "step": 1839
+    },
+    {
+      "epoch": 2.0353982300884956,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 4.9115044247787615e-06,
+      "loss": 0.785,
+      "step": 1840
+    },
+    {
+      "epoch": 2.036504424778761,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.908738938053098e-06,
+      "loss": 0.771,
+      "step": 1841
+    },
+    {
+      "epoch": 2.0376106194690267,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.905973451327435e-06,
+      "loss": 0.8014,
+      "step": 1842
+    },
+    {
+      "epoch": 2.038716814159292,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.903207964601771e-06,
+      "loss": 0.7512,
+      "step": 1843
+    },
+    {
+      "epoch": 2.0398230088495577,
+      "grad_norm": 0.1015625,
+      "learning_rate": 4.900442477876107e-06,
+      "loss": 0.788,
+      "step": 1844
+    },
+    {
+      "epoch": 2.040929203539823,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 4.897676991150443e-06,
+      "loss": 0.7955,
+      "step": 1845
+    },
+    {
+      "epoch": 2.0420353982300883,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.894911504424779e-06,
+      "loss": 0.8227,
+      "step": 1846
+    },
+    {
+      "epoch": 2.043141592920354,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 4.892146017699115e-06,
+      "loss": 0.7276,
+      "step": 1847
+    },
+    {
+      "epoch": 2.0442477876106193,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.889380530973451e-06,
+      "loss": 0.7395,
+      "step": 1848
+    },
+    {
+      "epoch": 2.045353982300885,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 4.8866150442477875e-06,
+      "loss": 0.7525,
+      "step": 1849
+    },
+    {
+      "epoch": 2.0464601769911503,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.8838495575221245e-06,
+      "loss": 0.7671,
+      "step": 1850
+    },
+    {
+      "epoch": 2.047566371681416,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.881084070796461e-06,
+      "loss": 0.7164,
+      "step": 1851
+    },
+    {
+      "epoch": 2.0486725663716814,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 4.878318584070797e-06,
+      "loss": 0.7568,
+      "step": 1852
+    },
+    {
+      "epoch": 2.049778761061947,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 4.875553097345133e-06,
+      "loss": 0.8035,
+      "step": 1853
+    },
+    {
+      "epoch": 2.0508849557522124,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.872787610619469e-06,
+      "loss": 0.7511,
+      "step": 1854
+    },
+    {
+      "epoch": 2.0519911504424777,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.870022123893806e-06,
+      "loss": 0.7601,
+      "step": 1855
+    },
+    {
+      "epoch": 2.0530973451327434,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.867256637168142e-06,
+      "loss": 0.7728,
+      "step": 1856
+    },
+    {
+      "epoch": 2.0542035398230087,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.864491150442478e-06,
+      "loss": 0.7757,
+      "step": 1857
+    },
+    {
+      "epoch": 2.0553097345132745,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.861725663716815e-06,
+      "loss": 0.7742,
+      "step": 1858
+    },
+    {
+      "epoch": 2.0564159292035398,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 4.858960176991151e-06,
+      "loss": 0.7539,
+      "step": 1859
+    },
+    {
+      "epoch": 2.0575221238938055,
+      "grad_norm": 0.123046875,
+      "learning_rate": 4.8561946902654874e-06,
+      "loss": 0.7929,
+      "step": 1860
+    },
+    {
+      "epoch": 2.058628318584071,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.8534292035398236e-06,
+      "loss": 0.7565,
+      "step": 1861
+    },
+    {
+      "epoch": 2.059734513274336,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.85066371681416e-06,
+      "loss": 0.772,
+      "step": 1862
+    },
+    {
+      "epoch": 2.060840707964602,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 4.847898230088496e-06,
+      "loss": 0.7779,
+      "step": 1863
+    },
+    {
+      "epoch": 2.061946902654867,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.845132743362832e-06,
+      "loss": 0.7598,
+      "step": 1864
+    },
+    {
+      "epoch": 2.063053097345133,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.842367256637169e-06,
+      "loss": 0.7551,
+      "step": 1865
+    },
+    {
+      "epoch": 2.064159292035398,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.839601769911505e-06,
+      "loss": 0.7632,
+      "step": 1866
+    },
+    {
+      "epoch": 2.065265486725664,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.836836283185841e-06,
+      "loss": 0.7393,
+      "step": 1867
+    },
+    {
+      "epoch": 2.066371681415929,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.834070796460177e-06,
+      "loss": 0.7961,
+      "step": 1868
+    },
+    {
+      "epoch": 2.0674778761061945,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.831305309734513e-06,
+      "loss": 0.7352,
+      "step": 1869
+    },
+    {
+      "epoch": 2.0685840707964602,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.8285398230088495e-06,
+      "loss": 0.7647,
+      "step": 1870
+    },
+    {
+      "epoch": 2.0696902654867255,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.825774336283186e-06,
+      "loss": 0.8431,
+      "step": 1871
+    },
+    {
+      "epoch": 2.0707964601769913,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 4.823008849557523e-06,
+      "loss": 0.7508,
+      "step": 1872
+    },
+    {
+      "epoch": 2.0719026548672566,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.820243362831859e-06,
+      "loss": 0.7759,
+      "step": 1873
+    },
+    {
+      "epoch": 2.0730088495575223,
+      "grad_norm": 0.126953125,
+      "learning_rate": 4.817477876106195e-06,
+      "loss": 0.7535,
+      "step": 1874
+    },
+    {
+      "epoch": 2.0741150442477876,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.814712389380531e-06,
+      "loss": 0.7337,
+      "step": 1875
+    },
+    {
+      "epoch": 2.0752212389380533,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.811946902654868e-06,
+      "loss": 0.7979,
+      "step": 1876
+    },
+    {
+      "epoch": 2.0763274336283186,
+      "grad_norm": 0.095703125,
+      "learning_rate": 4.809181415929204e-06,
+      "loss": 0.7282,
+      "step": 1877
+    },
+    {
+      "epoch": 2.077433628318584,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 4.80641592920354e-06,
+      "loss": 0.758,
+      "step": 1878
+    },
+    {
+      "epoch": 2.0785398230088497,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.803650442477876e-06,
+      "loss": 0.7796,
+      "step": 1879
+    },
+    {
+      "epoch": 2.079646017699115,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.800884955752213e-06,
+      "loss": 0.7693,
+      "step": 1880
+    },
+    {
+      "epoch": 2.0807522123893807,
+      "grad_norm": 0.119140625,
+      "learning_rate": 4.7981194690265495e-06,
+      "loss": 0.7598,
+      "step": 1881
+    },
+    {
+      "epoch": 2.081858407079646,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.795353982300886e-06,
+      "loss": 0.8005,
+      "step": 1882
+    },
+    {
+      "epoch": 2.0829646017699117,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.792588495575222e-06,
+      "loss": 0.7557,
+      "step": 1883
+    },
+    {
+      "epoch": 2.084070796460177,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.789823008849558e-06,
+      "loss": 0.796,
+      "step": 1884
+    },
+    {
+      "epoch": 2.0851769911504423,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.787057522123894e-06,
+      "loss": 0.7727,
+      "step": 1885
+    },
+    {
+      "epoch": 2.086283185840708,
+      "grad_norm": 0.130859375,
+      "learning_rate": 4.78429203539823e-06,
+      "loss": 0.8011,
+      "step": 1886
+    },
+    {
+      "epoch": 2.0873893805309733,
+      "grad_norm": 0.1015625,
+      "learning_rate": 4.781526548672566e-06,
+      "loss": 0.7165,
+      "step": 1887
+    },
+    {
+      "epoch": 2.088495575221239,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.778761061946903e-06,
+      "loss": 0.7757,
+      "step": 1888
+    },
+    {
+      "epoch": 2.0896017699115044,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.775995575221239e-06,
+      "loss": 0.7542,
+      "step": 1889
+    },
+    {
+      "epoch": 2.09070796460177,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.7732300884955754e-06,
+      "loss": 0.7627,
+      "step": 1890
+    },
+    {
+      "epoch": 2.0918141592920354,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.7704646017699116e-06,
+      "loss": 0.8335,
+      "step": 1891
+    },
+    {
+      "epoch": 2.0929203539823007,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.767699115044248e-06,
+      "loss": 0.7735,
+      "step": 1892
+    },
+    {
+      "epoch": 2.0940265486725664,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.764933628318585e-06,
+      "loss": 0.7839,
+      "step": 1893
+    },
+    {
+      "epoch": 2.0951327433628317,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 4.762168141592921e-06,
+      "loss": 0.7094,
+      "step": 1894
+    },
+    {
+      "epoch": 2.0962389380530975,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.759402654867257e-06,
+      "loss": 0.7571,
+      "step": 1895
+    },
+    {
+      "epoch": 2.0973451327433628,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.756637168141594e-06,
+      "loss": 0.7712,
+      "step": 1896
+    },
+    {
+      "epoch": 2.0984513274336285,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.75387168141593e-06,
+      "loss": 0.7228,
+      "step": 1897
+    },
+    {
+      "epoch": 2.099557522123894,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.751106194690266e-06,
+      "loss": 0.7637,
+      "step": 1898
+    },
+    {
+      "epoch": 2.100663716814159,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.748340707964602e-06,
+      "loss": 0.7724,
+      "step": 1899
+    },
+    {
+      "epoch": 2.101769911504425,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.745575221238938e-06,
+      "loss": 0.7791,
+      "step": 1900
+    },
+    {
+      "epoch": 2.10287610619469,
+      "grad_norm": 0.119140625,
+      "learning_rate": 4.7428097345132745e-06,
+      "loss": 0.8164,
+      "step": 1901
+    },
+    {
+      "epoch": 2.103982300884956,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.740044247787611e-06,
+      "loss": 0.7508,
+      "step": 1902
+    },
+    {
+      "epoch": 2.105088495575221,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 4.737278761061948e-06,
+      "loss": 0.8579,
+      "step": 1903
+    },
+    {
+      "epoch": 2.106194690265487,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.734513274336284e-06,
+      "loss": 0.741,
+      "step": 1904
+    },
+    {
+      "epoch": 2.107300884955752,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.73174778761062e-06,
+      "loss": 0.785,
+      "step": 1905
+    },
+    {
+      "epoch": 2.1084070796460175,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 4.728982300884956e-06,
+      "loss": 0.8063,
+      "step": 1906
+    },
+    {
+      "epoch": 2.109513274336283,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.726216814159292e-06,
+      "loss": 0.8016,
+      "step": 1907
+    },
+    {
+      "epoch": 2.1106194690265485,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.723451327433628e-06,
+      "loss": 0.7909,
+      "step": 1908
+    },
+    {
+      "epoch": 2.1117256637168142,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.720685840707964e-06,
+      "loss": 0.7564,
+      "step": 1909
+    },
+    {
+      "epoch": 2.1128318584070795,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 4.717920353982301e-06,
+      "loss": 0.7536,
+      "step": 1910
+    },
+    {
+      "epoch": 2.1139380530973453,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.7151548672566375e-06,
+      "loss": 0.7689,
+      "step": 1911
+    },
+    {
+      "epoch": 2.1150442477876106,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 4.712389380530974e-06,
+      "loss": 0.7882,
+      "step": 1912
+    },
+    {
+      "epoch": 2.1161504424778763,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 4.70962389380531e-06,
+      "loss": 0.7441,
+      "step": 1913
+    },
+    {
+      "epoch": 2.1172566371681416,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.706858407079647e-06,
+      "loss": 0.7633,
+      "step": 1914
+    },
+    {
+      "epoch": 2.118362831858407,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 4.704092920353983e-06,
+      "loss": 0.7686,
+      "step": 1915
+    },
+    {
+      "epoch": 2.1194690265486726,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.701327433628319e-06,
+      "loss": 0.7717,
+      "step": 1916
+    },
+    {
+      "epoch": 2.120575221238938,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.698561946902655e-06,
+      "loss": 0.7931,
+      "step": 1917
+    },
+    {
+      "epoch": 2.1216814159292037,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.695796460176992e-06,
+      "loss": 0.7794,
+      "step": 1918
+    },
+    {
+      "epoch": 2.122787610619469,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.693030973451328e-06,
+      "loss": 0.8011,
+      "step": 1919
+    },
+    {
+      "epoch": 2.1238938053097347,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.690265486725664e-06,
+      "loss": 0.8103,
+      "step": 1920
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 4.6875000000000004e-06,
+      "loss": 0.7779,
+      "step": 1921
+    },
+    {
+      "epoch": 2.1261061946902653,
+      "grad_norm": 0.1171875,
+      "learning_rate": 4.6847345132743366e-06,
+      "loss": 0.7945,
+      "step": 1922
+    },
+    {
+      "epoch": 2.127212389380531,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.681969026548673e-06,
+      "loss": 0.7627,
+      "step": 1923
+    },
+    {
+      "epoch": 2.1283185840707963,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.679203539823009e-06,
+      "loss": 0.7554,
+      "step": 1924
+    },
+    {
+      "epoch": 2.129424778761062,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.676438053097346e-06,
+      "loss": 0.7083,
+      "step": 1925
+    },
+    {
+      "epoch": 2.1305309734513274,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 4.673672566371682e-06,
+      "loss": 0.8026,
+      "step": 1926
+    },
+    {
+      "epoch": 2.131637168141593,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.670907079646018e-06,
+      "loss": 0.7678,
+      "step": 1927
+    },
+    {
+      "epoch": 2.1327433628318584,
+      "grad_norm": 0.1171875,
+      "learning_rate": 4.668141592920354e-06,
+      "loss": 0.785,
+      "step": 1928
+    },
+    {
+      "epoch": 2.133849557522124,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.66537610619469e-06,
+      "loss": 0.7682,
+      "step": 1929
+    },
+    {
+      "epoch": 2.1349557522123894,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.662610619469026e-06,
+      "loss": 0.7742,
+      "step": 1930
+    },
+    {
+      "epoch": 2.1360619469026547,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 4.659845132743363e-06,
+      "loss": 0.7559,
+      "step": 1931
+    },
+    {
+      "epoch": 2.1371681415929205,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 4.6570796460176995e-06,
+      "loss": 0.7707,
+      "step": 1932
+    },
+    {
+      "epoch": 2.1382743362831858,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.654314159292036e-06,
+      "loss": 0.782,
+      "step": 1933
+    },
+    {
+      "epoch": 2.1393805309734515,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.651548672566373e-06,
+      "loss": 0.8201,
+      "step": 1934
+    },
+    {
+      "epoch": 2.140486725663717,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.648783185840709e-06,
+      "loss": 0.7498,
+      "step": 1935
+    },
+    {
+      "epoch": 2.1415929203539825,
+      "grad_norm": 0.09228515625,
+      "learning_rate": 4.646017699115045e-06,
+      "loss": 0.7004,
+      "step": 1936
+    },
+    {
+      "epoch": 2.142699115044248,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 4.643252212389381e-06,
+      "loss": 0.7553,
+      "step": 1937
+    },
+    {
+      "epoch": 2.143805309734513,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.640486725663717e-06,
+      "loss": 0.8024,
+      "step": 1938
+    },
+    {
+      "epoch": 2.144911504424779,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 4.637721238938053e-06,
+      "loss": 0.8008,
+      "step": 1939
+    },
+    {
+      "epoch": 2.146017699115044,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.63495575221239e-06,
+      "loss": 0.7655,
+      "step": 1940
+    },
+    {
+      "epoch": 2.14712389380531,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.632190265486726e-06,
+      "loss": 0.7681,
+      "step": 1941
+    },
+    {
+      "epoch": 2.148230088495575,
+      "grad_norm": 0.125,
+      "learning_rate": 4.6294247787610625e-06,
+      "loss": 0.8482,
+      "step": 1942
+    },
+    {
+      "epoch": 2.149336283185841,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 4.626659292035399e-06,
+      "loss": 0.7423,
+      "step": 1943
+    },
+    {
+      "epoch": 2.150442477876106,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.623893805309735e-06,
+      "loss": 0.765,
+      "step": 1944
+    },
+    {
+      "epoch": 2.1515486725663715,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 4.621128318584071e-06,
+      "loss": 0.7763,
+      "step": 1945
+    },
+    {
+      "epoch": 2.1526548672566372,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 4.618362831858407e-06,
+      "loss": 0.7192,
+      "step": 1946
+    },
+    {
+      "epoch": 2.1537610619469025,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.615597345132743e-06,
+      "loss": 0.7764,
+      "step": 1947
+    },
+    {
+      "epoch": 2.1548672566371683,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.61283185840708e-06,
+      "loss": 0.7815,
+      "step": 1948
+    },
+    {
+      "epoch": 2.1559734513274336,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.610066371681416e-06,
+      "loss": 0.7518,
+      "step": 1949
+    },
+    {
+      "epoch": 2.1570796460176993,
+      "grad_norm": 0.115234375,
+      "learning_rate": 4.607300884955752e-06,
+      "loss": 0.7722,
+      "step": 1950
+    },
+    {
+      "epoch": 2.1581858407079646,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.604535398230089e-06,
+      "loss": 0.7735,
+      "step": 1951
+    },
+    {
+      "epoch": 2.15929203539823,
+      "grad_norm": 0.103515625,
+      "learning_rate": 4.6017699115044254e-06,
+      "loss": 0.7367,
+      "step": 1952
+    },
+    {
+      "epoch": 2.1603982300884956,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 4.5990044247787616e-06,
+      "loss": 0.762,
+      "step": 1953
+    },
+    {
+      "epoch": 2.161504424778761,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.596238938053098e-06,
+      "loss": 0.7805,
+      "step": 1954
+    },
+    {
+      "epoch": 2.1626106194690267,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 4.593473451327435e-06,
+      "loss": 0.7555,
+      "step": 1955
+    },
+    {
+      "epoch": 2.163716814159292,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.590707964601771e-06,
+      "loss": 0.7772,
+      "step": 1956
+    },
+    {
+      "epoch": 2.1648230088495577,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 4.587942477876107e-06,
+      "loss": 0.6967,
+      "step": 1957
+    },
+    {
+      "epoch": 2.165929203539823,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.585176991150443e-06,
+      "loss": 0.7812,
+      "step": 1958
+    },
+    {
+      "epoch": 2.1670353982300883,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 4.582411504424779e-06,
+      "loss": 0.7716,
+      "step": 1959
+    },
+    {
+      "epoch": 2.168141592920354,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.579646017699115e-06,
+      "loss": 0.8072,
+      "step": 1960
+    },
+    {
+      "epoch": 2.1692477876106193,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 4.576880530973451e-06,
+      "loss": 0.8168,
+      "step": 1961
+    },
+    {
+      "epoch": 2.170353982300885,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.5741150442477875e-06,
+      "loss": 0.7819,
+      "step": 1962
+    },
+    {
+      "epoch": 2.1714601769911503,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.5713495575221245e-06,
+      "loss": 0.7469,
+      "step": 1963
+    },
+    {
+      "epoch": 2.172566371681416,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.568584070796461e-06,
+      "loss": 0.7749,
+      "step": 1964
+    },
+    {
+      "epoch": 2.1736725663716814,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 4.565818584070797e-06,
+      "loss": 0.7678,
+      "step": 1965
+    },
+    {
+      "epoch": 2.1747787610619467,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.563053097345133e-06,
+      "loss": 0.7462,
+      "step": 1966
+    },
+    {
+      "epoch": 2.1758849557522124,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.560287610619469e-06,
+      "loss": 0.7856,
+      "step": 1967
+    },
+    {
+      "epoch": 2.1769911504424777,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.557522123893805e-06,
+      "loss": 0.781,
+      "step": 1968
+    },
+    {
+      "epoch": 2.1780973451327434,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 4.554756637168142e-06,
+      "loss": 0.782,
+      "step": 1969
+    },
+    {
+      "epoch": 2.1792035398230087,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.551991150442478e-06,
+      "loss": 0.7558,
+      "step": 1970
+    },
+    {
+      "epoch": 2.1803097345132745,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.549225663716814e-06,
+      "loss": 0.7557,
+      "step": 1971
+    },
+    {
+      "epoch": 2.1814159292035398,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 4.546460176991151e-06,
+      "loss": 0.7661,
+      "step": 1972
+    },
+    {
+      "epoch": 2.1825221238938055,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 4.5436946902654875e-06,
+      "loss": 0.8595,
+      "step": 1973
+    },
+    {
+      "epoch": 2.183628318584071,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.540929203539824e-06,
+      "loss": 0.7621,
+      "step": 1974
+    },
+    {
+      "epoch": 2.184734513274336,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.53816371681416e-06,
+      "loss": 0.7753,
+      "step": 1975
+    },
+    {
+      "epoch": 2.185840707964602,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.535398230088496e-06,
+      "loss": 0.7684,
+      "step": 1976
+    },
+    {
+      "epoch": 2.186946902654867,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.532632743362832e-06,
+      "loss": 0.7306,
+      "step": 1977
+    },
+    {
+      "epoch": 2.188053097345133,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.529867256637169e-06,
+      "loss": 0.7549,
+      "step": 1978
+    },
+    {
+      "epoch": 2.189159292035398,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.527101769911505e-06,
+      "loss": 0.7767,
+      "step": 1979
+    },
+    {
+      "epoch": 2.190265486725664,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.524336283185841e-06,
+      "loss": 0.7612,
+      "step": 1980
+    },
+    {
+      "epoch": 2.191371681415929,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 4.521570796460177e-06,
+      "loss": 0.7704,
+      "step": 1981
+    },
+    {
+      "epoch": 2.1924778761061945,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 4.5188053097345134e-06,
+      "loss": 0.7826,
+      "step": 1982
+    },
+    {
+      "epoch": 2.1935840707964602,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.5160398230088496e-06,
+      "loss": 0.77,
+      "step": 1983
+    },
+    {
+      "epoch": 2.1946902654867255,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.513274336283186e-06,
+      "loss": 0.7158,
+      "step": 1984
+    },
+    {
+      "epoch": 2.1957964601769913,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.510508849557522e-06,
+      "loss": 0.7333,
+      "step": 1985
+    },
+    {
+      "epoch": 2.1969026548672566,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.507743362831859e-06,
+      "loss": 0.7719,
+      "step": 1986
+    },
+    {
+      "epoch": 2.1980088495575223,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.504977876106195e-06,
+      "loss": 0.7861,
+      "step": 1987
+    },
+    {
+      "epoch": 2.1991150442477876,
+      "grad_norm": 0.1328125,
+      "learning_rate": 4.502212389380531e-06,
+      "loss": 0.7752,
+      "step": 1988
+    },
+    {
+      "epoch": 2.2002212389380533,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.499446902654868e-06,
+      "loss": 0.7643,
+      "step": 1989
+    },
+    {
+      "epoch": 2.2013274336283186,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.496681415929204e-06,
+      "loss": 0.737,
+      "step": 1990
+    },
+    {
+      "epoch": 2.202433628318584,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 4.49391592920354e-06,
+      "loss": 0.7514,
+      "step": 1991
+    },
+    {
+      "epoch": 2.2035398230088497,
+      "grad_norm": 0.123046875,
+      "learning_rate": 4.491150442477876e-06,
+      "loss": 0.8073,
+      "step": 1992
+    },
+    {
+      "epoch": 2.204646017699115,
+      "grad_norm": 0.1328125,
+      "learning_rate": 4.488384955752213e-06,
+      "loss": 0.8034,
+      "step": 1993
+    },
+    {
+      "epoch": 2.2057522123893807,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 4.4856194690265495e-06,
+      "loss": 0.7767,
+      "step": 1994
+    },
+    {
+      "epoch": 2.206858407079646,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.482853982300886e-06,
+      "loss": 0.7686,
+      "step": 1995
+    },
+    {
+      "epoch": 2.2079646017699117,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 4.480088495575222e-06,
+      "loss": 0.7836,
+      "step": 1996
+    },
+    {
+      "epoch": 2.209070796460177,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.477323008849558e-06,
+      "loss": 0.7569,
+      "step": 1997
+    },
+    {
+      "epoch": 2.2101769911504423,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.474557522123894e-06,
+      "loss": 0.7561,
+      "step": 1998
+    },
+    {
+      "epoch": 2.211283185840708,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.47179203539823e-06,
+      "loss": 0.7796,
+      "step": 1999
+    },
+    {
+      "epoch": 2.2123893805309733,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 4.469026548672566e-06,
+      "loss": 0.7688,
+      "step": 2000
+    },
+    {
+      "epoch": 2.213495575221239,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.466261061946903e-06,
+      "loss": 0.7861,
+      "step": 2001
+    },
+    {
+      "epoch": 2.2146017699115044,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.463495575221239e-06,
+      "loss": 0.7472,
+      "step": 2002
+    },
+    {
+      "epoch": 2.21570796460177,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 4.4607300884955755e-06,
+      "loss": 0.7388,
+      "step": 2003
+    },
+    {
+      "epoch": 2.2168141592920354,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.457964601769912e-06,
+      "loss": 0.7941,
+      "step": 2004
+    },
+    {
+      "epoch": 2.2179203539823007,
+      "grad_norm": 0.099609375,
+      "learning_rate": 4.455199115044248e-06,
+      "loss": 0.7751,
+      "step": 2005
+    },
+    {
+      "epoch": 2.2190265486725664,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 4.452433628318585e-06,
+      "loss": 0.8149,
+      "step": 2006
+    },
+    {
+      "epoch": 2.2201327433628317,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.449668141592921e-06,
+      "loss": 0.7543,
+      "step": 2007
+    },
+    {
+      "epoch": 2.2212389380530975,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 4.446902654867257e-06,
+      "loss": 0.7625,
+      "step": 2008
+    },
+    {
+      "epoch": 2.2223451327433628,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.444137168141593e-06,
+      "loss": 0.7727,
+      "step": 2009
+    },
+    {
+      "epoch": 2.2234513274336285,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.44137168141593e-06,
+      "loss": 0.7738,
+      "step": 2010
+    },
+    {
+      "epoch": 2.224557522123894,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.438606194690266e-06,
+      "loss": 0.7872,
+      "step": 2011
+    },
+    {
+      "epoch": 2.225663716814159,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 4.435840707964602e-06,
+      "loss": 0.782,
+      "step": 2012
+    },
+    {
+      "epoch": 2.226769911504425,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.4330752212389384e-06,
+      "loss": 0.8091,
+      "step": 2013
+    },
+    {
+      "epoch": 2.22787610619469,
+      "grad_norm": 0.123046875,
+      "learning_rate": 4.4303097345132746e-06,
+      "loss": 0.746,
+      "step": 2014
+    },
+    {
+      "epoch": 2.228982300884956,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.427544247787611e-06,
+      "loss": 0.791,
+      "step": 2015
+    },
+    {
+      "epoch": 2.230088495575221,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 4.424778761061948e-06,
+      "loss": 0.7684,
+      "step": 2016
+    },
+    {
+      "epoch": 2.231194690265487,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.422013274336284e-06,
+      "loss": 0.781,
+      "step": 2017
+    },
+    {
+      "epoch": 2.232300884955752,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.41924778761062e-06,
+      "loss": 0.7574,
+      "step": 2018
+    },
+    {
+      "epoch": 2.2334070796460175,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 4.416482300884956e-06,
+      "loss": 0.7444,
+      "step": 2019
+    },
+    {
+      "epoch": 2.234513274336283,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.413716814159292e-06,
+      "loss": 0.7728,
+      "step": 2020
+    },
+    {
+      "epoch": 2.2356194690265485,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.410951327433628e-06,
+      "loss": 0.7731,
+      "step": 2021
+    },
+    {
+      "epoch": 2.2367256637168142,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 4.408185840707964e-06,
+      "loss": 0.8071,
+      "step": 2022
+    },
+    {
+      "epoch": 2.2378318584070795,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 4.405420353982301e-06,
+      "loss": 0.7655,
+      "step": 2023
+    },
+    {
+      "epoch": 2.2389380530973453,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.4026548672566375e-06,
+      "loss": 0.7738,
+      "step": 2024
+    },
+    {
+      "epoch": 2.2400442477876106,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.399889380530974e-06,
+      "loss": 0.7674,
+      "step": 2025
+    },
+    {
+      "epoch": 2.241150442477876,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 4.39712389380531e-06,
+      "loss": 0.7415,
+      "step": 2026
+    },
+    {
+      "epoch": 2.2422566371681416,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.394358407079647e-06,
+      "loss": 0.8308,
+      "step": 2027
+    },
+    {
+      "epoch": 2.243362831858407,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 4.391592920353983e-06,
+      "loss": 0.7658,
+      "step": 2028
+    },
+    {
+      "epoch": 2.2444690265486726,
+      "grad_norm": 0.119140625,
+      "learning_rate": 4.388827433628319e-06,
+      "loss": 0.7998,
+      "step": 2029
+    },
+    {
+      "epoch": 2.245575221238938,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.386061946902655e-06,
+      "loss": 0.7467,
+      "step": 2030
+    },
+    {
+      "epoch": 2.2466814159292037,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.383296460176992e-06,
+      "loss": 0.8377,
+      "step": 2031
+    },
+    {
+      "epoch": 2.247787610619469,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 4.380530973451328e-06,
+      "loss": 0.7349,
+      "step": 2032
+    },
+    {
+      "epoch": 2.2488938053097347,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.377765486725664e-06,
+      "loss": 0.7675,
+      "step": 2033
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.7239,
+      "step": 2034
+    },
+    {
+      "epoch": 2.2511061946902653,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 4.372234513274337e-06,
+      "loss": 0.7601,
+      "step": 2035
+    },
+    {
+      "epoch": 2.252212389380531,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 4.369469026548673e-06,
+      "loss": 0.7186,
+      "step": 2036
+    },
+    {
+      "epoch": 2.2533185840707963,
+      "grad_norm": 0.103515625,
+      "learning_rate": 4.366703539823009e-06,
+      "loss": 0.7354,
+      "step": 2037
+    },
+    {
+      "epoch": 2.254424778761062,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 4.363938053097346e-06,
+      "loss": 0.793,
+      "step": 2038
+    },
+    {
+      "epoch": 2.2555309734513274,
+      "grad_norm": 0.119140625,
+      "learning_rate": 4.361172566371682e-06,
+      "loss": 0.7781,
+      "step": 2039
+    },
+    {
+      "epoch": 2.256637168141593,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.358407079646018e-06,
+      "loss": 0.7597,
+      "step": 2040
+    },
+    {
+      "epoch": 2.2577433628318584,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.355641592920354e-06,
+      "loss": 0.7915,
+      "step": 2041
+    },
+    {
+      "epoch": 2.258849557522124,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.35287610619469e-06,
+      "loss": 0.786,
+      "step": 2042
+    },
+    {
+      "epoch": 2.2599557522123894,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 4.3501106194690264e-06,
+      "loss": 0.8299,
+      "step": 2043
+    },
+    {
+      "epoch": 2.2610619469026547,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.347345132743363e-06,
+      "loss": 0.778,
+      "step": 2044
+    },
+    {
+      "epoch": 2.2621681415929205,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 4.3445796460176995e-06,
+      "loss": 0.8183,
+      "step": 2045
+    },
+    {
+      "epoch": 2.2632743362831858,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.341814159292036e-06,
+      "loss": 0.8094,
+      "step": 2046
+    },
+    {
+      "epoch": 2.2643805309734515,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.339048672566372e-06,
+      "loss": 0.7763,
+      "step": 2047
+    },
+    {
+      "epoch": 2.265486725663717,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.336283185840709e-06,
+      "loss": 0.7934,
+      "step": 2048
+    },
+    {
+      "epoch": 2.2665929203539825,
+      "grad_norm": 0.125,
+      "learning_rate": 4.333517699115045e-06,
+      "loss": 0.7863,
+      "step": 2049
+    },
+    {
+      "epoch": 2.267699115044248,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.330752212389381e-06,
+      "loss": 0.7578,
+      "step": 2050
+    },
+    {
+      "epoch": 2.268805309734513,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.327986725663717e-06,
+      "loss": 0.7653,
+      "step": 2051
+    },
+    {
+      "epoch": 2.269911504424779,
+      "grad_norm": 0.126953125,
+      "learning_rate": 4.325221238938053e-06,
+      "loss": 0.8119,
+      "step": 2052
+    },
+    {
+      "epoch": 2.271017699115044,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.32245575221239e-06,
+      "loss": 0.7836,
+      "step": 2053
+    },
+    {
+      "epoch": 2.27212389380531,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.319690265486726e-06,
+      "loss": 0.7355,
+      "step": 2054
+    },
+    {
+      "epoch": 2.273230088495575,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.3169247787610625e-06,
+      "loss": 0.792,
+      "step": 2055
+    },
+    {
+      "epoch": 2.274336283185841,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.314159292035399e-06,
+      "loss": 0.772,
+      "step": 2056
+    },
+    {
+      "epoch": 2.275442477876106,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.311393805309735e-06,
+      "loss": 0.7733,
+      "step": 2057
+    },
+    {
+      "epoch": 2.2765486725663715,
+      "grad_norm": 0.099609375,
+      "learning_rate": 4.308628318584071e-06,
+      "loss": 0.7409,
+      "step": 2058
+    },
+    {
+      "epoch": 2.2776548672566372,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 4.305862831858407e-06,
+      "loss": 0.7212,
+      "step": 2059
+    },
+    {
+      "epoch": 2.2787610619469025,
+      "grad_norm": 0.115234375,
+      "learning_rate": 4.303097345132743e-06,
+      "loss": 0.8494,
+      "step": 2060
+    },
+    {
+      "epoch": 2.2798672566371683,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 4.30033185840708e-06,
+      "loss": 0.7942,
+      "step": 2061
+    },
+    {
+      "epoch": 2.2809734513274336,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 4.297566371681416e-06,
+      "loss": 0.7371,
+      "step": 2062
+    },
+    {
+      "epoch": 2.2820796460176993,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 4.294800884955752e-06,
+      "loss": 0.7888,
+      "step": 2063
+    },
+    {
+      "epoch": 2.2831858407079646,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 4.2920353982300885e-06,
+      "loss": 0.7833,
+      "step": 2064
+    },
+    {
+      "epoch": 2.28429203539823,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.2892699115044255e-06,
+      "loss": 0.7913,
+      "step": 2065
+    },
+    {
+      "epoch": 2.2853982300884956,
+      "grad_norm": 0.115234375,
+      "learning_rate": 4.286504424778762e-06,
+      "loss": 0.7786,
+      "step": 2066
+    },
+    {
+      "epoch": 2.286504424778761,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 4.283738938053098e-06,
+      "loss": 0.8374,
+      "step": 2067
+    },
+    {
+      "epoch": 2.2876106194690267,
+      "grad_norm": 0.130859375,
+      "learning_rate": 4.280973451327434e-06,
+      "loss": 0.754,
+      "step": 2068
+    },
+    {
+      "epoch": 2.288716814159292,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.278207964601771e-06,
+      "loss": 0.7698,
+      "step": 2069
+    },
+    {
+      "epoch": 2.2898230088495577,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.275442477876107e-06,
+      "loss": 0.776,
+      "step": 2070
+    },
+    {
+      "epoch": 2.290929203539823,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 4.272676991150443e-06,
+      "loss": 0.7286,
+      "step": 2071
+    },
+    {
+      "epoch": 2.2920353982300883,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.269911504424779e-06,
+      "loss": 0.7595,
+      "step": 2072
+    },
+    {
+      "epoch": 2.293141592920354,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.267146017699115e-06,
+      "loss": 0.7683,
+      "step": 2073
+    },
+    {
+      "epoch": 2.2942477876106193,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 4.2643805309734514e-06,
+      "loss": 0.7527,
+      "step": 2074
+    },
+    {
+      "epoch": 2.295353982300885,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.2616150442477876e-06,
+      "loss": 0.7537,
+      "step": 2075
+    },
+    {
+      "epoch": 2.2964601769911503,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.2588495575221245e-06,
+      "loss": 0.7695,
+      "step": 2076
+    },
+    {
+      "epoch": 2.297566371681416,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 4.256084070796461e-06,
+      "loss": 0.763,
+      "step": 2077
+    },
+    {
+      "epoch": 2.2986725663716814,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 4.253318584070797e-06,
+      "loss": 0.7412,
+      "step": 2078
+    },
+    {
+      "epoch": 2.2997787610619467,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 4.250553097345133e-06,
+      "loss": 0.6837,
+      "step": 2079
+    },
+    {
+      "epoch": 2.3008849557522124,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 4.247787610619469e-06,
+      "loss": 0.7184,
+      "step": 2080
+    },
+    {
+      "epoch": 2.3019911504424777,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.245022123893805e-06,
+      "loss": 0.7952,
+      "step": 2081
+    },
+    {
+      "epoch": 2.3030973451327434,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.242256637168142e-06,
+      "loss": 0.7417,
+      "step": 2082
+    },
+    {
+      "epoch": 2.3042035398230087,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.239491150442478e-06,
+      "loss": 0.7988,
+      "step": 2083
+    },
+    {
+      "epoch": 2.3053097345132745,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 4.236725663716814e-06,
+      "loss": 0.7763,
+      "step": 2084
+    },
+    {
+      "epoch": 2.3064159292035398,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.2339601769911505e-06,
+      "loss": 0.7539,
+      "step": 2085
+    },
+    {
+      "epoch": 2.307522123893805,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.2311946902654875e-06,
+      "loss": 0.7693,
+      "step": 2086
+    },
+    {
+      "epoch": 2.308628318584071,
+      "grad_norm": 0.1171875,
+      "learning_rate": 4.228429203539824e-06,
+      "loss": 0.8046,
+      "step": 2087
+    },
+    {
+      "epoch": 2.309734513274336,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 4.22566371681416e-06,
+      "loss": 0.766,
+      "step": 2088
+    },
+    {
+      "epoch": 2.310840707964602,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.222898230088496e-06,
+      "loss": 0.8059,
+      "step": 2089
+    },
+    {
+      "epoch": 2.311946902654867,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 4.220132743362832e-06,
+      "loss": 0.7334,
+      "step": 2090
+    },
+    {
+      "epoch": 2.313053097345133,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.217367256637169e-06,
+      "loss": 0.801,
+      "step": 2091
+    },
+    {
+      "epoch": 2.314159292035398,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 4.214601769911505e-06,
+      "loss": 0.7761,
+      "step": 2092
+    },
+    {
+      "epoch": 2.315265486725664,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.211836283185841e-06,
+      "loss": 0.7543,
+      "step": 2093
+    },
+    {
+      "epoch": 2.316371681415929,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.209070796460177e-06,
+      "loss": 0.7484,
+      "step": 2094
+    },
+    {
+      "epoch": 2.317477876106195,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 4.2063053097345135e-06,
+      "loss": 0.819,
+      "step": 2095
+    },
+    {
+      "epoch": 2.3185840707964602,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.20353982300885e-06,
+      "loss": 0.7714,
+      "step": 2096
+    },
+    {
+      "epoch": 2.3196902654867255,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 4.200774336283186e-06,
+      "loss": 0.7243,
+      "step": 2097
+    },
+    {
+      "epoch": 2.3207964601769913,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.198008849557522e-06,
+      "loss": 0.8015,
+      "step": 2098
+    },
+    {
+      "epoch": 2.3219026548672566,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.195243362831859e-06,
+      "loss": 0.7713,
+      "step": 2099
+    },
+    {
+      "epoch": 2.3230088495575223,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 4.192477876106195e-06,
+      "loss": 0.7632,
+      "step": 2100
+    },
+    {
+      "epoch": 2.3241150442477876,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.189712389380531e-06,
+      "loss": 0.7823,
+      "step": 2101
+    },
+    {
+      "epoch": 2.3252212389380533,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.186946902654867e-06,
+      "loss": 0.7629,
+      "step": 2102
+    },
+    {
+      "epoch": 2.3263274336283186,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 4.184181415929204e-06,
+      "loss": 0.7823,
+      "step": 2103
+    },
+    {
+      "epoch": 2.327433628318584,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.18141592920354e-06,
+      "loss": 0.7994,
+      "step": 2104
+    },
+    {
+      "epoch": 2.3285398230088497,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.178650442477876e-06,
+      "loss": 0.7742,
+      "step": 2105
+    },
+    {
+      "epoch": 2.329646017699115,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 4.175884955752213e-06,
+      "loss": 0.778,
+      "step": 2106
+    },
+    {
+      "epoch": 2.3307522123893807,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.1731194690265495e-06,
+      "loss": 0.7618,
+      "step": 2107
+    },
+    {
+      "epoch": 2.331858407079646,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.170353982300886e-06,
+      "loss": 0.7421,
+      "step": 2108
+    },
+    {
+      "epoch": 2.3329646017699117,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.167588495575222e-06,
+      "loss": 0.8516,
+      "step": 2109
+    },
+    {
+      "epoch": 2.334070796460177,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 4.164823008849558e-06,
+      "loss": 0.7567,
+      "step": 2110
+    },
+    {
+      "epoch": 2.3351769911504423,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.162057522123894e-06,
+      "loss": 0.7531,
+      "step": 2111
+    },
+    {
+      "epoch": 2.336283185840708,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.15929203539823e-06,
+      "loss": 0.7477,
+      "step": 2112
+    },
+    {
+      "epoch": 2.3373893805309733,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.156526548672566e-06,
+      "loss": 0.75,
+      "step": 2113
+    },
+    {
+      "epoch": 2.338495575221239,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 4.153761061946903e-06,
+      "loss": 0.7808,
+      "step": 2114
+    },
+    {
+      "epoch": 2.3396017699115044,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.150995575221239e-06,
+      "loss": 0.7435,
+      "step": 2115
+    },
+    {
+      "epoch": 2.34070796460177,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.1482300884955755e-06,
+      "loss": 0.7744,
+      "step": 2116
+    },
+    {
+      "epoch": 2.3418141592920354,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.145464601769912e-06,
+      "loss": 0.7774,
+      "step": 2117
+    },
+    {
+      "epoch": 2.3429203539823007,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.142699115044248e-06,
+      "loss": 0.7737,
+      "step": 2118
+    },
+    {
+      "epoch": 2.3440265486725664,
+      "grad_norm": 0.1171875,
+      "learning_rate": 4.139933628318584e-06,
+      "loss": 0.7803,
+      "step": 2119
+    },
+    {
+      "epoch": 2.3451327433628317,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.137168141592921e-06,
+      "loss": 0.7661,
+      "step": 2120
+    },
+    {
+      "epoch": 2.3462389380530975,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 4.134402654867257e-06,
+      "loss": 0.7429,
+      "step": 2121
+    },
+    {
+      "epoch": 2.3473451327433628,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.131637168141593e-06,
+      "loss": 0.8049,
+      "step": 2122
+    },
+    {
+      "epoch": 2.3484513274336285,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 4.128871681415929e-06,
+      "loss": 0.7612,
+      "step": 2123
+    },
+    {
+      "epoch": 2.349557522123894,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.126106194690266e-06,
+      "loss": 0.7807,
+      "step": 2124
+    },
+    {
+      "epoch": 2.350663716814159,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.123340707964602e-06,
+      "loss": 0.7756,
+      "step": 2125
+    },
+    {
+      "epoch": 2.351769911504425,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 4.1205752212389385e-06,
+      "loss": 0.7361,
+      "step": 2126
+    },
+    {
+      "epoch": 2.35287610619469,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.117809734513275e-06,
+      "loss": 0.8198,
+      "step": 2127
+    },
+    {
+      "epoch": 2.353982300884956,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.115044247787611e-06,
+      "loss": 0.7218,
+      "step": 2128
+    },
+    {
+      "epoch": 2.355088495575221,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.112278761061948e-06,
+      "loss": 0.7758,
+      "step": 2129
+    },
+    {
+      "epoch": 2.356194690265487,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.109513274336284e-06,
+      "loss": 0.8069,
+      "step": 2130
+    },
+    {
+      "epoch": 2.357300884955752,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.10674778761062e-06,
+      "loss": 0.751,
+      "step": 2131
+    },
+    {
+      "epoch": 2.3584070796460175,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.103982300884956e-06,
+      "loss": 0.7745,
+      "step": 2132
+    },
+    {
+      "epoch": 2.359513274336283,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 4.101216814159292e-06,
+      "loss": 0.7856,
+      "step": 2133
+    },
+    {
+      "epoch": 2.3606194690265485,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 4.098451327433628e-06,
+      "loss": 0.8163,
+      "step": 2134
+    },
+    {
+      "epoch": 2.3617256637168142,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 4.0956858407079644e-06,
+      "loss": 0.7529,
+      "step": 2135
+    },
+    {
+      "epoch": 2.3628318584070795,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.092920353982301e-06,
+      "loss": 0.7598,
+      "step": 2136
+    },
+    {
+      "epoch": 2.3639380530973453,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.0901548672566375e-06,
+      "loss": 0.7892,
+      "step": 2137
+    },
+    {
+      "epoch": 2.3650442477876106,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 4.087389380530974e-06,
+      "loss": 0.7952,
+      "step": 2138
+    },
+    {
+      "epoch": 2.366150442477876,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.08462389380531e-06,
+      "loss": 0.7613,
+      "step": 2139
+    },
+    {
+      "epoch": 2.3672566371681416,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.081858407079646e-06,
+      "loss": 0.7926,
+      "step": 2140
+    },
+    {
+      "epoch": 2.368362831858407,
+      "grad_norm": 0.103515625,
+      "learning_rate": 4.079092920353983e-06,
+      "loss": 0.7445,
+      "step": 2141
+    },
+    {
+      "epoch": 2.3694690265486726,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.076327433628319e-06,
+      "loss": 0.784,
+      "step": 2142
+    },
+    {
+      "epoch": 2.370575221238938,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 4.073561946902655e-06,
+      "loss": 0.7774,
+      "step": 2143
+    },
+    {
+      "epoch": 2.3716814159292037,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.070796460176992e-06,
+      "loss": 0.7834,
+      "step": 2144
+    },
+    {
+      "epoch": 2.372787610619469,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.068030973451328e-06,
+      "loss": 0.7923,
+      "step": 2145
+    },
+    {
+      "epoch": 2.3738938053097347,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 4.065265486725664e-06,
+      "loss": 0.7481,
+      "step": 2146
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.0625000000000005e-06,
+      "loss": 0.7544,
+      "step": 2147
+    },
+    {
+      "epoch": 2.3761061946902653,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.059734513274337e-06,
+      "loss": 0.7754,
+      "step": 2148
+    },
+    {
+      "epoch": 2.377212389380531,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.056969026548673e-06,
+      "loss": 0.7807,
+      "step": 2149
+    },
+    {
+      "epoch": 2.3783185840707963,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 4.054203539823009e-06,
+      "loss": 0.7597,
+      "step": 2150
+    },
+    {
+      "epoch": 2.379424778761062,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.051438053097346e-06,
+      "loss": 0.7413,
+      "step": 2151
+    },
+    {
+      "epoch": 2.3805309734513274,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 4.048672566371682e-06,
+      "loss": 0.7698,
+      "step": 2152
+    },
+    {
+      "epoch": 2.381637168141593,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.045907079646018e-06,
+      "loss": 0.7583,
+      "step": 2153
+    },
+    {
+      "epoch": 2.3827433628318584,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 4.043141592920354e-06,
+      "loss": 0.7902,
+      "step": 2154
+    },
+    {
+      "epoch": 2.383849557522124,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 4.04037610619469e-06,
+      "loss": 0.7444,
+      "step": 2155
+    },
+    {
+      "epoch": 2.3849557522123894,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.0376106194690265e-06,
+      "loss": 0.7725,
+      "step": 2156
+    },
+    {
+      "epoch": 2.3860619469026547,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 4.034845132743363e-06,
+      "loss": 0.7878,
+      "step": 2157
+    },
+    {
+      "epoch": 2.3871681415929205,
+      "grad_norm": 0.09765625,
+      "learning_rate": 4.0320796460176996e-06,
+      "loss": 0.7435,
+      "step": 2158
+    },
+    {
+      "epoch": 2.3882743362831858,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.029314159292036e-06,
+      "loss": 0.8202,
+      "step": 2159
+    },
+    {
+      "epoch": 2.3893805309734515,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.026548672566372e-06,
+      "loss": 0.7327,
+      "step": 2160
+    },
+    {
+      "epoch": 2.390486725663717,
+      "grad_norm": 0.119140625,
+      "learning_rate": 4.023783185840708e-06,
+      "loss": 0.7678,
+      "step": 2161
+    },
+    {
+      "epoch": 2.3915929203539825,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 4.021017699115045e-06,
+      "loss": 0.7748,
+      "step": 2162
+    },
+    {
+      "epoch": 2.392699115044248,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.018252212389381e-06,
+      "loss": 0.7778,
+      "step": 2163
+    },
+    {
+      "epoch": 2.393805309734513,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.015486725663717e-06,
+      "loss": 0.7402,
+      "step": 2164
+    },
+    {
+      "epoch": 2.394911504424779,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 4.012721238938053e-06,
+      "loss": 0.7566,
+      "step": 2165
+    },
+    {
+      "epoch": 2.396017699115044,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 4.00995575221239e-06,
+      "loss": 0.7851,
+      "step": 2166
+    },
+    {
+      "epoch": 2.39712389380531,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.007190265486726e-06,
+      "loss": 0.7562,
+      "step": 2167
+    },
+    {
+      "epoch": 2.398230088495575,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.0044247787610625e-06,
+      "loss": 0.7639,
+      "step": 2168
+    },
+    {
+      "epoch": 2.399336283185841,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.001659292035399e-06,
+      "loss": 0.7501,
+      "step": 2169
+    },
+    {
+      "epoch": 2.400442477876106,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.998893805309735e-06,
+      "loss": 0.7584,
+      "step": 2170
+    },
+    {
+      "epoch": 2.4015486725663715,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.996128318584071e-06,
+      "loss": 0.7279,
+      "step": 2171
+    },
+    {
+      "epoch": 2.4026548672566372,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.993362831858407e-06,
+      "loss": 0.7815,
+      "step": 2172
+    },
+    {
+      "epoch": 2.4037610619469025,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 3.990597345132743e-06,
+      "loss": 0.7499,
+      "step": 2173
+    },
+    {
+      "epoch": 2.4048672566371683,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.98783185840708e-06,
+      "loss": 0.7765,
+      "step": 2174
+    },
+    {
+      "epoch": 2.4059734513274336,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.985066371681416e-06,
+      "loss": 0.804,
+      "step": 2175
+    },
+    {
+      "epoch": 2.4070796460176993,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.982300884955752e-06,
+      "loss": 0.7739,
+      "step": 2176
+    },
+    {
+      "epoch": 2.4081858407079646,
+      "grad_norm": 0.103515625,
+      "learning_rate": 3.9795353982300885e-06,
+      "loss": 0.7588,
+      "step": 2177
+    },
+    {
+      "epoch": 2.40929203539823,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 3.976769911504425e-06,
+      "loss": 0.7367,
+      "step": 2178
+    },
+    {
+      "epoch": 2.4103982300884956,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.974004424778762e-06,
+      "loss": 0.7083,
+      "step": 2179
+    },
+    {
+      "epoch": 2.411504424778761,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.971238938053098e-06,
+      "loss": 0.7451,
+      "step": 2180
+    },
+    {
+      "epoch": 2.4126106194690267,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 3.968473451327434e-06,
+      "loss": 0.7788,
+      "step": 2181
+    },
+    {
+      "epoch": 2.413716814159292,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.965707964601771e-06,
+      "loss": 0.7641,
+      "step": 2182
+    },
+    {
+      "epoch": 2.4148230088495577,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.962942477876107e-06,
+      "loss": 0.7446,
+      "step": 2183
+    },
+    {
+      "epoch": 2.415929203539823,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.960176991150443e-06,
+      "loss": 0.7799,
+      "step": 2184
+    },
+    {
+      "epoch": 2.4170353982300883,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.957411504424779e-06,
+      "loss": 0.7336,
+      "step": 2185
+    },
+    {
+      "epoch": 2.418141592920354,
+      "grad_norm": 0.12109375,
+      "learning_rate": 3.954646017699115e-06,
+      "loss": 0.7699,
+      "step": 2186
+    },
+    {
+      "epoch": 2.4192477876106193,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.9518805309734515e-06,
+      "loss": 0.7411,
+      "step": 2187
+    },
+    {
+      "epoch": 2.420353982300885,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 3.949115044247788e-06,
+      "loss": 0.7705,
+      "step": 2188
+    },
+    {
+      "epoch": 2.4214601769911503,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.9463495575221246e-06,
+      "loss": 0.7487,
+      "step": 2189
+    },
+    {
+      "epoch": 2.422566371681416,
+      "grad_norm": 0.107421875,
+      "learning_rate": 3.943584070796461e-06,
+      "loss": 0.7946,
+      "step": 2190
+    },
+    {
+      "epoch": 2.4236725663716814,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.940818584070797e-06,
+      "loss": 0.7598,
+      "step": 2191
+    },
+    {
+      "epoch": 2.4247787610619467,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.938053097345133e-06,
+      "loss": 0.791,
+      "step": 2192
+    },
+    {
+      "epoch": 2.4258849557522124,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.935287610619469e-06,
+      "loss": 0.789,
+      "step": 2193
+    },
+    {
+      "epoch": 2.4269911504424777,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.932522123893805e-06,
+      "loss": 0.7758,
+      "step": 2194
+    },
+    {
+      "epoch": 2.4280973451327434,
+      "grad_norm": 0.126953125,
+      "learning_rate": 3.929756637168141e-06,
+      "loss": 0.8354,
+      "step": 2195
+    },
+    {
+      "epoch": 2.4292035398230087,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.926991150442478e-06,
+      "loss": 0.7728,
+      "step": 2196
+    },
+    {
+      "epoch": 2.4303097345132745,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 3.924225663716814e-06,
+      "loss": 0.7867,
+      "step": 2197
+    },
+    {
+      "epoch": 2.4314159292035398,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.9214601769911505e-06,
+      "loss": 0.747,
+      "step": 2198
+    },
+    {
+      "epoch": 2.432522123893805,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.9186946902654875e-06,
+      "loss": 0.7275,
+      "step": 2199
+    },
+    {
+      "epoch": 2.433628318584071,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.915929203539824e-06,
+      "loss": 0.7894,
+      "step": 2200
+    },
+    {
+      "epoch": 2.434734513274336,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.91316371681416e-06,
+      "loss": 0.7704,
+      "step": 2201
+    },
+    {
+      "epoch": 2.435840707964602,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.910398230088496e-06,
+      "loss": 0.8445,
+      "step": 2202
+    },
+    {
+      "epoch": 2.436946902654867,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.907632743362832e-06,
+      "loss": 0.8033,
+      "step": 2203
+    },
+    {
+      "epoch": 2.438053097345133,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.904867256637169e-06,
+      "loss": 0.7817,
+      "step": 2204
+    },
+    {
+      "epoch": 2.439159292035398,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.902101769911505e-06,
+      "loss": 0.8007,
+      "step": 2205
+    },
+    {
+      "epoch": 2.440265486725664,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.899336283185841e-06,
+      "loss": 0.7688,
+      "step": 2206
+    },
+    {
+      "epoch": 2.441371681415929,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 3.896570796460177e-06,
+      "loss": 0.7746,
+      "step": 2207
+    },
+    {
+      "epoch": 2.442477876106195,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 3.8938053097345135e-06,
+      "loss": 0.7497,
+      "step": 2208
+    },
+    {
+      "epoch": 2.4435840707964602,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.89103982300885e-06,
+      "loss": 0.7858,
+      "step": 2209
+    },
+    {
+      "epoch": 2.4446902654867255,
+      "grad_norm": 0.107421875,
+      "learning_rate": 3.888274336283186e-06,
+      "loss": 0.778,
+      "step": 2210
+    },
+    {
+      "epoch": 2.4457964601769913,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.885508849557522e-06,
+      "loss": 0.7778,
+      "step": 2211
+    },
+    {
+      "epoch": 2.4469026548672566,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.882743362831859e-06,
+      "loss": 0.7368,
+      "step": 2212
+    },
+    {
+      "epoch": 2.4480088495575223,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.879977876106195e-06,
+      "loss": 0.775,
+      "step": 2213
+    },
+    {
+      "epoch": 2.4491150442477876,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.877212389380531e-06,
+      "loss": 0.7285,
+      "step": 2214
+    },
+    {
+      "epoch": 2.4502212389380533,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.874446902654867e-06,
+      "loss": 0.8135,
+      "step": 2215
+    },
+    {
+      "epoch": 2.4513274336283186,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 3.871681415929203e-06,
+      "loss": 0.7797,
+      "step": 2216
+    },
+    {
+      "epoch": 2.452433628318584,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.86891592920354e-06,
+      "loss": 0.777,
+      "step": 2217
+    },
+    {
+      "epoch": 2.4535398230088497,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.8661504424778764e-06,
+      "loss": 0.8014,
+      "step": 2218
+    },
+    {
+      "epoch": 2.454646017699115,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.8633849557522126e-06,
+      "loss": 0.7906,
+      "step": 2219
+    },
+    {
+      "epoch": 2.4557522123893807,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.8606194690265495e-06,
+      "loss": 0.7628,
+      "step": 2220
+    },
+    {
+      "epoch": 2.456858407079646,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.857853982300886e-06,
+      "loss": 0.7668,
+      "step": 2221
+    },
+    {
+      "epoch": 2.4579646017699117,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.855088495575222e-06,
+      "loss": 0.7817,
+      "step": 2222
+    },
+    {
+      "epoch": 2.459070796460177,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.852323008849558e-06,
+      "loss": 0.7711,
+      "step": 2223
+    },
+    {
+      "epoch": 2.4601769911504423,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 3.849557522123894e-06,
+      "loss": 0.7411,
+      "step": 2224
+    },
+    {
+      "epoch": 2.461283185840708,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 3.84679203539823e-06,
+      "loss": 0.7766,
+      "step": 2225
+    },
+    {
+      "epoch": 2.4623893805309733,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.844026548672566e-06,
+      "loss": 0.7806,
+      "step": 2226
+    },
+    {
+      "epoch": 2.463495575221239,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.841261061946903e-06,
+      "loss": 0.7713,
+      "step": 2227
+    },
+    {
+      "epoch": 2.4646017699115044,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 3.838495575221239e-06,
+      "loss": 0.7208,
+      "step": 2228
+    },
+    {
+      "epoch": 2.46570796460177,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.8357300884955755e-06,
+      "loss": 0.76,
+      "step": 2229
+    },
+    {
+      "epoch": 2.4668141592920354,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.832964601769912e-06,
+      "loss": 0.7487,
+      "step": 2230
+    },
+    {
+      "epoch": 2.4679203539823007,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.830199115044248e-06,
+      "loss": 0.7623,
+      "step": 2231
+    },
+    {
+      "epoch": 2.4690265486725664,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.827433628318584e-06,
+      "loss": 0.8855,
+      "step": 2232
+    },
+    {
+      "epoch": 2.4701327433628317,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.82466814159292e-06,
+      "loss": 0.8102,
+      "step": 2233
+    },
+    {
+      "epoch": 2.4712389380530975,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.821902654867257e-06,
+      "loss": 0.7549,
+      "step": 2234
+    },
+    {
+      "epoch": 2.4723451327433628,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.819137168141593e-06,
+      "loss": 0.8482,
+      "step": 2235
+    },
+    {
+      "epoch": 2.4734513274336285,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.816371681415929e-06,
+      "loss": 0.8275,
+      "step": 2236
+    },
+    {
+      "epoch": 2.474557522123894,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 3.813606194690266e-06,
+      "loss": 0.7511,
+      "step": 2237
+    },
+    {
+      "epoch": 2.475663716814159,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.810840707964602e-06,
+      "loss": 0.7424,
+      "step": 2238
+    },
+    {
+      "epoch": 2.476769911504425,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.808075221238938e-06,
+      "loss": 0.7898,
+      "step": 2239
+    },
+    {
+      "epoch": 2.47787610619469,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.8053097345132746e-06,
+      "loss": 0.8169,
+      "step": 2240
+    },
+    {
+      "epoch": 2.478982300884956,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.8025442477876107e-06,
+      "loss": 0.7463,
+      "step": 2241
+    },
+    {
+      "epoch": 2.480088495575221,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 3.7997787610619473e-06,
+      "loss": 0.743,
+      "step": 2242
+    },
+    {
+      "epoch": 2.481194690265487,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 3.797013274336284e-06,
+      "loss": 0.7448,
+      "step": 2243
+    },
+    {
+      "epoch": 2.482300884955752,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.79424778761062e-06,
+      "loss": 0.758,
+      "step": 2244
+    },
+    {
+      "epoch": 2.4834070796460175,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.791482300884956e-06,
+      "loss": 0.7918,
+      "step": 2245
+    },
+    {
+      "epoch": 2.484513274336283,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.7887168141592922e-06,
+      "loss": 0.7895,
+      "step": 2246
+    },
+    {
+      "epoch": 2.4856194690265485,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 3.7859513274336283e-06,
+      "loss": 0.7391,
+      "step": 2247
+    },
+    {
+      "epoch": 2.4867256637168142,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.783185840707965e-06,
+      "loss": 0.7993,
+      "step": 2248
+    },
+    {
+      "epoch": 2.4878318584070795,
+      "grad_norm": 0.107421875,
+      "learning_rate": 3.7804203539823014e-06,
+      "loss": 0.7699,
+      "step": 2249
+    },
+    {
+      "epoch": 2.4889380530973453,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 3.7776548672566376e-06,
+      "loss": 0.76,
+      "step": 2250
+    },
+    {
+      "epoch": 2.4900442477876106,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.7748893805309737e-06,
+      "loss": 0.7614,
+      "step": 2251
+    },
+    {
+      "epoch": 2.491150442477876,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 3.7721238938053102e-06,
+      "loss": 0.7936,
+      "step": 2252
+    },
+    {
+      "epoch": 2.4922566371681416,
+      "grad_norm": 0.12109375,
+      "learning_rate": 3.7693584070796464e-06,
+      "loss": 0.8247,
+      "step": 2253
+    },
+    {
+      "epoch": 2.493362831858407,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.7665929203539825e-06,
+      "loss": 0.8,
+      "step": 2254
+    },
+    {
+      "epoch": 2.4944690265486726,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.7638274336283186e-06,
+      "loss": 0.7685,
+      "step": 2255
+    },
+    {
+      "epoch": 2.495575221238938,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.7610619469026547e-06,
+      "loss": 0.7786,
+      "step": 2256
+    },
+    {
+      "epoch": 2.4966814159292037,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.7582964601769917e-06,
+      "loss": 0.7732,
+      "step": 2257
+    },
+    {
+      "epoch": 2.497787610619469,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.755530973451328e-06,
+      "loss": 0.7763,
+      "step": 2258
+    },
+    {
+      "epoch": 2.4988938053097347,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 3.752765486725664e-06,
+      "loss": 0.7181,
+      "step": 2259
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.8055,
+      "step": 2260
+    },
+    {
+      "epoch": 2.5011061946902657,
+      "grad_norm": 0.126953125,
+      "learning_rate": 3.7472345132743366e-06,
+      "loss": 0.8143,
+      "step": 2261
+    },
+    {
+      "epoch": 2.502212389380531,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 3.7444690265486728e-06,
+      "loss": 0.7515,
+      "step": 2262
+    },
+    {
+      "epoch": 2.5033185840707963,
+      "grad_norm": 0.125,
+      "learning_rate": 3.741703539823009e-06,
+      "loss": 0.8439,
+      "step": 2263
+    },
+    {
+      "epoch": 2.504424778761062,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.738938053097346e-06,
+      "loss": 0.7767,
+      "step": 2264
+    },
+    {
+      "epoch": 2.5055309734513274,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.736172566371682e-06,
+      "loss": 0.7788,
+      "step": 2265
+    },
+    {
+      "epoch": 2.5066371681415927,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.733407079646018e-06,
+      "loss": 0.7687,
+      "step": 2266
+    },
+    {
+      "epoch": 2.5077433628318584,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.7306415929203542e-06,
+      "loss": 0.8525,
+      "step": 2267
+    },
+    {
+      "epoch": 2.508849557522124,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.7278761061946904e-06,
+      "loss": 0.7555,
+      "step": 2268
+    },
+    {
+      "epoch": 2.5099557522123894,
+      "grad_norm": 0.126953125,
+      "learning_rate": 3.725110619469027e-06,
+      "loss": 0.802,
+      "step": 2269
+    },
+    {
+      "epoch": 2.5110619469026547,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.722345132743363e-06,
+      "loss": 0.7642,
+      "step": 2270
+    },
+    {
+      "epoch": 2.5121681415929205,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 3.719579646017699e-06,
+      "loss": 0.7416,
+      "step": 2271
+    },
+    {
+      "epoch": 2.5132743362831858,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.7168141592920357e-06,
+      "loss": 0.7791,
+      "step": 2272
+    },
+    {
+      "epoch": 2.5143805309734515,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.7140486725663723e-06,
+      "loss": 0.7638,
+      "step": 2273
+    },
+    {
+      "epoch": 2.515486725663717,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.7112831858407084e-06,
+      "loss": 0.7954,
+      "step": 2274
+    },
+    {
+      "epoch": 2.5165929203539825,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.7085176991150445e-06,
+      "loss": 0.7605,
+      "step": 2275
+    },
+    {
+      "epoch": 2.517699115044248,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.7057522123893807e-06,
+      "loss": 0.704,
+      "step": 2276
+    },
+    {
+      "epoch": 2.518805309734513,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.7029867256637168e-06,
+      "loss": 0.7748,
+      "step": 2277
+    },
+    {
+      "epoch": 2.519911504424779,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.7002212389380533e-06,
+      "loss": 0.7408,
+      "step": 2278
+    },
+    {
+      "epoch": 2.521017699115044,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.69745575221239e-06,
+      "loss": 0.7759,
+      "step": 2279
+    },
+    {
+      "epoch": 2.52212389380531,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.694690265486726e-06,
+      "loss": 0.7699,
+      "step": 2280
+    },
+    {
+      "epoch": 2.523230088495575,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.6919247787610626e-06,
+      "loss": 0.8192,
+      "step": 2281
+    },
+    {
+      "epoch": 2.524336283185841,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 3.6891592920353987e-06,
+      "loss": 0.7173,
+      "step": 2282
+    },
+    {
+      "epoch": 2.525442477876106,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.686393805309735e-06,
+      "loss": 0.8052,
+      "step": 2283
+    },
+    {
+      "epoch": 2.5265486725663715,
+      "grad_norm": 0.1015625,
+      "learning_rate": 3.683628318584071e-06,
+      "loss": 0.7501,
+      "step": 2284
+    },
+    {
+      "epoch": 2.5276548672566372,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 3.680862831858407e-06,
+      "loss": 0.7899,
+      "step": 2285
+    },
+    {
+      "epoch": 2.5287610619469025,
+      "grad_norm": 0.119140625,
+      "learning_rate": 3.6780973451327436e-06,
+      "loss": 0.7858,
+      "step": 2286
+    },
+    {
+      "epoch": 2.5298672566371683,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.67533185840708e-06,
+      "loss": 0.776,
+      "step": 2287
+    },
+    {
+      "epoch": 2.5309734513274336,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.6725663716814163e-06,
+      "loss": 0.8493,
+      "step": 2288
+    },
+    {
+      "epoch": 2.5320796460176993,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 3.6698008849557524e-06,
+      "loss": 0.7302,
+      "step": 2289
+    },
+    {
+      "epoch": 2.5331858407079646,
+      "grad_norm": 0.103515625,
+      "learning_rate": 3.667035398230089e-06,
+      "loss": 0.7547,
+      "step": 2290
+    },
+    {
+      "epoch": 2.53429203539823,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.664269911504425e-06,
+      "loss": 0.7894,
+      "step": 2291
+    },
+    {
+      "epoch": 2.5353982300884956,
+      "grad_norm": 0.12109375,
+      "learning_rate": 3.661504424778761e-06,
+      "loss": 0.8005,
+      "step": 2292
+    },
+    {
+      "epoch": 2.536504424778761,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.6587389380530973e-06,
+      "loss": 0.6834,
+      "step": 2293
+    },
+    {
+      "epoch": 2.5376106194690267,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 3.6559734513274343e-06,
+      "loss": 0.7679,
+      "step": 2294
+    },
+    {
+      "epoch": 2.538716814159292,
+      "grad_norm": 0.103515625,
+      "learning_rate": 3.6532079646017704e-06,
+      "loss": 0.741,
+      "step": 2295
+    },
+    {
+      "epoch": 2.5398230088495577,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.6504424778761066e-06,
+      "loss": 0.7697,
+      "step": 2296
+    },
+    {
+      "epoch": 2.540929203539823,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 3.6476769911504427e-06,
+      "loss": 0.749,
+      "step": 2297
+    },
+    {
+      "epoch": 2.5420353982300883,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.6449115044247792e-06,
+      "loss": 0.788,
+      "step": 2298
+    },
+    {
+      "epoch": 2.543141592920354,
+      "grad_norm": 0.107421875,
+      "learning_rate": 3.6421460176991154e-06,
+      "loss": 0.7912,
+      "step": 2299
+    },
+    {
+      "epoch": 2.5442477876106193,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.6393805309734515e-06,
+      "loss": 0.7613,
+      "step": 2300
+    },
+    {
+      "epoch": 2.545353982300885,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 3.6366150442477876e-06,
+      "loss": 0.7685,
+      "step": 2301
+    },
+    {
+      "epoch": 2.5464601769911503,
+      "grad_norm": 0.123046875,
+      "learning_rate": 3.6338495575221246e-06,
+      "loss": 0.8087,
+      "step": 2302
+    },
+    {
+      "epoch": 2.547566371681416,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.6310840707964607e-06,
+      "loss": 0.7709,
+      "step": 2303
+    },
+    {
+      "epoch": 2.5486725663716814,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 3.628318584070797e-06,
+      "loss": 0.7379,
+      "step": 2304
+    },
+    {
+      "epoch": 2.5497787610619467,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.625553097345133e-06,
+      "loss": 0.7988,
+      "step": 2305
+    },
+    {
+      "epoch": 2.5508849557522124,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.622787610619469e-06,
+      "loss": 0.7546,
+      "step": 2306
+    },
+    {
+      "epoch": 2.551991150442478,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 3.6200221238938056e-06,
+      "loss": 0.7329,
+      "step": 2307
+    },
+    {
+      "epoch": 2.5530973451327434,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.6172566371681418e-06,
+      "loss": 0.8194,
+      "step": 2308
+    },
+    {
+      "epoch": 2.5542035398230087,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.6144911504424783e-06,
+      "loss": 0.7745,
+      "step": 2309
+    },
+    {
+      "epoch": 2.5553097345132745,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 3.611725663716815e-06,
+      "loss": 0.7692,
+      "step": 2310
+    },
+    {
+      "epoch": 2.5564159292035398,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.608960176991151e-06,
+      "loss": 0.7851,
+      "step": 2311
+    },
+    {
+      "epoch": 2.557522123893805,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.606194690265487e-06,
+      "loss": 0.7717,
+      "step": 2312
+    },
+    {
+      "epoch": 2.558628318584071,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 3.6034292035398232e-06,
+      "loss": 0.8096,
+      "step": 2313
+    },
+    {
+      "epoch": 2.5597345132743365,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.6006637168141594e-06,
+      "loss": 0.762,
+      "step": 2314
+    },
+    {
+      "epoch": 2.560840707964602,
+      "grad_norm": 0.1015625,
+      "learning_rate": 3.5978982300884955e-06,
+      "loss": 0.7342,
+      "step": 2315
+    },
+    {
+      "epoch": 2.561946902654867,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.595132743362832e-06,
+      "loss": 0.7764,
+      "step": 2316
+    },
+    {
+      "epoch": 2.563053097345133,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 3.5923672566371686e-06,
+      "loss": 0.7275,
+      "step": 2317
+    },
+    {
+      "epoch": 2.564159292035398,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.5896017699115047e-06,
+      "loss": 0.7446,
+      "step": 2318
+    },
+    {
+      "epoch": 2.5652654867256635,
+      "grad_norm": 0.126953125,
+      "learning_rate": 3.5868362831858413e-06,
+      "loss": 0.7688,
+      "step": 2319
+    },
+    {
+      "epoch": 2.566371681415929,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 3.5840707964601774e-06,
+      "loss": 0.7119,
+      "step": 2320
+    },
+    {
+      "epoch": 2.567477876106195,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.5813053097345135e-06,
+      "loss": 0.7609,
+      "step": 2321
+    },
+    {
+      "epoch": 2.5685840707964602,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.5785398230088496e-06,
+      "loss": 0.7727,
+      "step": 2322
+    },
+    {
+      "epoch": 2.5696902654867255,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.5757743362831858e-06,
+      "loss": 0.7741,
+      "step": 2323
+    },
+    {
+      "epoch": 2.5707964601769913,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.5730088495575223e-06,
+      "loss": 0.7593,
+      "step": 2324
+    },
+    {
+      "epoch": 2.5719026548672566,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.570243362831859e-06,
+      "loss": 0.7676,
+      "step": 2325
+    },
+    {
+      "epoch": 2.573008849557522,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.567477876106195e-06,
+      "loss": 0.7472,
+      "step": 2326
+    },
+    {
+      "epoch": 2.5741150442477876,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.564712389380531e-06,
+      "loss": 0.8022,
+      "step": 2327
+    },
+    {
+      "epoch": 2.5752212389380533,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.5619469026548677e-06,
+      "loss": 0.767,
+      "step": 2328
+    },
+    {
+      "epoch": 2.5763274336283186,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.559181415929204e-06,
+      "loss": 0.7512,
+      "step": 2329
+    },
+    {
+      "epoch": 2.577433628318584,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.55641592920354e-06,
+      "loss": 0.8395,
+      "step": 2330
+    },
+    {
+      "epoch": 2.5785398230088497,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.553650442477876e-06,
+      "loss": 0.7185,
+      "step": 2331
+    },
+    {
+      "epoch": 2.579646017699115,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.550884955752213e-06,
+      "loss": 0.7754,
+      "step": 2332
+    },
+    {
+      "epoch": 2.5807522123893807,
+      "grad_norm": 0.1015625,
+      "learning_rate": 3.548119469026549e-06,
+      "loss": 0.7446,
+      "step": 2333
+    },
+    {
+      "epoch": 2.581858407079646,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.5453539823008853e-06,
+      "loss": 0.7691,
+      "step": 2334
+    },
+    {
+      "epoch": 2.5829646017699117,
+      "grad_norm": 0.107421875,
+      "learning_rate": 3.5425884955752214e-06,
+      "loss": 0.7447,
+      "step": 2335
+    },
+    {
+      "epoch": 2.584070796460177,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 3.539823008849558e-06,
+      "loss": 0.8414,
+      "step": 2336
+    },
+    {
+      "epoch": 2.5851769911504423,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 3.537057522123894e-06,
+      "loss": 0.7883,
+      "step": 2337
+    },
+    {
+      "epoch": 2.586283185840708,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.53429203539823e-06,
+      "loss": 0.7557,
+      "step": 2338
+    },
+    {
+      "epoch": 2.5873893805309733,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 3.5315265486725663e-06,
+      "loss": 0.7514,
+      "step": 2339
+    },
+    {
+      "epoch": 2.588495575221239,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.5287610619469033e-06,
+      "loss": 0.7463,
+      "step": 2340
+    },
+    {
+      "epoch": 2.5896017699115044,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.5259955752212394e-06,
+      "loss": 0.7747,
+      "step": 2341
+    },
+    {
+      "epoch": 2.59070796460177,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.5232300884955756e-06,
+      "loss": 0.7773,
+      "step": 2342
+    },
+    {
+      "epoch": 2.5918141592920354,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 3.5204646017699117e-06,
+      "loss": 0.8396,
+      "step": 2343
+    },
+    {
+      "epoch": 2.5929203539823007,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.517699115044248e-06,
+      "loss": 0.7764,
+      "step": 2344
+    },
+    {
+      "epoch": 2.5940265486725664,
+      "grad_norm": 0.119140625,
+      "learning_rate": 3.5149336283185844e-06,
+      "loss": 0.7856,
+      "step": 2345
+    },
+    {
+      "epoch": 2.5951327433628317,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.5121681415929205e-06,
+      "loss": 0.7651,
+      "step": 2346
+    },
+    {
+      "epoch": 2.5962389380530975,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 3.509402654867257e-06,
+      "loss": 0.7326,
+      "step": 2347
+    },
+    {
+      "epoch": 2.5973451327433628,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 3.5066371681415936e-06,
+      "loss": 0.749,
+      "step": 2348
+    },
+    {
+      "epoch": 2.5984513274336285,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.5038716814159297e-06,
+      "loss": 0.8155,
+      "step": 2349
+    },
+    {
+      "epoch": 2.599557522123894,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.501106194690266e-06,
+      "loss": 0.759,
+      "step": 2350
+    },
+    {
+      "epoch": 2.600663716814159,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.498340707964602e-06,
+      "loss": 0.7975,
+      "step": 2351
+    },
+    {
+      "epoch": 2.601769911504425,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.495575221238938e-06,
+      "loss": 0.7684,
+      "step": 2352
+    },
+    {
+      "epoch": 2.60287610619469,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.4928097345132746e-06,
+      "loss": 0.8249,
+      "step": 2353
+    },
+    {
+      "epoch": 2.603982300884956,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.4900442477876108e-06,
+      "loss": 0.7311,
+      "step": 2354
+    },
+    {
+      "epoch": 2.605088495575221,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.4872787610619473e-06,
+      "loss": 0.7686,
+      "step": 2355
+    },
+    {
+      "epoch": 2.606194690265487,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 3.4845132743362834e-06,
+      "loss": 0.755,
+      "step": 2356
+    },
+    {
+      "epoch": 2.607300884955752,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.48174778761062e-06,
+      "loss": 0.7885,
+      "step": 2357
+    },
+    {
+      "epoch": 2.6084070796460175,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.478982300884956e-06,
+      "loss": 0.7568,
+      "step": 2358
+    },
+    {
+      "epoch": 2.609513274336283,
+      "grad_norm": 0.107421875,
+      "learning_rate": 3.4762168141592922e-06,
+      "loss": 0.7547,
+      "step": 2359
+    },
+    {
+      "epoch": 2.6106194690265485,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 3.4734513274336284e-06,
+      "loss": 0.7532,
+      "step": 2360
+    },
+    {
+      "epoch": 2.6117256637168142,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.4706858407079645e-06,
+      "loss": 0.7883,
+      "step": 2361
+    },
+    {
+      "epoch": 2.6128318584070795,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.4679203539823015e-06,
+      "loss": 0.7652,
+      "step": 2362
+    },
+    {
+      "epoch": 2.6139380530973453,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.4651548672566376e-06,
+      "loss": 0.8039,
+      "step": 2363
+    },
+    {
+      "epoch": 2.6150442477876106,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 3.4623893805309737e-06,
+      "loss": 0.8056,
+      "step": 2364
+    },
+    {
+      "epoch": 2.616150442477876,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 3.45962389380531e-06,
+      "loss": 0.8017,
+      "step": 2365
+    },
+    {
+      "epoch": 2.6172566371681416,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.4568584070796464e-06,
+      "loss": 0.7157,
+      "step": 2366
+    },
+    {
+      "epoch": 2.6183628318584073,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.4540929203539825e-06,
+      "loss": 0.7503,
+      "step": 2367
+    },
+    {
+      "epoch": 2.6194690265486726,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.4513274336283186e-06,
+      "loss": 0.8025,
+      "step": 2368
+    },
+    {
+      "epoch": 2.620575221238938,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 3.4485619469026548e-06,
+      "loss": 0.7418,
+      "step": 2369
+    },
+    {
+      "epoch": 2.6216814159292037,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 3.4457964601769917e-06,
+      "loss": 0.7625,
+      "step": 2370
+    },
+    {
+      "epoch": 2.622787610619469,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.443030973451328e-06,
+      "loss": 0.7686,
+      "step": 2371
+    },
+    {
+      "epoch": 2.6238938053097343,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.440265486725664e-06,
+      "loss": 0.8117,
+      "step": 2372
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.4375e-06,
+      "loss": 0.8006,
+      "step": 2373
+    },
+    {
+      "epoch": 2.6261061946902657,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.4347345132743367e-06,
+      "loss": 0.7841,
+      "step": 2374
+    },
+    {
+      "epoch": 2.627212389380531,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.431969026548673e-06,
+      "loss": 0.7664,
+      "step": 2375
+    },
+    {
+      "epoch": 2.6283185840707963,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.429203539823009e-06,
+      "loss": 0.7188,
+      "step": 2376
+    },
+    {
+      "epoch": 2.629424778761062,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.4264380530973455e-06,
+      "loss": 0.7688,
+      "step": 2377
+    },
+    {
+      "epoch": 2.6305309734513274,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.423672566371682e-06,
+      "loss": 0.7756,
+      "step": 2378
+    },
+    {
+      "epoch": 2.6316371681415927,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.420907079646018e-06,
+      "loss": 0.7922,
+      "step": 2379
+    },
+    {
+      "epoch": 2.6327433628318584,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.4181415929203543e-06,
+      "loss": 0.7619,
+      "step": 2380
+    },
+    {
+      "epoch": 2.633849557522124,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.4153761061946904e-06,
+      "loss": 0.7587,
+      "step": 2381
+    },
+    {
+      "epoch": 2.6349557522123894,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.4126106194690265e-06,
+      "loss": 0.7529,
+      "step": 2382
+    },
+    {
+      "epoch": 2.6360619469026547,
+      "grad_norm": 0.1015625,
+      "learning_rate": 3.409845132743363e-06,
+      "loss": 0.7296,
+      "step": 2383
+    },
+    {
+      "epoch": 2.6371681415929205,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.407079646017699e-06,
+      "loss": 0.7776,
+      "step": 2384
+    },
+    {
+      "epoch": 2.6382743362831858,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 3.4043141592920358e-06,
+      "loss": 0.8025,
+      "step": 2385
+    },
+    {
+      "epoch": 2.6393805309734515,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.4015486725663723e-06,
+      "loss": 0.7827,
+      "step": 2386
+    },
+    {
+      "epoch": 2.640486725663717,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.3987831858407084e-06,
+      "loss": 0.773,
+      "step": 2387
+    },
+    {
+      "epoch": 2.6415929203539825,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 3.3960176991150446e-06,
+      "loss": 0.7548,
+      "step": 2388
+    },
+    {
+      "epoch": 2.642699115044248,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.3932522123893807e-06,
+      "loss": 0.7402,
+      "step": 2389
+    },
+    {
+      "epoch": 2.643805309734513,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.390486725663717e-06,
+      "loss": 0.7699,
+      "step": 2390
+    },
+    {
+      "epoch": 2.644911504424779,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.3877212389380534e-06,
+      "loss": 0.796,
+      "step": 2391
+    },
+    {
+      "epoch": 2.646017699115044,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.38495575221239e-06,
+      "loss": 0.7598,
+      "step": 2392
+    },
+    {
+      "epoch": 2.64712389380531,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.382190265486726e-06,
+      "loss": 0.7687,
+      "step": 2393
+    },
+    {
+      "epoch": 2.648230088495575,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 3.379424778761062e-06,
+      "loss": 0.806,
+      "step": 2394
+    },
+    {
+      "epoch": 2.649336283185841,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.3766592920353987e-06,
+      "loss": 0.8095,
+      "step": 2395
+    },
+    {
+      "epoch": 2.650442477876106,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 3.373893805309735e-06,
+      "loss": 0.7476,
+      "step": 2396
+    },
+    {
+      "epoch": 2.6515486725663715,
+      "grad_norm": 0.134765625,
+      "learning_rate": 3.371128318584071e-06,
+      "loss": 0.7408,
+      "step": 2397
+    },
+    {
+      "epoch": 2.6526548672566372,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.368362831858407e-06,
+      "loss": 0.7591,
+      "step": 2398
+    },
+    {
+      "epoch": 2.6537610619469025,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.365597345132743e-06,
+      "loss": 0.761,
+      "step": 2399
+    },
+    {
+      "epoch": 2.6548672566371683,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.36283185840708e-06,
+      "loss": 0.7572,
+      "step": 2400
+    },
+    {
+      "epoch": 2.6559734513274336,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.3600663716814163e-06,
+      "loss": 0.8115,
+      "step": 2401
+    },
+    {
+      "epoch": 2.6570796460176993,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.3573008849557524e-06,
+      "loss": 0.752,
+      "step": 2402
+    },
+    {
+      "epoch": 2.6581858407079646,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.354535398230089e-06,
+      "loss": 0.7666,
+      "step": 2403
+    },
+    {
+      "epoch": 2.65929203539823,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.351769911504425e-06,
+      "loss": 0.7992,
+      "step": 2404
+    },
+    {
+      "epoch": 2.6603982300884956,
+      "grad_norm": 0.125,
+      "learning_rate": 3.3490044247787612e-06,
+      "loss": 0.7463,
+      "step": 2405
+    },
+    {
+      "epoch": 2.661504424778761,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.3462389380530974e-06,
+      "loss": 0.7937,
+      "step": 2406
+    },
+    {
+      "epoch": 2.6626106194690267,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.3434734513274343e-06,
+      "loss": 0.7559,
+      "step": 2407
+    },
+    {
+      "epoch": 2.663716814159292,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.3407079646017705e-06,
+      "loss": 0.8587,
+      "step": 2408
+    },
+    {
+      "epoch": 2.6648230088495577,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 3.3379424778761066e-06,
+      "loss": 0.7538,
+      "step": 2409
+    },
+    {
+      "epoch": 2.665929203539823,
+      "grad_norm": 0.123046875,
+      "learning_rate": 3.3351769911504427e-06,
+      "loss": 0.7253,
+      "step": 2410
+    },
+    {
+      "epoch": 2.6670353982300883,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 3.332411504424779e-06,
+      "loss": 0.7538,
+      "step": 2411
+    },
+    {
+      "epoch": 2.668141592920354,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.3296460176991154e-06,
+      "loss": 0.7951,
+      "step": 2412
+    },
+    {
+      "epoch": 2.6692477876106193,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.3268805309734515e-06,
+      "loss": 0.7451,
+      "step": 2413
+    },
+    {
+      "epoch": 2.670353982300885,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 3.3241150442477876e-06,
+      "loss": 0.7847,
+      "step": 2414
+    },
+    {
+      "epoch": 2.6714601769911503,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.321349557522124e-06,
+      "loss": 0.7929,
+      "step": 2415
+    },
+    {
+      "epoch": 2.672566371681416,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.3185840707964607e-06,
+      "loss": 0.7695,
+      "step": 2416
+    },
+    {
+      "epoch": 2.6736725663716814,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 3.315818584070797e-06,
+      "loss": 0.7645,
+      "step": 2417
+    },
+    {
+      "epoch": 2.6747787610619467,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 3.313053097345133e-06,
+      "loss": 0.8093,
+      "step": 2418
+    },
+    {
+      "epoch": 2.6758849557522124,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.310287610619469e-06,
+      "loss": 0.7731,
+      "step": 2419
+    },
+    {
+      "epoch": 2.676991150442478,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 3.3075221238938052e-06,
+      "loss": 0.7986,
+      "step": 2420
+    },
+    {
+      "epoch": 2.6780973451327434,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 3.304756637168142e-06,
+      "loss": 0.7538,
+      "step": 2421
+    },
+    {
+      "epoch": 2.6792035398230087,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.3019911504424783e-06,
+      "loss": 0.764,
+      "step": 2422
+    },
+    {
+      "epoch": 2.6803097345132745,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 3.2992256637168145e-06,
+      "loss": 0.7712,
+      "step": 2423
+    },
+    {
+      "epoch": 2.6814159292035398,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.296460176991151e-06,
+      "loss": 0.8029,
+      "step": 2424
+    },
+    {
+      "epoch": 2.682522123893805,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.293694690265487e-06,
+      "loss": 0.7788,
+      "step": 2425
+    },
+    {
+      "epoch": 2.683628318584071,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.2909292035398233e-06,
+      "loss": 0.7485,
+      "step": 2426
+    },
+    {
+      "epoch": 2.6847345132743365,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.2881637168141594e-06,
+      "loss": 0.7477,
+      "step": 2427
+    },
+    {
+      "epoch": 2.685840707964602,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.2853982300884955e-06,
+      "loss": 0.7926,
+      "step": 2428
+    },
+    {
+      "epoch": 2.686946902654867,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.282632743362832e-06,
+      "loss": 0.7902,
+      "step": 2429
+    },
+    {
+      "epoch": 2.688053097345133,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.2798672566371686e-06,
+      "loss": 0.7606,
+      "step": 2430
+    },
+    {
+      "epoch": 2.689159292035398,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.2771017699115047e-06,
+      "loss": 0.7647,
+      "step": 2431
+    },
+    {
+      "epoch": 2.6902654867256635,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 3.274336283185841e-06,
+      "loss": 0.8484,
+      "step": 2432
+    },
+    {
+      "epoch": 2.691371681415929,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 3.2715707964601774e-06,
+      "loss": 0.7867,
+      "step": 2433
+    },
+    {
+      "epoch": 2.692477876106195,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.2688053097345136e-06,
+      "loss": 0.8294,
+      "step": 2434
+    },
+    {
+      "epoch": 2.6935840707964602,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.2660398230088497e-06,
+      "loss": 0.7912,
+      "step": 2435
+    },
+    {
+      "epoch": 2.6946902654867255,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.263274336283186e-06,
+      "loss": 0.7663,
+      "step": 2436
+    },
+    {
+      "epoch": 2.6957964601769913,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.260508849557522e-06,
+      "loss": 0.7467,
+      "step": 2437
+    },
+    {
+      "epoch": 2.6969026548672566,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.257743362831859e-06,
+      "loss": 0.8142,
+      "step": 2438
+    },
+    {
+      "epoch": 2.698008849557522,
+      "grad_norm": 0.1015625,
+      "learning_rate": 3.254977876106195e-06,
+      "loss": 0.7561,
+      "step": 2439
+    },
+    {
+      "epoch": 2.6991150442477876,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.252212389380531e-06,
+      "loss": 0.8329,
+      "step": 2440
+    },
+    {
+      "epoch": 2.7002212389380533,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.2494469026548677e-06,
+      "loss": 0.7774,
+      "step": 2441
+    },
+    {
+      "epoch": 2.7013274336283186,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.246681415929204e-06,
+      "loss": 0.7814,
+      "step": 2442
+    },
+    {
+      "epoch": 2.702433628318584,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.24391592920354e-06,
+      "loss": 0.7653,
+      "step": 2443
+    },
+    {
+      "epoch": 2.7035398230088497,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.241150442477876e-06,
+      "loss": 0.7786,
+      "step": 2444
+    },
+    {
+      "epoch": 2.704646017699115,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.238384955752213e-06,
+      "loss": 0.7432,
+      "step": 2445
+    },
+    {
+      "epoch": 2.7057522123893807,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.235619469026549e-06,
+      "loss": 0.8105,
+      "step": 2446
+    },
+    {
+      "epoch": 2.706858407079646,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.2328539823008853e-06,
+      "loss": 0.7337,
+      "step": 2447
+    },
+    {
+      "epoch": 2.7079646017699117,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 3.2300884955752214e-06,
+      "loss": 0.738,
+      "step": 2448
+    },
+    {
+      "epoch": 2.709070796460177,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.2273230088495576e-06,
+      "loss": 0.8209,
+      "step": 2449
+    },
+    {
+      "epoch": 2.7101769911504423,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.224557522123894e-06,
+      "loss": 0.7394,
+      "step": 2450
+    },
+    {
+      "epoch": 2.711283185840708,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.2217920353982302e-06,
+      "loss": 0.7909,
+      "step": 2451
+    },
+    {
+      "epoch": 2.7123893805309733,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.2190265486725664e-06,
+      "loss": 0.748,
+      "step": 2452
+    },
+    {
+      "epoch": 2.713495575221239,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.2162610619469033e-06,
+      "loss": 0.7764,
+      "step": 2453
+    },
+    {
+      "epoch": 2.7146017699115044,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 3.2134955752212395e-06,
+      "loss": 0.7376,
+      "step": 2454
+    },
+    {
+      "epoch": 2.71570796460177,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.2107300884955756e-06,
+      "loss": 0.7906,
+      "step": 2455
+    },
+    {
+      "epoch": 2.7168141592920354,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 3.2079646017699117e-06,
+      "loss": 0.7244,
+      "step": 2456
+    },
+    {
+      "epoch": 2.7179203539823007,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.205199115044248e-06,
+      "loss": 0.7889,
+      "step": 2457
+    },
+    {
+      "epoch": 2.7190265486725664,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.2024336283185844e-06,
+      "loss": 0.7765,
+      "step": 2458
+    },
+    {
+      "epoch": 2.7201327433628317,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 3.1996681415929205e-06,
+      "loss": 0.7752,
+      "step": 2459
+    },
+    {
+      "epoch": 2.7212389380530975,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.196902654867257e-06,
+      "loss": 0.7337,
+      "step": 2460
+    },
+    {
+      "epoch": 2.7223451327433628,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 3.194137168141593e-06,
+      "loss": 0.7912,
+      "step": 2461
+    },
+    {
+      "epoch": 2.7234513274336285,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.1913716814159297e-06,
+      "loss": 0.7822,
+      "step": 2462
+    },
+    {
+      "epoch": 2.724557522123894,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.188606194690266e-06,
+      "loss": 0.795,
+      "step": 2463
+    },
+    {
+      "epoch": 2.725663716814159,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.185840707964602e-06,
+      "loss": 0.7607,
+      "step": 2464
+    },
+    {
+      "epoch": 2.726769911504425,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.183075221238938e-06,
+      "loss": 0.7267,
+      "step": 2465
+    },
+    {
+      "epoch": 2.72787610619469,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.1803097345132742e-06,
+      "loss": 0.8173,
+      "step": 2466
+    },
+    {
+      "epoch": 2.728982300884956,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 3.177544247787611e-06,
+      "loss": 0.7931,
+      "step": 2467
+    },
+    {
+      "epoch": 2.730088495575221,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.1747787610619473e-06,
+      "loss": 0.7707,
+      "step": 2468
+    },
+    {
+      "epoch": 2.731194690265487,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 3.1720132743362835e-06,
+      "loss": 0.7684,
+      "step": 2469
+    },
+    {
+      "epoch": 2.732300884955752,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.1692477876106196e-06,
+      "loss": 0.823,
+      "step": 2470
+    },
+    {
+      "epoch": 2.7334070796460175,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.166482300884956e-06,
+      "loss": 0.7913,
+      "step": 2471
+    },
+    {
+      "epoch": 2.734513274336283,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.1637168141592923e-06,
+      "loss": 0.7604,
+      "step": 2472
+    },
+    {
+      "epoch": 2.7356194690265485,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 3.1609513274336284e-06,
+      "loss": 0.7729,
+      "step": 2473
+    },
+    {
+      "epoch": 2.7367256637168142,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.1581858407079645e-06,
+      "loss": 0.782,
+      "step": 2474
+    },
+    {
+      "epoch": 2.7378318584070795,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.1554203539823015e-06,
+      "loss": 0.7738,
+      "step": 2475
+    },
+    {
+      "epoch": 2.7389380530973453,
+      "grad_norm": 0.119140625,
+      "learning_rate": 3.1526548672566376e-06,
+      "loss": 0.8098,
+      "step": 2476
+    },
+    {
+      "epoch": 2.7400442477876106,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.1498893805309737e-06,
+      "loss": 0.7716,
+      "step": 2477
+    },
+    {
+      "epoch": 2.741150442477876,
+      "grad_norm": 0.1015625,
+      "learning_rate": 3.14712389380531e-06,
+      "loss": 0.7059,
+      "step": 2478
+    },
+    {
+      "epoch": 2.7422566371681416,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 3.1443584070796464e-06,
+      "loss": 0.7795,
+      "step": 2479
+    },
+    {
+      "epoch": 2.7433628318584073,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 3.1415929203539825e-06,
+      "loss": 0.8453,
+      "step": 2480
+    },
+    {
+      "epoch": 2.7444690265486726,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 3.1388274336283187e-06,
+      "loss": 0.7756,
+      "step": 2481
+    },
+    {
+      "epoch": 2.745575221238938,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.136061946902655e-06,
+      "loss": 0.7622,
+      "step": 2482
+    },
+    {
+      "epoch": 2.7466814159292037,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.1332964601769918e-06,
+      "loss": 0.8567,
+      "step": 2483
+    },
+    {
+      "epoch": 2.747787610619469,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.130530973451328e-06,
+      "loss": 0.7552,
+      "step": 2484
+    },
+    {
+      "epoch": 2.7488938053097343,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.127765486725664e-06,
+      "loss": 0.7517,
+      "step": 2485
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 3.125e-06,
+      "loss": 0.7711,
+      "step": 2486
+    },
+    {
+      "epoch": 2.7511061946902657,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.1222345132743363e-06,
+      "loss": 0.7731,
+      "step": 2487
+    },
+    {
+      "epoch": 2.752212389380531,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.119469026548673e-06,
+      "loss": 0.7377,
+      "step": 2488
+    },
+    {
+      "epoch": 2.7533185840707963,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 3.116703539823009e-06,
+      "loss": 0.7707,
+      "step": 2489
+    },
+    {
+      "epoch": 2.754424778761062,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.1139380530973455e-06,
+      "loss": 0.7778,
+      "step": 2490
+    },
+    {
+      "epoch": 2.7555309734513274,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 3.111172566371682e-06,
+      "loss": 0.74,
+      "step": 2491
+    },
+    {
+      "epoch": 2.7566371681415927,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 3.108407079646018e-06,
+      "loss": 0.8017,
+      "step": 2492
+    },
+    {
+      "epoch": 2.7577433628318584,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 3.1056415929203543e-06,
+      "loss": 0.7959,
+      "step": 2493
+    },
+    {
+      "epoch": 2.758849557522124,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.1028761061946904e-06,
+      "loss": 0.784,
+      "step": 2494
+    },
+    {
+      "epoch": 2.7599557522123894,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.1001106194690266e-06,
+      "loss": 0.7573,
+      "step": 2495
+    },
+    {
+      "epoch": 2.7610619469026547,
+      "grad_norm": 0.099609375,
+      "learning_rate": 3.097345132743363e-06,
+      "loss": 0.7235,
+      "step": 2496
+    },
+    {
+      "epoch": 2.7621681415929205,
+      "grad_norm": 0.107421875,
+      "learning_rate": 3.0945796460176992e-06,
+      "loss": 0.7746,
+      "step": 2497
+    },
+    {
+      "epoch": 2.7632743362831858,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.0918141592920358e-06,
+      "loss": 0.737,
+      "step": 2498
+    },
+    {
+      "epoch": 2.7643805309734515,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 3.089048672566372e-06,
+      "loss": 0.7887,
+      "step": 2499
+    },
+    {
+      "epoch": 2.765486725663717,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.0862831858407085e-06,
+      "loss": 0.7436,
+      "step": 2500
+    },
+    {
+      "epoch": 2.7665929203539825,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 3.0835176991150446e-06,
+      "loss": 0.768,
+      "step": 2501
+    },
+    {
+      "epoch": 2.767699115044248,
+      "grad_norm": 0.103515625,
+      "learning_rate": 3.0807522123893807e-06,
+      "loss": 0.7283,
+      "step": 2502
+    },
+    {
+      "epoch": 2.768805309734513,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 3.077986725663717e-06,
+      "loss": 0.7945,
+      "step": 2503
+    },
+    {
+      "epoch": 2.769911504424779,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.075221238938053e-06,
+      "loss": 0.7602,
+      "step": 2504
+    },
+    {
+      "epoch": 2.771017699115044,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.07245575221239e-06,
+      "loss": 0.7633,
+      "step": 2505
+    },
+    {
+      "epoch": 2.77212389380531,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 3.069690265486726e-06,
+      "loss": 0.7445,
+      "step": 2506
+    },
+    {
+      "epoch": 2.773230088495575,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 3.066924778761062e-06,
+      "loss": 0.7845,
+      "step": 2507
+    },
+    {
+      "epoch": 2.774336283185841,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.0641592920353987e-06,
+      "loss": 0.7766,
+      "step": 2508
+    },
+    {
+      "epoch": 2.775442477876106,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 3.061393805309735e-06,
+      "loss": 0.749,
+      "step": 2509
+    },
+    {
+      "epoch": 2.7765486725663715,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.058628318584071e-06,
+      "loss": 0.7598,
+      "step": 2510
+    },
+    {
+      "epoch": 2.7776548672566372,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.055862831858407e-06,
+      "loss": 0.758,
+      "step": 2511
+    },
+    {
+      "epoch": 2.7787610619469025,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.0530973451327432e-06,
+      "loss": 0.8028,
+      "step": 2512
+    },
+    {
+      "epoch": 2.7798672566371683,
+      "grad_norm": 0.103515625,
+      "learning_rate": 3.05033185840708e-06,
+      "loss": 0.7597,
+      "step": 2513
+    },
+    {
+      "epoch": 2.7809734513274336,
+      "grad_norm": 0.119140625,
+      "learning_rate": 3.0475663716814163e-06,
+      "loss": 0.796,
+      "step": 2514
+    },
+    {
+      "epoch": 2.7820796460176993,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 3.0448008849557525e-06,
+      "loss": 0.8348,
+      "step": 2515
+    },
+    {
+      "epoch": 2.7831858407079646,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.0420353982300886e-06,
+      "loss": 0.7594,
+      "step": 2516
+    },
+    {
+      "epoch": 2.78429203539823,
+      "grad_norm": 0.193359375,
+      "learning_rate": 3.039269911504425e-06,
+      "loss": 0.7736,
+      "step": 2517
+    },
+    {
+      "epoch": 2.7853982300884956,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 3.0365044247787613e-06,
+      "loss": 0.7082,
+      "step": 2518
+    },
+    {
+      "epoch": 2.786504424778761,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.0337389380530974e-06,
+      "loss": 0.8234,
+      "step": 2519
+    },
+    {
+      "epoch": 2.7876106194690267,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.030973451327434e-06,
+      "loss": 0.7222,
+      "step": 2520
+    },
+    {
+      "epoch": 2.788716814159292,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 3.0282079646017705e-06,
+      "loss": 0.782,
+      "step": 2521
+    },
+    {
+      "epoch": 2.7898230088495577,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 3.0254424778761066e-06,
+      "loss": 0.76,
+      "step": 2522
+    },
+    {
+      "epoch": 2.790929203539823,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.0226769911504427e-06,
+      "loss": 0.7949,
+      "step": 2523
+    },
+    {
+      "epoch": 2.7920353982300883,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 3.019911504424779e-06,
+      "loss": 0.7327,
+      "step": 2524
+    },
+    {
+      "epoch": 2.793141592920354,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 3.017146017699115e-06,
+      "loss": 0.7711,
+      "step": 2525
+    },
+    {
+      "epoch": 2.7942477876106193,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.0143805309734515e-06,
+      "loss": 0.7194,
+      "step": 2526
+    },
+    {
+      "epoch": 2.795353982300885,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.0116150442477877e-06,
+      "loss": 0.752,
+      "step": 2527
+    },
+    {
+      "epoch": 2.7964601769911503,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.0088495575221242e-06,
+      "loss": 0.7692,
+      "step": 2528
+    },
+    {
+      "epoch": 2.797566371681416,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.0060840707964608e-06,
+      "loss": 0.7849,
+      "step": 2529
+    },
+    {
+      "epoch": 2.7986725663716814,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.003318584070797e-06,
+      "loss": 0.7531,
+      "step": 2530
+    },
+    {
+      "epoch": 2.7997787610619467,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.000553097345133e-06,
+      "loss": 0.8197,
+      "step": 2531
+    },
+    {
+      "epoch": 2.8008849557522124,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.997787610619469e-06,
+      "loss": 0.7467,
+      "step": 2532
+    },
+    {
+      "epoch": 2.801991150442478,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.9950221238938053e-06,
+      "loss": 0.7392,
+      "step": 2533
+    },
+    {
+      "epoch": 2.8030973451327434,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 2.992256637168142e-06,
+      "loss": 0.7336,
+      "step": 2534
+    },
+    {
+      "epoch": 2.8042035398230087,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.9894911504424784e-06,
+      "loss": 0.8075,
+      "step": 2535
+    },
+    {
+      "epoch": 2.8053097345132745,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.9867256637168145e-06,
+      "loss": 0.7802,
+      "step": 2536
+    },
+    {
+      "epoch": 2.8064159292035398,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 2.9839601769911506e-06,
+      "loss": 0.7427,
+      "step": 2537
+    },
+    {
+      "epoch": 2.807522123893805,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.981194690265487e-06,
+      "loss": 0.8185,
+      "step": 2538
+    },
+    {
+      "epoch": 2.808628318584071,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.9784292035398233e-06,
+      "loss": 0.7892,
+      "step": 2539
+    },
+    {
+      "epoch": 2.8097345132743365,
+      "grad_norm": 0.115234375,
+      "learning_rate": 2.9756637168141594e-06,
+      "loss": 0.7736,
+      "step": 2540
+    },
+    {
+      "epoch": 2.810840707964602,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.9728982300884956e-06,
+      "loss": 0.7489,
+      "step": 2541
+    },
+    {
+      "epoch": 2.811946902654867,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.9701327433628317e-06,
+      "loss": 0.7665,
+      "step": 2542
+    },
+    {
+      "epoch": 2.813053097345133,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.9673672566371686e-06,
+      "loss": 0.7592,
+      "step": 2543
+    },
+    {
+      "epoch": 2.814159292035398,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.9646017699115048e-06,
+      "loss": 0.7802,
+      "step": 2544
+    },
+    {
+      "epoch": 2.8152654867256635,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 2.961836283185841e-06,
+      "loss": 0.7823,
+      "step": 2545
+    },
+    {
+      "epoch": 2.816371681415929,
+      "grad_norm": 0.115234375,
+      "learning_rate": 2.9590707964601775e-06,
+      "loss": 0.7607,
+      "step": 2546
+    },
+    {
+      "epoch": 2.817477876106195,
+      "grad_norm": 0.115234375,
+      "learning_rate": 2.9563053097345136e-06,
+      "loss": 0.7916,
+      "step": 2547
+    },
+    {
+      "epoch": 2.8185840707964602,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.9535398230088497e-06,
+      "loss": 0.8373,
+      "step": 2548
+    },
+    {
+      "epoch": 2.8196902654867255,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.950774336283186e-06,
+      "loss": 0.7805,
+      "step": 2549
+    },
+    {
+      "epoch": 2.8207964601769913,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.948008849557522e-06,
+      "loss": 0.8358,
+      "step": 2550
+    },
+    {
+      "epoch": 2.8219026548672566,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.945243362831859e-06,
+      "loss": 0.7752,
+      "step": 2551
+    },
+    {
+      "epoch": 2.823008849557522,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.942477876106195e-06,
+      "loss": 0.794,
+      "step": 2552
+    },
+    {
+      "epoch": 2.8241150442477876,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.939712389380531e-06,
+      "loss": 0.7589,
+      "step": 2553
+    },
+    {
+      "epoch": 2.8252212389380533,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.9369469026548673e-06,
+      "loss": 0.7685,
+      "step": 2554
+    },
+    {
+      "epoch": 2.8263274336283186,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.934181415929204e-06,
+      "loss": 0.7513,
+      "step": 2555
+    },
+    {
+      "epoch": 2.827433628318584,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.93141592920354e-06,
+      "loss": 0.743,
+      "step": 2556
+    },
+    {
+      "epoch": 2.8285398230088497,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.928650442477876e-06,
+      "loss": 0.7491,
+      "step": 2557
+    },
+    {
+      "epoch": 2.829646017699115,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.925884955752213e-06,
+      "loss": 0.7363,
+      "step": 2558
+    },
+    {
+      "epoch": 2.8307522123893807,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 2.923119469026549e-06,
+      "loss": 0.7487,
+      "step": 2559
+    },
+    {
+      "epoch": 2.831858407079646,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.9203539823008853e-06,
+      "loss": 0.7554,
+      "step": 2560
+    },
+    {
+      "epoch": 2.8329646017699117,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.9175884955752215e-06,
+      "loss": 0.7933,
+      "step": 2561
+    },
+    {
+      "epoch": 2.834070796460177,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.9148230088495576e-06,
+      "loss": 0.7904,
+      "step": 2562
+    },
+    {
+      "epoch": 2.8351769911504423,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.9120575221238937e-06,
+      "loss": 0.758,
+      "step": 2563
+    },
+    {
+      "epoch": 2.836283185840708,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.9092920353982303e-06,
+      "loss": 0.7974,
+      "step": 2564
+    },
+    {
+      "epoch": 2.8373893805309733,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.9065265486725664e-06,
+      "loss": 0.7383,
+      "step": 2565
+    },
+    {
+      "epoch": 2.838495575221239,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.903761061946903e-06,
+      "loss": 0.8321,
+      "step": 2566
+    },
+    {
+      "epoch": 2.8396017699115044,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.9009955752212395e-06,
+      "loss": 0.7976,
+      "step": 2567
+    },
+    {
+      "epoch": 2.84070796460177,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.8982300884955756e-06,
+      "loss": 0.7554,
+      "step": 2568
+    },
+    {
+      "epoch": 2.8418141592920354,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.8954646017699117e-06,
+      "loss": 0.7654,
+      "step": 2569
+    },
+    {
+      "epoch": 2.8429203539823007,
+      "grad_norm": 0.103515625,
+      "learning_rate": 2.892699115044248e-06,
+      "loss": 0.7178,
+      "step": 2570
+    },
+    {
+      "epoch": 2.8440265486725664,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.889933628318584e-06,
+      "loss": 0.8131,
+      "step": 2571
+    },
+    {
+      "epoch": 2.8451327433628317,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.8871681415929205e-06,
+      "loss": 0.7606,
+      "step": 2572
+    },
+    {
+      "epoch": 2.8462389380530975,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.884402654867257e-06,
+      "loss": 0.8261,
+      "step": 2573
+    },
+    {
+      "epoch": 2.8473451327433628,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.8816371681415932e-06,
+      "loss": 0.7768,
+      "step": 2574
+    },
+    {
+      "epoch": 2.8484513274336285,
+      "grad_norm": 0.119140625,
+      "learning_rate": 2.8788716814159293e-06,
+      "loss": 0.7861,
+      "step": 2575
+    },
+    {
+      "epoch": 2.849557522123894,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.876106194690266e-06,
+      "loss": 0.7568,
+      "step": 2576
+    },
+    {
+      "epoch": 2.850663716814159,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.873340707964602e-06,
+      "loss": 0.7969,
+      "step": 2577
+    },
+    {
+      "epoch": 2.851769911504425,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.870575221238938e-06,
+      "loss": 0.8032,
+      "step": 2578
+    },
+    {
+      "epoch": 2.85287610619469,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.8678097345132743e-06,
+      "loss": 0.7372,
+      "step": 2579
+    },
+    {
+      "epoch": 2.853982300884956,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.8650442477876104e-06,
+      "loss": 0.7744,
+      "step": 2580
+    },
+    {
+      "epoch": 2.855088495575221,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.8622787610619474e-06,
+      "loss": 0.7524,
+      "step": 2581
+    },
+    {
+      "epoch": 2.856194690265487,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.8595132743362835e-06,
+      "loss": 0.8006,
+      "step": 2582
+    },
+    {
+      "epoch": 2.857300884955752,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.8567477876106196e-06,
+      "loss": 0.7525,
+      "step": 2583
+    },
+    {
+      "epoch": 2.8584070796460175,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.853982300884956e-06,
+      "loss": 0.8799,
+      "step": 2584
+    },
+    {
+      "epoch": 2.859513274336283,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.8512168141592923e-06,
+      "loss": 0.7757,
+      "step": 2585
+    },
+    {
+      "epoch": 2.8606194690265485,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.8484513274336284e-06,
+      "loss": 0.7772,
+      "step": 2586
+    },
+    {
+      "epoch": 2.8617256637168142,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.8456858407079645e-06,
+      "loss": 0.7849,
+      "step": 2587
+    },
+    {
+      "epoch": 2.8628318584070795,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.8429203539823015e-06,
+      "loss": 0.7975,
+      "step": 2588
+    },
+    {
+      "epoch": 2.8639380530973453,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.8401548672566376e-06,
+      "loss": 0.7774,
+      "step": 2589
+    },
+    {
+      "epoch": 2.8650442477876106,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.8373893805309738e-06,
+      "loss": 0.7873,
+      "step": 2590
+    },
+    {
+      "epoch": 2.866150442477876,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.83462389380531e-06,
+      "loss": 0.7961,
+      "step": 2591
+    },
+    {
+      "epoch": 2.8672566371681416,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.831858407079646e-06,
+      "loss": 0.763,
+      "step": 2592
+    },
+    {
+      "epoch": 2.8683628318584073,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.8290929203539826e-06,
+      "loss": 0.7591,
+      "step": 2593
+    },
+    {
+      "epoch": 2.8694690265486726,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.8263274336283187e-06,
+      "loss": 0.7583,
+      "step": 2594
+    },
+    {
+      "epoch": 2.870575221238938,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.823561946902655e-06,
+      "loss": 0.8039,
+      "step": 2595
+    },
+    {
+      "epoch": 2.8716814159292037,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.820796460176992e-06,
+      "loss": 0.7299,
+      "step": 2596
+    },
+    {
+      "epoch": 2.872787610619469,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 2.818030973451328e-06,
+      "loss": 0.7747,
+      "step": 2597
+    },
+    {
+      "epoch": 2.8738938053097343,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.815265486725664e-06,
+      "loss": 0.7524,
+      "step": 2598
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.8125e-06,
+      "loss": 0.7915,
+      "step": 2599
+    },
+    {
+      "epoch": 2.8761061946902657,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.8097345132743363e-06,
+      "loss": 0.7859,
+      "step": 2600
+    },
+    {
+      "epoch": 2.877212389380531,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.806969026548673e-06,
+      "loss": 0.7501,
+      "step": 2601
+    },
+    {
+      "epoch": 2.8783185840707963,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.804203539823009e-06,
+      "loss": 0.7767,
+      "step": 2602
+    },
+    {
+      "epoch": 2.879424778761062,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.8014380530973455e-06,
+      "loss": 0.7602,
+      "step": 2603
+    },
+    {
+      "epoch": 2.8805309734513274,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.7986725663716817e-06,
+      "loss": 0.7582,
+      "step": 2604
+    },
+    {
+      "epoch": 2.8816371681415927,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 2.795907079646018e-06,
+      "loss": 0.775,
+      "step": 2605
+    },
+    {
+      "epoch": 2.8827433628318584,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.7931415929203543e-06,
+      "loss": 0.7454,
+      "step": 2606
+    },
+    {
+      "epoch": 2.883849557522124,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.7903761061946905e-06,
+      "loss": 0.7724,
+      "step": 2607
+    },
+    {
+      "epoch": 2.8849557522123894,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.7876106194690266e-06,
+      "loss": 0.7716,
+      "step": 2608
+    },
+    {
+      "epoch": 2.8860619469026547,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.7848451327433627e-06,
+      "loss": 0.7485,
+      "step": 2609
+    },
+    {
+      "epoch": 2.8871681415929205,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.7820796460176993e-06,
+      "loss": 0.7318,
+      "step": 2610
+    },
+    {
+      "epoch": 2.8882743362831858,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.779314159292036e-06,
+      "loss": 0.7535,
+      "step": 2611
+    },
+    {
+      "epoch": 2.8893805309734515,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.776548672566372e-06,
+      "loss": 0.7504,
+      "step": 2612
+    },
+    {
+      "epoch": 2.890486725663717,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.773783185840708e-06,
+      "loss": 0.7657,
+      "step": 2613
+    },
+    {
+      "epoch": 2.8915929203539825,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.7710176991150446e-06,
+      "loss": 0.7919,
+      "step": 2614
+    },
+    {
+      "epoch": 2.892699115044248,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.7682522123893807e-06,
+      "loss": 0.8021,
+      "step": 2615
+    },
+    {
+      "epoch": 2.893805309734513,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.765486725663717e-06,
+      "loss": 0.802,
+      "step": 2616
+    },
+    {
+      "epoch": 2.894911504424779,
+      "grad_norm": 0.1015625,
+      "learning_rate": 2.762721238938053e-06,
+      "loss": 0.7401,
+      "step": 2617
+    },
+    {
+      "epoch": 2.896017699115044,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.75995575221239e-06,
+      "loss": 0.8255,
+      "step": 2618
+    },
+    {
+      "epoch": 2.89712389380531,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.757190265486726e-06,
+      "loss": 0.783,
+      "step": 2619
+    },
+    {
+      "epoch": 2.898230088495575,
+      "grad_norm": 0.099609375,
+      "learning_rate": 2.754424778761062e-06,
+      "loss": 0.7362,
+      "step": 2620
+    },
+    {
+      "epoch": 2.899336283185841,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.7516592920353983e-06,
+      "loss": 0.8153,
+      "step": 2621
+    },
+    {
+      "epoch": 2.900442477876106,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.748893805309735e-06,
+      "loss": 0.772,
+      "step": 2622
+    },
+    {
+      "epoch": 2.9015486725663715,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.746128318584071e-06,
+      "loss": 0.7682,
+      "step": 2623
+    },
+    {
+      "epoch": 2.9026548672566372,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.743362831858407e-06,
+      "loss": 0.7462,
+      "step": 2624
+    },
+    {
+      "epoch": 2.9037610619469025,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.7405973451327433e-06,
+      "loss": 0.7736,
+      "step": 2625
+    },
+    {
+      "epoch": 2.9048672566371683,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.7378318584070802e-06,
+      "loss": 0.7347,
+      "step": 2626
+    },
+    {
+      "epoch": 2.9059734513274336,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.7350663716814164e-06,
+      "loss": 0.8035,
+      "step": 2627
+    },
+    {
+      "epoch": 2.9070796460176993,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.7323008849557525e-06,
+      "loss": 0.7663,
+      "step": 2628
+    },
+    {
+      "epoch": 2.9081858407079646,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.7295353982300886e-06,
+      "loss": 0.7724,
+      "step": 2629
+    },
+    {
+      "epoch": 2.90929203539823,
+      "grad_norm": 0.103515625,
+      "learning_rate": 2.7267699115044247e-06,
+      "loss": 0.7701,
+      "step": 2630
+    },
+    {
+      "epoch": 2.9103982300884956,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.7240044247787613e-06,
+      "loss": 0.7918,
+      "step": 2631
+    },
+    {
+      "epoch": 2.911504424778761,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.7212389380530974e-06,
+      "loss": 0.762,
+      "step": 2632
+    },
+    {
+      "epoch": 2.9126106194690267,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.718473451327434e-06,
+      "loss": 0.7591,
+      "step": 2633
+    },
+    {
+      "epoch": 2.913716814159292,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.7157079646017705e-06,
+      "loss": 0.766,
+      "step": 2634
+    },
+    {
+      "epoch": 2.9148230088495577,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.7129424778761066e-06,
+      "loss": 0.7912,
+      "step": 2635
+    },
+    {
+      "epoch": 2.915929203539823,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.7101769911504428e-06,
+      "loss": 0.8579,
+      "step": 2636
+    },
+    {
+      "epoch": 2.9170353982300883,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.707411504424779e-06,
+      "loss": 0.7368,
+      "step": 2637
+    },
+    {
+      "epoch": 2.918141592920354,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.704646017699115e-06,
+      "loss": 0.7491,
+      "step": 2638
+    },
+    {
+      "epoch": 2.9192477876106193,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.7018805309734516e-06,
+      "loss": 0.8138,
+      "step": 2639
+    },
+    {
+      "epoch": 2.920353982300885,
+      "grad_norm": 0.119140625,
+      "learning_rate": 2.6991150442477877e-06,
+      "loss": 0.7566,
+      "step": 2640
+    },
+    {
+      "epoch": 2.9214601769911503,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.6963495575221242e-06,
+      "loss": 0.7863,
+      "step": 2641
+    },
+    {
+      "epoch": 2.922566371681416,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.6935840707964604e-06,
+      "loss": 0.8069,
+      "step": 2642
+    },
+    {
+      "epoch": 2.9236725663716814,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.690818584070797e-06,
+      "loss": 0.7207,
+      "step": 2643
+    },
+    {
+      "epoch": 2.9247787610619467,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.688053097345133e-06,
+      "loss": 0.7555,
+      "step": 2644
+    },
+    {
+      "epoch": 2.9258849557522124,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.685287610619469e-06,
+      "loss": 0.7643,
+      "step": 2645
+    },
+    {
+      "epoch": 2.926991150442478,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.6825221238938053e-06,
+      "loss": 0.7705,
+      "step": 2646
+    },
+    {
+      "epoch": 2.9280973451327434,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.6797566371681414e-06,
+      "loss": 0.84,
+      "step": 2647
+    },
+    {
+      "epoch": 2.9292035398230087,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.6769911504424784e-06,
+      "loss": 0.7756,
+      "step": 2648
+    },
+    {
+      "epoch": 2.9303097345132745,
+      "grad_norm": 0.1015625,
+      "learning_rate": 2.6742256637168145e-06,
+      "loss": 0.7423,
+      "step": 2649
+    },
+    {
+      "epoch": 2.9314159292035398,
+      "grad_norm": 0.125,
+      "learning_rate": 2.6714601769911507e-06,
+      "loss": 0.7659,
+      "step": 2650
+    },
+    {
+      "epoch": 2.932522123893805,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.668694690265487e-06,
+      "loss": 0.7525,
+      "step": 2651
+    },
+    {
+      "epoch": 2.933628318584071,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.6659292035398233e-06,
+      "loss": 0.7784,
+      "step": 2652
+    },
+    {
+      "epoch": 2.9347345132743365,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.6631637168141595e-06,
+      "loss": 0.7236,
+      "step": 2653
+    },
+    {
+      "epoch": 2.935840707964602,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.6603982300884956e-06,
+      "loss": 0.8299,
+      "step": 2654
+    },
+    {
+      "epoch": 2.936946902654867,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.6576327433628317e-06,
+      "loss": 0.7475,
+      "step": 2655
+    },
+    {
+      "epoch": 2.938053097345133,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.6548672566371687e-06,
+      "loss": 0.7602,
+      "step": 2656
+    },
+    {
+      "epoch": 2.939159292035398,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.652101769911505e-06,
+      "loss": 0.7974,
+      "step": 2657
+    },
+    {
+      "epoch": 2.9402654867256635,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.649336283185841e-06,
+      "loss": 0.7683,
+      "step": 2658
+    },
+    {
+      "epoch": 2.941371681415929,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.646570796460177e-06,
+      "loss": 0.7651,
+      "step": 2659
+    },
+    {
+      "epoch": 2.942477876106195,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.6438053097345136e-06,
+      "loss": 0.7807,
+      "step": 2660
+    },
+    {
+      "epoch": 2.9435840707964602,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.6410398230088497e-06,
+      "loss": 0.7892,
+      "step": 2661
+    },
+    {
+      "epoch": 2.9446902654867255,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.638274336283186e-06,
+      "loss": 0.7492,
+      "step": 2662
+    },
+    {
+      "epoch": 2.9457964601769913,
+      "grad_norm": 0.1015625,
+      "learning_rate": 2.635508849557522e-06,
+      "loss": 0.7247,
+      "step": 2663
+    },
+    {
+      "epoch": 2.9469026548672566,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.632743362831859e-06,
+      "loss": 0.7487,
+      "step": 2664
+    },
+    {
+      "epoch": 2.948008849557522,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.629977876106195e-06,
+      "loss": 0.7853,
+      "step": 2665
+    },
+    {
+      "epoch": 2.9491150442477876,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.627212389380531e-06,
+      "loss": 0.8029,
+      "step": 2666
+    },
+    {
+      "epoch": 2.9502212389380533,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.6244469026548673e-06,
+      "loss": 0.7528,
+      "step": 2667
+    },
+    {
+      "epoch": 2.9513274336283186,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.6216814159292035e-06,
+      "loss": 0.744,
+      "step": 2668
+    },
+    {
+      "epoch": 2.952433628318584,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.61891592920354e-06,
+      "loss": 0.7763,
+      "step": 2669
+    },
+    {
+      "epoch": 2.9535398230088497,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.616150442477876e-06,
+      "loss": 0.7355,
+      "step": 2670
+    },
+    {
+      "epoch": 2.954646017699115,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.6133849557522127e-06,
+      "loss": 0.7574,
+      "step": 2671
+    },
+    {
+      "epoch": 2.9557522123893807,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.6106194690265492e-06,
+      "loss": 0.7428,
+      "step": 2672
+    },
+    {
+      "epoch": 2.956858407079646,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.6078539823008854e-06,
+      "loss": 0.7437,
+      "step": 2673
+    },
+    {
+      "epoch": 2.9579646017699117,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.6050884955752215e-06,
+      "loss": 0.796,
+      "step": 2674
+    },
+    {
+      "epoch": 2.959070796460177,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.6023230088495576e-06,
+      "loss": 0.8005,
+      "step": 2675
+    },
+    {
+      "epoch": 2.9601769911504423,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.5995575221238937e-06,
+      "loss": 0.7542,
+      "step": 2676
+    },
+    {
+      "epoch": 2.961283185840708,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.5967920353982303e-06,
+      "loss": 0.7926,
+      "step": 2677
+    },
+    {
+      "epoch": 2.9623893805309733,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.5940265486725664e-06,
+      "loss": 0.772,
+      "step": 2678
+    },
+    {
+      "epoch": 2.963495575221239,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.591261061946903e-06,
+      "loss": 0.7729,
+      "step": 2679
+    },
+    {
+      "epoch": 2.9646017699115044,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.588495575221239e-06,
+      "loss": 0.8041,
+      "step": 2680
+    },
+    {
+      "epoch": 2.96570796460177,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.5857300884955756e-06,
+      "loss": 0.7746,
+      "step": 2681
+    },
+    {
+      "epoch": 2.9668141592920354,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.5829646017699118e-06,
+      "loss": 0.7561,
+      "step": 2682
+    },
+    {
+      "epoch": 2.9679203539823007,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.580199115044248e-06,
+      "loss": 0.7789,
+      "step": 2683
+    },
+    {
+      "epoch": 2.9690265486725664,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.577433628318584e-06,
+      "loss": 0.7476,
+      "step": 2684
+    },
+    {
+      "epoch": 2.9701327433628317,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.57466814159292e-06,
+      "loss": 0.7733,
+      "step": 2685
+    },
+    {
+      "epoch": 2.9712389380530975,
+      "grad_norm": 0.103515625,
+      "learning_rate": 2.571902654867257e-06,
+      "loss": 0.7434,
+      "step": 2686
+    },
+    {
+      "epoch": 2.9723451327433628,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.5691371681415932e-06,
+      "loss": 0.8044,
+      "step": 2687
+    },
+    {
+      "epoch": 2.9734513274336285,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.5663716814159294e-06,
+      "loss": 0.783,
+      "step": 2688
+    },
+    {
+      "epoch": 2.974557522123894,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 2.563606194690266e-06,
+      "loss": 0.7533,
+      "step": 2689
+    },
+    {
+      "epoch": 2.975663716814159,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.560840707964602e-06,
+      "loss": 0.7351,
+      "step": 2690
+    },
+    {
+      "epoch": 2.976769911504425,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.558075221238938e-06,
+      "loss": 0.75,
+      "step": 2691
+    },
+    {
+      "epoch": 2.97787610619469,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.5553097345132743e-06,
+      "loss": 0.7659,
+      "step": 2692
+    },
+    {
+      "epoch": 2.978982300884956,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.5525442477876104e-06,
+      "loss": 0.7907,
+      "step": 2693
+    },
+    {
+      "epoch": 2.980088495575221,
+      "grad_norm": 0.115234375,
+      "learning_rate": 2.5497787610619474e-06,
+      "loss": 0.7494,
+      "step": 2694
+    },
+    {
+      "epoch": 2.981194690265487,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.5470132743362835e-06,
+      "loss": 0.7548,
+      "step": 2695
+    },
+    {
+      "epoch": 2.982300884955752,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 2.5442477876106196e-06,
+      "loss": 0.7395,
+      "step": 2696
+    },
+    {
+      "epoch": 2.9834070796460175,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.5414823008849558e-06,
+      "loss": 0.783,
+      "step": 2697
+    },
+    {
+      "epoch": 2.984513274336283,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.5387168141592923e-06,
+      "loss": 0.7398,
+      "step": 2698
+    },
+    {
+      "epoch": 2.9856194690265485,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.5359513274336284e-06,
+      "loss": 0.8038,
+      "step": 2699
+    },
+    {
+      "epoch": 2.9867256637168142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.5331858407079646e-06,
+      "loss": 0.7897,
+      "step": 2700
+    },
+    {
+      "epoch": 2.9878318584070795,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.5304203539823015e-06,
+      "loss": 0.7616,
+      "step": 2701
+    },
+    {
+      "epoch": 2.9889380530973453,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.5276548672566377e-06,
+      "loss": 0.7309,
+      "step": 2702
+    },
+    {
+      "epoch": 2.9900442477876106,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.524889380530974e-06,
+      "loss": 0.7755,
+      "step": 2703
+    },
+    {
+      "epoch": 2.991150442477876,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.52212389380531e-06,
+      "loss": 0.7615,
+      "step": 2704
+    },
+    {
+      "epoch": 2.9922566371681416,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.519358407079646e-06,
+      "loss": 0.7519,
+      "step": 2705
+    },
+    {
+      "epoch": 2.9933628318584073,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.5165929203539826e-06,
+      "loss": 0.7537,
+      "step": 2706
+    },
+    {
+      "epoch": 2.9944690265486726,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.5138274336283187e-06,
+      "loss": 0.7562,
+      "step": 2707
+    },
+    {
+      "epoch": 2.995575221238938,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.511061946902655e-06,
+      "loss": 0.7406,
+      "step": 2708
+    },
+    {
+      "epoch": 2.9966814159292037,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.5082964601769914e-06,
+      "loss": 0.7834,
+      "step": 2709
+    },
+    {
+      "epoch": 2.997787610619469,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.505530973451328e-06,
+      "loss": 0.7617,
+      "step": 2710
+    },
+    {
+      "epoch": 2.9988938053097343,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.502765486725664e-06,
+      "loss": 0.7631,
+      "step": 2711
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8323,
+      "step": 2712
+    },
+    {
+      "epoch": 3.0011061946902653,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 2.4972345132743363e-06,
+      "loss": 0.7353,
+      "step": 2713
+    },
+    {
+      "epoch": 3.002212389380531,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.494469026548673e-06,
+      "loss": 0.7365,
+      "step": 2714
+    },
+    {
+      "epoch": 3.0033185840707963,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.491703539823009e-06,
+      "loss": 0.8122,
+      "step": 2715
+    },
+    {
+      "epoch": 3.004424778761062,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.488938053097345e-06,
+      "loss": 0.7858,
+      "step": 2716
+    },
+    {
+      "epoch": 3.0055309734513274,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.4861725663716817e-06,
+      "loss": 0.8404,
+      "step": 2717
+    },
+    {
+      "epoch": 3.006637168141593,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.483407079646018e-06,
+      "loss": 0.7802,
+      "step": 2718
+    },
+    {
+      "epoch": 3.0077433628318584,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.4806415929203544e-06,
+      "loss": 0.7585,
+      "step": 2719
+    },
+    {
+      "epoch": 3.0088495575221237,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.4778761061946905e-06,
+      "loss": 0.7818,
+      "step": 2720
+    },
+    {
+      "epoch": 3.0099557522123894,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.4751106194690266e-06,
+      "loss": 0.7688,
+      "step": 2721
+    },
+    {
+      "epoch": 3.0110619469026547,
+      "grad_norm": 0.115234375,
+      "learning_rate": 2.472345132743363e-06,
+      "loss": 0.7416,
+      "step": 2722
+    },
+    {
+      "epoch": 3.0121681415929205,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.4695796460176993e-06,
+      "loss": 0.7626,
+      "step": 2723
+    },
+    {
+      "epoch": 3.0132743362831858,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.4668141592920354e-06,
+      "loss": 0.7587,
+      "step": 2724
+    },
+    {
+      "epoch": 3.0143805309734515,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.464048672566372e-06,
+      "loss": 0.7763,
+      "step": 2725
+    },
+    {
+      "epoch": 3.015486725663717,
+      "grad_norm": 0.103515625,
+      "learning_rate": 2.461283185840708e-06,
+      "loss": 0.7405,
+      "step": 2726
+    },
+    {
+      "epoch": 3.0165929203539825,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.4585176991150446e-06,
+      "loss": 0.7701,
+      "step": 2727
+    },
+    {
+      "epoch": 3.017699115044248,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.4557522123893808e-06,
+      "loss": 0.7994,
+      "step": 2728
+    },
+    {
+      "epoch": 3.018805309734513,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.4529867256637173e-06,
+      "loss": 0.7854,
+      "step": 2729
+    },
+    {
+      "epoch": 3.019911504424779,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.4502212389380534e-06,
+      "loss": 0.7875,
+      "step": 2730
+    },
+    {
+      "epoch": 3.021017699115044,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.4474557522123896e-06,
+      "loss": 0.7883,
+      "step": 2731
+    },
+    {
+      "epoch": 3.02212389380531,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.4446902654867257e-06,
+      "loss": 0.7275,
+      "step": 2732
+    },
+    {
+      "epoch": 3.023230088495575,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 2.4419247787610622e-06,
+      "loss": 0.7312,
+      "step": 2733
+    },
+    {
+      "epoch": 3.024336283185841,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.4391592920353984e-06,
+      "loss": 0.7491,
+      "step": 2734
+    },
+    {
+      "epoch": 3.025442477876106,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.4363938053097345e-06,
+      "loss": 0.7622,
+      "step": 2735
+    },
+    {
+      "epoch": 3.0265486725663715,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.433628318584071e-06,
+      "loss": 0.7987,
+      "step": 2736
+    },
+    {
+      "epoch": 3.0276548672566372,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 2.4308628318584076e-06,
+      "loss": 0.7522,
+      "step": 2737
+    },
+    {
+      "epoch": 3.0287610619469025,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.4280973451327437e-06,
+      "loss": 0.7829,
+      "step": 2738
+    },
+    {
+      "epoch": 3.0298672566371683,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.42533185840708e-06,
+      "loss": 0.7701,
+      "step": 2739
+    },
+    {
+      "epoch": 3.0309734513274336,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.422566371681416e-06,
+      "loss": 0.7666,
+      "step": 2740
+    },
+    {
+      "epoch": 3.0320796460176993,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.4198008849557525e-06,
+      "loss": 0.7525,
+      "step": 2741
+    },
+    {
+      "epoch": 3.0331858407079646,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.4170353982300886e-06,
+      "loss": 0.7576,
+      "step": 2742
+    },
+    {
+      "epoch": 3.03429203539823,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.4142699115044248e-06,
+      "loss": 0.7514,
+      "step": 2743
+    },
+    {
+      "epoch": 3.0353982300884956,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.4115044247787613e-06,
+      "loss": 0.7925,
+      "step": 2744
+    },
+    {
+      "epoch": 3.036504424778761,
+      "grad_norm": 0.103515625,
+      "learning_rate": 2.4087389380530974e-06,
+      "loss": 0.7516,
+      "step": 2745
+    },
+    {
+      "epoch": 3.0376106194690267,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.405973451327434e-06,
+      "loss": 0.7799,
+      "step": 2746
+    },
+    {
+      "epoch": 3.038716814159292,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 2.40320796460177e-06,
+      "loss": 0.721,
+      "step": 2747
+    },
+    {
+      "epoch": 3.0398230088495577,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 2.4004424778761067e-06,
+      "loss": 0.7176,
+      "step": 2748
+    },
+    {
+      "epoch": 3.040929203539823,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.397676991150443e-06,
+      "loss": 0.8004,
+      "step": 2749
+    },
+    {
+      "epoch": 3.0420353982300883,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.394911504424779e-06,
+      "loss": 0.7476,
+      "step": 2750
+    },
+    {
+      "epoch": 3.043141592920354,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.392146017699115e-06,
+      "loss": 0.8049,
+      "step": 2751
+    },
+    {
+      "epoch": 3.0442477876106193,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.3893805309734516e-06,
+      "loss": 0.7859,
+      "step": 2752
+    },
+    {
+      "epoch": 3.045353982300885,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.3866150442477877e-06,
+      "loss": 0.7638,
+      "step": 2753
+    },
+    {
+      "epoch": 3.0464601769911503,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.383849557522124e-06,
+      "loss": 0.759,
+      "step": 2754
+    },
+    {
+      "epoch": 3.047566371681416,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.3810840707964604e-06,
+      "loss": 0.8131,
+      "step": 2755
+    },
+    {
+      "epoch": 3.0486725663716814,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.378318584070797e-06,
+      "loss": 0.7922,
+      "step": 2756
+    },
+    {
+      "epoch": 3.049778761061947,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.375553097345133e-06,
+      "loss": 0.7695,
+      "step": 2757
+    },
+    {
+      "epoch": 3.0508849557522124,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.372787610619469e-06,
+      "loss": 0.7616,
+      "step": 2758
+    },
+    {
+      "epoch": 3.0519911504424777,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.3700221238938053e-06,
+      "loss": 0.7728,
+      "step": 2759
+    },
+    {
+      "epoch": 3.0530973451327434,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.367256637168142e-06,
+      "loss": 0.7438,
+      "step": 2760
+    },
+    {
+      "epoch": 3.0542035398230087,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.364491150442478e-06,
+      "loss": 0.8006,
+      "step": 2761
+    },
+    {
+      "epoch": 3.0553097345132745,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.361725663716814e-06,
+      "loss": 0.8048,
+      "step": 2762
+    },
+    {
+      "epoch": 3.0564159292035398,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.3589601769911507e-06,
+      "loss": 0.7804,
+      "step": 2763
+    },
+    {
+      "epoch": 3.0575221238938055,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 2.356194690265487e-06,
+      "loss": 0.768,
+      "step": 2764
+    },
+    {
+      "epoch": 3.058628318584071,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.3534292035398234e-06,
+      "loss": 0.7316,
+      "step": 2765
+    },
+    {
+      "epoch": 3.059734513274336,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.3506637168141595e-06,
+      "loss": 0.7479,
+      "step": 2766
+    },
+    {
+      "epoch": 3.060840707964602,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.347898230088496e-06,
+      "loss": 0.8223,
+      "step": 2767
+    },
+    {
+      "epoch": 3.061946902654867,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.345132743362832e-06,
+      "loss": 0.8086,
+      "step": 2768
+    },
+    {
+      "epoch": 3.063053097345133,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.3423672566371683e-06,
+      "loss": 0.7698,
+      "step": 2769
+    },
+    {
+      "epoch": 3.064159292035398,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.3396017699115044e-06,
+      "loss": 0.7939,
+      "step": 2770
+    },
+    {
+      "epoch": 3.065265486725664,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.336836283185841e-06,
+      "loss": 0.8147,
+      "step": 2771
+    },
+    {
+      "epoch": 3.066371681415929,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.334070796460177e-06,
+      "loss": 0.7843,
+      "step": 2772
+    },
+    {
+      "epoch": 3.0674778761061945,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.331305309734513e-06,
+      "loss": 0.7709,
+      "step": 2773
+    },
+    {
+      "epoch": 3.0685840707964602,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.3285398230088498e-06,
+      "loss": 0.7551,
+      "step": 2774
+    },
+    {
+      "epoch": 3.0696902654867255,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.3257743362831863e-06,
+      "loss": 0.7861,
+      "step": 2775
+    },
+    {
+      "epoch": 3.0707964601769913,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.3230088495575224e-06,
+      "loss": 0.781,
+      "step": 2776
+    },
+    {
+      "epoch": 3.0719026548672566,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.3202433628318586e-06,
+      "loss": 0.7659,
+      "step": 2777
+    },
+    {
+      "epoch": 3.0730088495575223,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.317477876106195e-06,
+      "loss": 0.811,
+      "step": 2778
+    },
+    {
+      "epoch": 3.0741150442477876,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.3147123893805312e-06,
+      "loss": 0.7508,
+      "step": 2779
+    },
+    {
+      "epoch": 3.0752212389380533,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.3119469026548674e-06,
+      "loss": 0.7642,
+      "step": 2780
+    },
+    {
+      "epoch": 3.0763274336283186,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.3091814159292035e-06,
+      "loss": 0.7858,
+      "step": 2781
+    },
+    {
+      "epoch": 3.077433628318584,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.30641592920354e-06,
+      "loss": 0.8014,
+      "step": 2782
+    },
+    {
+      "epoch": 3.0785398230088497,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.303650442477876e-06,
+      "loss": 0.8055,
+      "step": 2783
+    },
+    {
+      "epoch": 3.079646017699115,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.3008849557522127e-06,
+      "loss": 0.8464,
+      "step": 2784
+    },
+    {
+      "epoch": 3.0807522123893807,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.298119469026549e-06,
+      "loss": 0.7589,
+      "step": 2785
+    },
+    {
+      "epoch": 3.081858407079646,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.2953539823008854e-06,
+      "loss": 0.8282,
+      "step": 2786
+    },
+    {
+      "epoch": 3.0829646017699117,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.2925884955752215e-06,
+      "loss": 0.7437,
+      "step": 2787
+    },
+    {
+      "epoch": 3.084070796460177,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.2898230088495576e-06,
+      "loss": 0.7701,
+      "step": 2788
+    },
+    {
+      "epoch": 3.0851769911504423,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.2870575221238938e-06,
+      "loss": 0.7392,
+      "step": 2789
+    },
+    {
+      "epoch": 3.086283185840708,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.2842920353982303e-06,
+      "loss": 0.8021,
+      "step": 2790
+    },
+    {
+      "epoch": 3.0873893805309733,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.2815265486725664e-06,
+      "loss": 0.7547,
+      "step": 2791
+    },
+    {
+      "epoch": 3.088495575221239,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.2787610619469026e-06,
+      "loss": 0.7846,
+      "step": 2792
+    },
+    {
+      "epoch": 3.0896017699115044,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.275995575221239e-06,
+      "loss": 0.7786,
+      "step": 2793
+    },
+    {
+      "epoch": 3.09070796460177,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.2732300884955757e-06,
+      "loss": 0.7572,
+      "step": 2794
+    },
+    {
+      "epoch": 3.0918141592920354,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.270464601769912e-06,
+      "loss": 0.7836,
+      "step": 2795
+    },
+    {
+      "epoch": 3.0929203539823007,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.267699115044248e-06,
+      "loss": 0.8459,
+      "step": 2796
+    },
+    {
+      "epoch": 3.0940265486725664,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.2649336283185845e-06,
+      "loss": 0.7843,
+      "step": 2797
+    },
+    {
+      "epoch": 3.0951327433628317,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.2621681415929206e-06,
+      "loss": 0.7609,
+      "step": 2798
+    },
+    {
+      "epoch": 3.0962389380530975,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.2594026548672567e-06,
+      "loss": 0.7579,
+      "step": 2799
+    },
+    {
+      "epoch": 3.0973451327433628,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.256637168141593e-06,
+      "loss": 0.7803,
+      "step": 2800
+    },
+    {
+      "epoch": 3.0984513274336285,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.2538716814159294e-06,
+      "loss": 0.7708,
+      "step": 2801
+    },
+    {
+      "epoch": 3.099557522123894,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.2511061946902655e-06,
+      "loss": 0.7491,
+      "step": 2802
+    },
+    {
+      "epoch": 3.100663716814159,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.248340707964602e-06,
+      "loss": 0.739,
+      "step": 2803
+    },
+    {
+      "epoch": 3.101769911504425,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.245575221238938e-06,
+      "loss": 0.7812,
+      "step": 2804
+    },
+    {
+      "epoch": 3.10287610619469,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.2428097345132747e-06,
+      "loss": 0.7738,
+      "step": 2805
+    },
+    {
+      "epoch": 3.103982300884956,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.240044247787611e-06,
+      "loss": 0.7838,
+      "step": 2806
+    },
+    {
+      "epoch": 3.105088495575221,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.237278761061947e-06,
+      "loss": 0.7915,
+      "step": 2807
+    },
+    {
+      "epoch": 3.106194690265487,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.234513274336283e-06,
+      "loss": 0.7902,
+      "step": 2808
+    },
+    {
+      "epoch": 3.107300884955752,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.2317477876106197e-06,
+      "loss": 0.734,
+      "step": 2809
+    },
+    {
+      "epoch": 3.1084070796460175,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 2.228982300884956e-06,
+      "loss": 0.7441,
+      "step": 2810
+    },
+    {
+      "epoch": 3.109513274336283,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.2262168141592924e-06,
+      "loss": 0.7977,
+      "step": 2811
+    },
+    {
+      "epoch": 3.1106194690265485,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.2234513274336285e-06,
+      "loss": 0.7591,
+      "step": 2812
+    },
+    {
+      "epoch": 3.1117256637168142,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.220685840707965e-06,
+      "loss": 0.7634,
+      "step": 2813
+    },
+    {
+      "epoch": 3.1128318584070795,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.217920353982301e-06,
+      "loss": 0.7796,
+      "step": 2814
+    },
+    {
+      "epoch": 3.1139380530973453,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.2151548672566373e-06,
+      "loss": 0.8021,
+      "step": 2815
+    },
+    {
+      "epoch": 3.1150442477876106,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.212389380530974e-06,
+      "loss": 0.781,
+      "step": 2816
+    },
+    {
+      "epoch": 3.1161504424778763,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.20962389380531e-06,
+      "loss": 0.7094,
+      "step": 2817
+    },
+    {
+      "epoch": 3.1172566371681416,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 2.206858407079646e-06,
+      "loss": 0.7593,
+      "step": 2818
+    },
+    {
+      "epoch": 3.118362831858407,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.204092920353982e-06,
+      "loss": 0.7821,
+      "step": 2819
+    },
+    {
+      "epoch": 3.1194690265486726,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.2013274336283188e-06,
+      "loss": 0.7766,
+      "step": 2820
+    },
+    {
+      "epoch": 3.120575221238938,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.198561946902655e-06,
+      "loss": 0.776,
+      "step": 2821
+    },
+    {
+      "epoch": 3.1216814159292037,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.1957964601769914e-06,
+      "loss": 0.784,
+      "step": 2822
+    },
+    {
+      "epoch": 3.122787610619469,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 2.1930309734513276e-06,
+      "loss": 0.8007,
+      "step": 2823
+    },
+    {
+      "epoch": 3.1238938053097347,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.190265486725664e-06,
+      "loss": 0.8002,
+      "step": 2824
+    },
+    {
+      "epoch": 3.125,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.1875000000000002e-06,
+      "loss": 0.7038,
+      "step": 2825
+    },
+    {
+      "epoch": 3.1261061946902653,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.1847345132743364e-06,
+      "loss": 0.7475,
+      "step": 2826
+    },
+    {
+      "epoch": 3.127212389380531,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.181969026548673e-06,
+      "loss": 0.8141,
+      "step": 2827
+    },
+    {
+      "epoch": 3.1283185840707963,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.179203539823009e-06,
+      "loss": 0.7839,
+      "step": 2828
+    },
+    {
+      "epoch": 3.129424778761062,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.176438053097345e-06,
+      "loss": 0.7681,
+      "step": 2829
+    },
+    {
+      "epoch": 3.1305309734513274,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.1736725663716817e-06,
+      "loss": 0.7944,
+      "step": 2830
+    },
+    {
+      "epoch": 3.131637168141593,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.170907079646018e-06,
+      "loss": 0.7803,
+      "step": 2831
+    },
+    {
+      "epoch": 3.1327433628318584,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.1681415929203544e-06,
+      "loss": 0.7756,
+      "step": 2832
+    },
+    {
+      "epoch": 3.133849557522124,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.1653761061946905e-06,
+      "loss": 0.7811,
+      "step": 2833
+    },
+    {
+      "epoch": 3.1349557522123894,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.1626106194690266e-06,
+      "loss": 0.7733,
+      "step": 2834
+    },
+    {
+      "epoch": 3.1360619469026547,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.159845132743363e-06,
+      "loss": 0.7738,
+      "step": 2835
+    },
+    {
+      "epoch": 3.1371681415929205,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.1570796460176993e-06,
+      "loss": 0.7946,
+      "step": 2836
+    },
+    {
+      "epoch": 3.1382743362831858,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.1543141592920354e-06,
+      "loss": 0.7766,
+      "step": 2837
+    },
+    {
+      "epoch": 3.1393805309734515,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.1515486725663716e-06,
+      "loss": 0.7893,
+      "step": 2838
+    },
+    {
+      "epoch": 3.140486725663717,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.148783185840708e-06,
+      "loss": 0.7826,
+      "step": 2839
+    },
+    {
+      "epoch": 3.1415929203539825,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.1460176991150442e-06,
+      "loss": 0.7852,
+      "step": 2840
+    },
+    {
+      "epoch": 3.142699115044248,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.143252212389381e-06,
+      "loss": 0.7772,
+      "step": 2841
+    },
+    {
+      "epoch": 3.143805309734513,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.140486725663717e-06,
+      "loss": 0.8138,
+      "step": 2842
+    },
+    {
+      "epoch": 3.144911504424779,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.1377212389380535e-06,
+      "loss": 0.7907,
+      "step": 2843
+    },
+    {
+      "epoch": 3.146017699115044,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.1349557522123896e-06,
+      "loss": 0.8189,
+      "step": 2844
+    },
+    {
+      "epoch": 3.14712389380531,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.1321902654867257e-06,
+      "loss": 0.7855,
+      "step": 2845
+    },
+    {
+      "epoch": 3.148230088495575,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.1294247787610623e-06,
+      "loss": 0.7913,
+      "step": 2846
+    },
+    {
+      "epoch": 3.149336283185841,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.1266592920353984e-06,
+      "loss": 0.7784,
+      "step": 2847
+    },
+    {
+      "epoch": 3.150442477876106,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 2.1238938053097345e-06,
+      "loss": 0.7606,
+      "step": 2848
+    },
+    {
+      "epoch": 3.1515486725663715,
+      "grad_norm": 0.125,
+      "learning_rate": 2.121128318584071e-06,
+      "loss": 0.8237,
+      "step": 2849
+    },
+    {
+      "epoch": 3.1526548672566372,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.118362831858407e-06,
+      "loss": 0.7448,
+      "step": 2850
+    },
+    {
+      "epoch": 3.1537610619469025,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.1155973451327437e-06,
+      "loss": 0.8151,
+      "step": 2851
+    },
+    {
+      "epoch": 3.1548672566371683,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.11283185840708e-06,
+      "loss": 0.7845,
+      "step": 2852
+    },
+    {
+      "epoch": 3.1559734513274336,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.110066371681416e-06,
+      "loss": 0.7629,
+      "step": 2853
+    },
+    {
+      "epoch": 3.1570796460176993,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.1073008849557525e-06,
+      "loss": 0.8094,
+      "step": 2854
+    },
+    {
+      "epoch": 3.1581858407079646,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.1045353982300887e-06,
+      "loss": 0.8006,
+      "step": 2855
+    },
+    {
+      "epoch": 3.15929203539823,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.101769911504425e-06,
+      "loss": 0.773,
+      "step": 2856
+    },
+    {
+      "epoch": 3.1603982300884956,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.099004424778761e-06,
+      "loss": 0.737,
+      "step": 2857
+    },
+    {
+      "epoch": 3.161504424778761,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.0962389380530975e-06,
+      "loss": 0.7619,
+      "step": 2858
+    },
+    {
+      "epoch": 3.1626106194690267,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.0934734513274336e-06,
+      "loss": 0.7561,
+      "step": 2859
+    },
+    {
+      "epoch": 3.163716814159292,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.09070796460177e-06,
+      "loss": 0.737,
+      "step": 2860
+    },
+    {
+      "epoch": 3.1648230088495577,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.0879424778761067e-06,
+      "loss": 0.7534,
+      "step": 2861
+    },
+    {
+      "epoch": 3.165929203539823,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.085176991150443e-06,
+      "loss": 0.8883,
+      "step": 2862
+    },
+    {
+      "epoch": 3.1670353982300883,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.082411504424779e-06,
+      "loss": 0.7557,
+      "step": 2863
+    },
+    {
+      "epoch": 3.168141592920354,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.079646017699115e-06,
+      "loss": 0.7934,
+      "step": 2864
+    },
+    {
+      "epoch": 3.1692477876106193,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.0768805309734516e-06,
+      "loss": 0.767,
+      "step": 2865
+    },
+    {
+      "epoch": 3.170353982300885,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.0741150442477878e-06,
+      "loss": 0.7835,
+      "step": 2866
+    },
+    {
+      "epoch": 3.1714601769911503,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.071349557522124e-06,
+      "loss": 0.8144,
+      "step": 2867
+    },
+    {
+      "epoch": 3.172566371681416,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.0685840707964604e-06,
+      "loss": 0.7783,
+      "step": 2868
+    },
+    {
+      "epoch": 3.1736725663716814,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.0658185840707966e-06,
+      "loss": 0.8025,
+      "step": 2869
+    },
+    {
+      "epoch": 3.1747787610619467,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.063053097345133e-06,
+      "loss": 0.7893,
+      "step": 2870
+    },
+    {
+      "epoch": 3.1758849557522124,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.0602876106194692e-06,
+      "loss": 0.7664,
+      "step": 2871
+    },
+    {
+      "epoch": 3.1769911504424777,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.0575221238938054e-06,
+      "loss": 0.717,
+      "step": 2872
+    },
+    {
+      "epoch": 3.1780973451327434,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 2.054756637168142e-06,
+      "loss": 0.7099,
+      "step": 2873
+    },
+    {
+      "epoch": 3.1792035398230087,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.051991150442478e-06,
+      "loss": 0.7594,
+      "step": 2874
+    },
+    {
+      "epoch": 3.1803097345132745,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.049225663716814e-06,
+      "loss": 0.7685,
+      "step": 2875
+    },
+    {
+      "epoch": 3.1814159292035398,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.0464601769911507e-06,
+      "loss": 0.7627,
+      "step": 2876
+    },
+    {
+      "epoch": 3.1825221238938055,
+      "grad_norm": 0.125,
+      "learning_rate": 2.043694690265487e-06,
+      "loss": 0.8107,
+      "step": 2877
+    },
+    {
+      "epoch": 3.183628318584071,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.040929203539823e-06,
+      "loss": 0.7667,
+      "step": 2878
+    },
+    {
+      "epoch": 3.184734513274336,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.0381637168141595e-06,
+      "loss": 0.809,
+      "step": 2879
+    },
+    {
+      "epoch": 3.185840707964602,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.035398230088496e-06,
+      "loss": 0.7528,
+      "step": 2880
+    },
+    {
+      "epoch": 3.186946902654867,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.032632743362832e-06,
+      "loss": 0.7706,
+      "step": 2881
+    },
+    {
+      "epoch": 3.188053097345133,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.0298672566371683e-06,
+      "loss": 0.764,
+      "step": 2882
+    },
+    {
+      "epoch": 3.189159292035398,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.0271017699115044e-06,
+      "loss": 0.789,
+      "step": 2883
+    },
+    {
+      "epoch": 3.190265486725664,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 2.024336283185841e-06,
+      "loss": 0.7503,
+      "step": 2884
+    },
+    {
+      "epoch": 3.191371681415929,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.021570796460177e-06,
+      "loss": 0.7369,
+      "step": 2885
+    },
+    {
+      "epoch": 3.1924778761061945,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.0188053097345132e-06,
+      "loss": 0.7999,
+      "step": 2886
+    },
+    {
+      "epoch": 3.1935840707964602,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.0160398230088498e-06,
+      "loss": 0.7879,
+      "step": 2887
+    },
+    {
+      "epoch": 3.1946902654867255,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.013274336283186e-06,
+      "loss": 0.7614,
+      "step": 2888
+    },
+    {
+      "epoch": 3.1957964601769913,
+      "grad_norm": 0.115234375,
+      "learning_rate": 2.0105088495575225e-06,
+      "loss": 0.7465,
+      "step": 2889
+    },
+    {
+      "epoch": 3.1969026548672566,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.0077433628318586e-06,
+      "loss": 0.7801,
+      "step": 2890
+    },
+    {
+      "epoch": 3.1980088495575223,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.004977876106195e-06,
+      "loss": 0.803,
+      "step": 2891
+    },
+    {
+      "epoch": 3.1991150442477876,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.0022123893805313e-06,
+      "loss": 0.7847,
+      "step": 2892
+    },
+    {
+      "epoch": 3.2002212389380533,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.9994469026548674e-06,
+      "loss": 0.786,
+      "step": 2893
+    },
+    {
+      "epoch": 3.2013274336283186,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.9966814159292035e-06,
+      "loss": 0.8162,
+      "step": 2894
+    },
+    {
+      "epoch": 3.202433628318584,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.99391592920354e-06,
+      "loss": 0.7222,
+      "step": 2895
+    },
+    {
+      "epoch": 3.2035398230088497,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.991150442477876e-06,
+      "loss": 0.7881,
+      "step": 2896
+    },
+    {
+      "epoch": 3.204646017699115,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.9883849557522123e-06,
+      "loss": 0.7555,
+      "step": 2897
+    },
+    {
+      "epoch": 3.2057522123893807,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.985619469026549e-06,
+      "loss": 0.761,
+      "step": 2898
+    },
+    {
+      "epoch": 3.206858407079646,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.9828539823008854e-06,
+      "loss": 0.7351,
+      "step": 2899
+    },
+    {
+      "epoch": 3.2079646017699117,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 1.9800884955752215e-06,
+      "loss": 0.7738,
+      "step": 2900
+    },
+    {
+      "epoch": 3.209070796460177,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.9773230088495577e-06,
+      "loss": 0.7453,
+      "step": 2901
+    },
+    {
+      "epoch": 3.2101769911504423,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.974557522123894e-06,
+      "loss": 0.757,
+      "step": 2902
+    },
+    {
+      "epoch": 3.211283185840708,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.9717920353982303e-06,
+      "loss": 0.7606,
+      "step": 2903
+    },
+    {
+      "epoch": 3.2123893805309733,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.9690265486725665e-06,
+      "loss": 0.7194,
+      "step": 2904
+    },
+    {
+      "epoch": 3.213495575221239,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.9662610619469026e-06,
+      "loss": 0.7648,
+      "step": 2905
+    },
+    {
+      "epoch": 3.2146017699115044,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.963495575221239e-06,
+      "loss": 0.7249,
+      "step": 2906
+    },
+    {
+      "epoch": 3.21570796460177,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.9607300884955753e-06,
+      "loss": 0.7637,
+      "step": 2907
+    },
+    {
+      "epoch": 3.2168141592920354,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.957964601769912e-06,
+      "loss": 0.7764,
+      "step": 2908
+    },
+    {
+      "epoch": 3.2179203539823007,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.955199115044248e-06,
+      "loss": 0.7902,
+      "step": 2909
+    },
+    {
+      "epoch": 3.2190265486725664,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.9524336283185845e-06,
+      "loss": 0.7839,
+      "step": 2910
+    },
+    {
+      "epoch": 3.2201327433628317,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.9496681415929206e-06,
+      "loss": 0.7768,
+      "step": 2911
+    },
+    {
+      "epoch": 3.2212389380530975,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.9469026548672567e-06,
+      "loss": 0.7338,
+      "step": 2912
+    },
+    {
+      "epoch": 3.2223451327433628,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.944137168141593e-06,
+      "loss": 0.7739,
+      "step": 2913
+    },
+    {
+      "epoch": 3.2234513274336285,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.9413716814159294e-06,
+      "loss": 0.7928,
+      "step": 2914
+    },
+    {
+      "epoch": 3.224557522123894,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.9386061946902655e-06,
+      "loss": 0.7403,
+      "step": 2915
+    },
+    {
+      "epoch": 3.225663716814159,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.9358407079646017e-06,
+      "loss": 0.78,
+      "step": 2916
+    },
+    {
+      "epoch": 3.226769911504425,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.9330752212389382e-06,
+      "loss": 0.7632,
+      "step": 2917
+    },
+    {
+      "epoch": 3.22787610619469,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.9303097345132748e-06,
+      "loss": 0.7885,
+      "step": 2918
+    },
+    {
+      "epoch": 3.228982300884956,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.927544247787611e-06,
+      "loss": 0.7794,
+      "step": 2919
+    },
+    {
+      "epoch": 3.230088495575221,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.924778761061947e-06,
+      "loss": 0.7903,
+      "step": 2920
+    },
+    {
+      "epoch": 3.231194690265487,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.922013274336283e-06,
+      "loss": 0.7684,
+      "step": 2921
+    },
+    {
+      "epoch": 3.232300884955752,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.9192477876106197e-06,
+      "loss": 0.7718,
+      "step": 2922
+    },
+    {
+      "epoch": 3.2334070796460175,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.916482300884956e-06,
+      "loss": 0.7546,
+      "step": 2923
+    },
+    {
+      "epoch": 3.234513274336283,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.913716814159292e-06,
+      "loss": 0.7633,
+      "step": 2924
+    },
+    {
+      "epoch": 3.2356194690265485,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.9109513274336285e-06,
+      "loss": 0.7436,
+      "step": 2925
+    },
+    {
+      "epoch": 3.2367256637168142,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.9081858407079646e-06,
+      "loss": 0.7871,
+      "step": 2926
+    },
+    {
+      "epoch": 3.2378318584070795,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.905420353982301e-06,
+      "loss": 0.8066,
+      "step": 2927
+    },
+    {
+      "epoch": 3.2389380530973453,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.9026548672566373e-06,
+      "loss": 0.7831,
+      "step": 2928
+    },
+    {
+      "epoch": 3.2400442477876106,
+      "grad_norm": 0.099609375,
+      "learning_rate": 1.8998893805309736e-06,
+      "loss": 0.7719,
+      "step": 2929
+    },
+    {
+      "epoch": 3.241150442477876,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.89712389380531e-06,
+      "loss": 0.7883,
+      "step": 2930
+    },
+    {
+      "epoch": 3.2422566371681416,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.8943584070796461e-06,
+      "loss": 0.8061,
+      "step": 2931
+    },
+    {
+      "epoch": 3.243362831858407,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.8915929203539824e-06,
+      "loss": 0.7544,
+      "step": 2932
+    },
+    {
+      "epoch": 3.2444690265486726,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.8888274336283188e-06,
+      "loss": 0.7397,
+      "step": 2933
+    },
+    {
+      "epoch": 3.245575221238938,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.8860619469026551e-06,
+      "loss": 0.7969,
+      "step": 2934
+    },
+    {
+      "epoch": 3.2466814159292037,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.8832964601769912e-06,
+      "loss": 0.738,
+      "step": 2935
+    },
+    {
+      "epoch": 3.247787610619469,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.8805309734513274e-06,
+      "loss": 0.7945,
+      "step": 2936
+    },
+    {
+      "epoch": 3.2488938053097347,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 1.877765486725664e-06,
+      "loss": 0.7263,
+      "step": 2937
+    },
+    {
+      "epoch": 3.25,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.7602,
+      "step": 2938
+    },
+    {
+      "epoch": 3.2511061946902653,
+      "grad_norm": 0.10546875,
+      "learning_rate": 1.8722345132743364e-06,
+      "loss": 0.7344,
+      "step": 2939
+    },
+    {
+      "epoch": 3.252212389380531,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.869469026548673e-06,
+      "loss": 0.7388,
+      "step": 2940
+    },
+    {
+      "epoch": 3.2533185840707963,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.866703539823009e-06,
+      "loss": 0.774,
+      "step": 2941
+    },
+    {
+      "epoch": 3.254424778761062,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.8639380530973452e-06,
+      "loss": 0.7454,
+      "step": 2942
+    },
+    {
+      "epoch": 3.2555309734513274,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.8611725663716815e-06,
+      "loss": 0.7708,
+      "step": 2943
+    },
+    {
+      "epoch": 3.256637168141593,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.8584070796460179e-06,
+      "loss": 0.7667,
+      "step": 2944
+    },
+    {
+      "epoch": 3.2577433628318584,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.8556415929203542e-06,
+      "loss": 0.7468,
+      "step": 2945
+    },
+    {
+      "epoch": 3.258849557522124,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.8528761061946903e-06,
+      "loss": 0.7674,
+      "step": 2946
+    },
+    {
+      "epoch": 3.2599557522123894,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.8501106194690267e-06,
+      "loss": 0.7551,
+      "step": 2947
+    },
+    {
+      "epoch": 3.2610619469026547,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.847345132743363e-06,
+      "loss": 0.7462,
+      "step": 2948
+    },
+    {
+      "epoch": 3.2621681415929205,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.8445796460176993e-06,
+      "loss": 0.7763,
+      "step": 2949
+    },
+    {
+      "epoch": 3.2632743362831858,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.8418141592920355e-06,
+      "loss": 0.7796,
+      "step": 2950
+    },
+    {
+      "epoch": 3.2643805309734515,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.8390486725663718e-06,
+      "loss": 0.7605,
+      "step": 2951
+    },
+    {
+      "epoch": 3.265486725663717,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.8362831858407081e-06,
+      "loss": 0.7623,
+      "step": 2952
+    },
+    {
+      "epoch": 3.2665929203539825,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.8335176991150445e-06,
+      "loss": 0.7485,
+      "step": 2953
+    },
+    {
+      "epoch": 3.267699115044248,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.8307522123893806e-06,
+      "loss": 0.7708,
+      "step": 2954
+    },
+    {
+      "epoch": 3.268805309734513,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.8279867256637172e-06,
+      "loss": 0.802,
+      "step": 2955
+    },
+    {
+      "epoch": 3.269911504424779,
+      "grad_norm": 0.1015625,
+      "learning_rate": 1.8252212389380533e-06,
+      "loss": 0.7382,
+      "step": 2956
+    },
+    {
+      "epoch": 3.271017699115044,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.8224557522123896e-06,
+      "loss": 0.7506,
+      "step": 2957
+    },
+    {
+      "epoch": 3.27212389380531,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.8196902654867257e-06,
+      "loss": 0.8172,
+      "step": 2958
+    },
+    {
+      "epoch": 3.273230088495575,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.8169247787610623e-06,
+      "loss": 0.7783,
+      "step": 2959
+    },
+    {
+      "epoch": 3.274336283185841,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.8141592920353984e-06,
+      "loss": 0.7499,
+      "step": 2960
+    },
+    {
+      "epoch": 3.275442477876106,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.8113938053097345e-06,
+      "loss": 0.767,
+      "step": 2961
+    },
+    {
+      "epoch": 3.2765486725663715,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.8086283185840709e-06,
+      "loss": 0.7402,
+      "step": 2962
+    },
+    {
+      "epoch": 3.2776548672566372,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.8058628318584074e-06,
+      "loss": 0.7699,
+      "step": 2963
+    },
+    {
+      "epoch": 3.2787610619469025,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.8030973451327436e-06,
+      "loss": 0.7677,
+      "step": 2964
+    },
+    {
+      "epoch": 3.2798672566371683,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.8003318584070797e-06,
+      "loss": 0.7644,
+      "step": 2965
+    },
+    {
+      "epoch": 3.2809734513274336,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.797566371681416e-06,
+      "loss": 0.8009,
+      "step": 2966
+    },
+    {
+      "epoch": 3.2820796460176993,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.7948008849557524e-06,
+      "loss": 0.7547,
+      "step": 2967
+    },
+    {
+      "epoch": 3.2831858407079646,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.7920353982300887e-06,
+      "loss": 0.7688,
+      "step": 2968
+    },
+    {
+      "epoch": 3.28429203539823,
+      "grad_norm": 0.09765625,
+      "learning_rate": 1.7892699115044248e-06,
+      "loss": 0.7168,
+      "step": 2969
+    },
+    {
+      "epoch": 3.2853982300884956,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.7865044247787612e-06,
+      "loss": 0.7853,
+      "step": 2970
+    },
+    {
+      "epoch": 3.286504424778761,
+      "grad_norm": 0.125,
+      "learning_rate": 1.7837389380530975e-06,
+      "loss": 0.7943,
+      "step": 2971
+    },
+    {
+      "epoch": 3.2876106194690267,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.7809734513274338e-06,
+      "loss": 0.7802,
+      "step": 2972
+    },
+    {
+      "epoch": 3.288716814159292,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.77820796460177e-06,
+      "loss": 0.7915,
+      "step": 2973
+    },
+    {
+      "epoch": 3.2898230088495577,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.7754424778761065e-06,
+      "loss": 0.7523,
+      "step": 2974
+    },
+    {
+      "epoch": 3.290929203539823,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.7726769911504426e-06,
+      "loss": 0.8403,
+      "step": 2975
+    },
+    {
+      "epoch": 3.2920353982300883,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.769911504424779e-06,
+      "loss": 0.7436,
+      "step": 2976
+    },
+    {
+      "epoch": 3.293141592920354,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.767146017699115e-06,
+      "loss": 0.7487,
+      "step": 2977
+    },
+    {
+      "epoch": 3.2942477876106193,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.7643805309734517e-06,
+      "loss": 0.7969,
+      "step": 2978
+    },
+    {
+      "epoch": 3.295353982300885,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.7616150442477878e-06,
+      "loss": 0.7556,
+      "step": 2979
+    },
+    {
+      "epoch": 3.2964601769911503,
+      "grad_norm": 0.0947265625,
+      "learning_rate": 1.758849557522124e-06,
+      "loss": 0.7382,
+      "step": 2980
+    },
+    {
+      "epoch": 3.297566371681416,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.7560840707964602e-06,
+      "loss": 0.7446,
+      "step": 2981
+    },
+    {
+      "epoch": 3.2986725663716814,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.7533185840707968e-06,
+      "loss": 0.7206,
+      "step": 2982
+    },
+    {
+      "epoch": 3.2997787610619467,
+      "grad_norm": 0.125,
+      "learning_rate": 1.750553097345133e-06,
+      "loss": 0.8439,
+      "step": 2983
+    },
+    {
+      "epoch": 3.3008849557522124,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.747787610619469e-06,
+      "loss": 0.7918,
+      "step": 2984
+    },
+    {
+      "epoch": 3.3019911504424777,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 1.7450221238938054e-06,
+      "loss": 0.7106,
+      "step": 2985
+    },
+    {
+      "epoch": 3.3030973451327434,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.7422566371681417e-06,
+      "loss": 0.764,
+      "step": 2986
+    },
+    {
+      "epoch": 3.3042035398230087,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.739491150442478e-06,
+      "loss": 0.7522,
+      "step": 2987
+    },
+    {
+      "epoch": 3.3053097345132745,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 1.7367256637168142e-06,
+      "loss": 0.746,
+      "step": 2988
+    },
+    {
+      "epoch": 3.3064159292035398,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.7339601769911507e-06,
+      "loss": 0.7917,
+      "step": 2989
+    },
+    {
+      "epoch": 3.307522123893805,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.7311946902654869e-06,
+      "loss": 0.7682,
+      "step": 2990
+    },
+    {
+      "epoch": 3.308628318584071,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.7284292035398232e-06,
+      "loss": 0.7659,
+      "step": 2991
+    },
+    {
+      "epoch": 3.309734513274336,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.7256637168141593e-06,
+      "loss": 0.7351,
+      "step": 2992
+    },
+    {
+      "epoch": 3.310840707964602,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.7228982300884959e-06,
+      "loss": 0.756,
+      "step": 2993
+    },
+    {
+      "epoch": 3.311946902654867,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.720132743362832e-06,
+      "loss": 0.8222,
+      "step": 2994
+    },
+    {
+      "epoch": 3.313053097345133,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.7173672566371683e-06,
+      "loss": 0.7635,
+      "step": 2995
+    },
+    {
+      "epoch": 3.314159292035398,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.7146017699115045e-06,
+      "loss": 0.7967,
+      "step": 2996
+    },
+    {
+      "epoch": 3.315265486725664,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.711836283185841e-06,
+      "loss": 0.7629,
+      "step": 2997
+    },
+    {
+      "epoch": 3.316371681415929,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.7090707964601771e-06,
+      "loss": 0.7521,
+      "step": 2998
+    },
+    {
+      "epoch": 3.317477876106195,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.7063053097345133e-06,
+      "loss": 0.7631,
+      "step": 2999
+    },
+    {
+      "epoch": 3.3185840707964602,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.7035398230088496e-06,
+      "loss": 0.8189,
+      "step": 3000
+    },
+    {
+      "epoch": 3.3196902654867255,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.7007743362831862e-06,
+      "loss": 0.7463,
+      "step": 3001
+    },
+    {
+      "epoch": 3.3207964601769913,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.6980088495575223e-06,
+      "loss": 0.7441,
+      "step": 3002
+    },
+    {
+      "epoch": 3.3219026548672566,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.6952433628318584e-06,
+      "loss": 0.7705,
+      "step": 3003
+    },
+    {
+      "epoch": 3.3230088495575223,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.692477876106195e-06,
+      "loss": 0.7763,
+      "step": 3004
+    },
+    {
+      "epoch": 3.3241150442477876,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.689712389380531e-06,
+      "loss": 0.7616,
+      "step": 3005
+    },
+    {
+      "epoch": 3.3252212389380533,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.6869469026548674e-06,
+      "loss": 0.7593,
+      "step": 3006
+    },
+    {
+      "epoch": 3.3263274336283186,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.6841814159292035e-06,
+      "loss": 0.7687,
+      "step": 3007
+    },
+    {
+      "epoch": 3.327433628318584,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.68141592920354e-06,
+      "loss": 0.7595,
+      "step": 3008
+    },
+    {
+      "epoch": 3.3285398230088497,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.6786504424778762e-06,
+      "loss": 0.8121,
+      "step": 3009
+    },
+    {
+      "epoch": 3.329646017699115,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.6758849557522126e-06,
+      "loss": 0.8113,
+      "step": 3010
+    },
+    {
+      "epoch": 3.3307522123893807,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.6731194690265487e-06,
+      "loss": 0.8839,
+      "step": 3011
+    },
+    {
+      "epoch": 3.331858407079646,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.6703539823008852e-06,
+      "loss": 0.7943,
+      "step": 3012
+    },
+    {
+      "epoch": 3.3329646017699117,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.6675884955752214e-06,
+      "loss": 0.7401,
+      "step": 3013
+    },
+    {
+      "epoch": 3.334070796460177,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.6648230088495577e-06,
+      "loss": 0.7391,
+      "step": 3014
+    },
+    {
+      "epoch": 3.3351769911504423,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.6620575221238938e-06,
+      "loss": 0.7725,
+      "step": 3015
+    },
+    {
+      "epoch": 3.336283185840708,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.6592920353982304e-06,
+      "loss": 0.8269,
+      "step": 3016
+    },
+    {
+      "epoch": 3.3373893805309733,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.6565265486725665e-06,
+      "loss": 0.8016,
+      "step": 3017
+    },
+    {
+      "epoch": 3.338495575221239,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.6537610619469026e-06,
+      "loss": 0.7656,
+      "step": 3018
+    },
+    {
+      "epoch": 3.3396017699115044,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.6509955752212392e-06,
+      "loss": 0.8293,
+      "step": 3019
+    },
+    {
+      "epoch": 3.34070796460177,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.6482300884955755e-06,
+      "loss": 0.7724,
+      "step": 3020
+    },
+    {
+      "epoch": 3.3418141592920354,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.6454646017699116e-06,
+      "loss": 0.8,
+      "step": 3021
+    },
+    {
+      "epoch": 3.3429203539823007,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.6426991150442478e-06,
+      "loss": 0.7376,
+      "step": 3022
+    },
+    {
+      "epoch": 3.3440265486725664,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.6399336283185843e-06,
+      "loss": 0.7594,
+      "step": 3023
+    },
+    {
+      "epoch": 3.3451327433628317,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 1.6371681415929204e-06,
+      "loss": 0.7712,
+      "step": 3024
+    },
+    {
+      "epoch": 3.3462389380530975,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.6344026548672568e-06,
+      "loss": 0.7826,
+      "step": 3025
+    },
+    {
+      "epoch": 3.3473451327433628,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.631637168141593e-06,
+      "loss": 0.8136,
+      "step": 3026
+    },
+    {
+      "epoch": 3.3484513274336285,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.6288716814159295e-06,
+      "loss": 0.7503,
+      "step": 3027
+    },
+    {
+      "epoch": 3.349557522123894,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.6261061946902656e-06,
+      "loss": 0.7463,
+      "step": 3028
+    },
+    {
+      "epoch": 3.350663716814159,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.623340707964602e-06,
+      "loss": 0.7692,
+      "step": 3029
+    },
+    {
+      "epoch": 3.351769911504425,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.620575221238938e-06,
+      "loss": 0.7478,
+      "step": 3030
+    },
+    {
+      "epoch": 3.35287610619469,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.6178097345132746e-06,
+      "loss": 0.8248,
+      "step": 3031
+    },
+    {
+      "epoch": 3.353982300884956,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.6150442477876107e-06,
+      "loss": 0.8745,
+      "step": 3032
+    },
+    {
+      "epoch": 3.355088495575221,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.612278761061947e-06,
+      "loss": 0.7555,
+      "step": 3033
+    },
+    {
+      "epoch": 3.356194690265487,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.6095132743362832e-06,
+      "loss": 0.7665,
+      "step": 3034
+    },
+    {
+      "epoch": 3.357300884955752,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.6067477876106197e-06,
+      "loss": 0.7723,
+      "step": 3035
+    },
+    {
+      "epoch": 3.3584070796460175,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.6039823008849559e-06,
+      "loss": 0.7564,
+      "step": 3036
+    },
+    {
+      "epoch": 3.359513274336283,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 1.6012168141592922e-06,
+      "loss": 0.7207,
+      "step": 3037
+    },
+    {
+      "epoch": 3.3606194690265485,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.5984513274336285e-06,
+      "loss": 0.7483,
+      "step": 3038
+    },
+    {
+      "epoch": 3.3617256637168142,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.5956858407079649e-06,
+      "loss": 0.7782,
+      "step": 3039
+    },
+    {
+      "epoch": 3.3628318584070795,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.592920353982301e-06,
+      "loss": 0.7785,
+      "step": 3040
+    },
+    {
+      "epoch": 3.3639380530973453,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.5901548672566371e-06,
+      "loss": 0.7876,
+      "step": 3041
+    },
+    {
+      "epoch": 3.3650442477876106,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.5873893805309737e-06,
+      "loss": 0.8199,
+      "step": 3042
+    },
+    {
+      "epoch": 3.366150442477876,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.5846238938053098e-06,
+      "loss": 0.8004,
+      "step": 3043
+    },
+    {
+      "epoch": 3.3672566371681416,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.5818584070796461e-06,
+      "loss": 0.7821,
+      "step": 3044
+    },
+    {
+      "epoch": 3.368362831858407,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.5790929203539823e-06,
+      "loss": 0.8074,
+      "step": 3045
+    },
+    {
+      "epoch": 3.3694690265486726,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.5763274336283188e-06,
+      "loss": 0.7672,
+      "step": 3046
+    },
+    {
+      "epoch": 3.370575221238938,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.573561946902655e-06,
+      "loss": 0.7739,
+      "step": 3047
+    },
+    {
+      "epoch": 3.3716814159292037,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.5707964601769913e-06,
+      "loss": 0.7628,
+      "step": 3048
+    },
+    {
+      "epoch": 3.372787610619469,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.5680309734513274e-06,
+      "loss": 0.7639,
+      "step": 3049
+    },
+    {
+      "epoch": 3.3738938053097347,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.565265486725664e-06,
+      "loss": 0.7579,
+      "step": 3050
+    },
+    {
+      "epoch": 3.375,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.5625e-06,
+      "loss": 0.7965,
+      "step": 3051
+    },
+    {
+      "epoch": 3.3761061946902653,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.5597345132743364e-06,
+      "loss": 0.8198,
+      "step": 3052
+    },
+    {
+      "epoch": 3.377212389380531,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.5569690265486728e-06,
+      "loss": 0.7436,
+      "step": 3053
+    },
+    {
+      "epoch": 3.3783185840707963,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.554203539823009e-06,
+      "loss": 0.7583,
+      "step": 3054
+    },
+    {
+      "epoch": 3.379424778761062,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.5514380530973452e-06,
+      "loss": 0.751,
+      "step": 3055
+    },
+    {
+      "epoch": 3.3805309734513274,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.5486725663716816e-06,
+      "loss": 0.8007,
+      "step": 3056
+    },
+    {
+      "epoch": 3.381637168141593,
+      "grad_norm": 0.10546875,
+      "learning_rate": 1.5459070796460179e-06,
+      "loss": 0.7488,
+      "step": 3057
+    },
+    {
+      "epoch": 3.3827433628318584,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.5431415929203542e-06,
+      "loss": 0.7603,
+      "step": 3058
+    },
+    {
+      "epoch": 3.383849557522124,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.5403761061946904e-06,
+      "loss": 0.767,
+      "step": 3059
+    },
+    {
+      "epoch": 3.3849557522123894,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.5376106194690265e-06,
+      "loss": 0.7882,
+      "step": 3060
+    },
+    {
+      "epoch": 3.3860619469026547,
+      "grad_norm": 0.1015625,
+      "learning_rate": 1.534845132743363e-06,
+      "loss": 0.7316,
+      "step": 3061
+    },
+    {
+      "epoch": 3.3871681415929205,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.5320796460176994e-06,
+      "loss": 0.7543,
+      "step": 3062
+    },
+    {
+      "epoch": 3.3882743362831858,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.5293141592920355e-06,
+      "loss": 0.818,
+      "step": 3063
+    },
+    {
+      "epoch": 3.3893805309734515,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.5265486725663716e-06,
+      "loss": 0.7762,
+      "step": 3064
+    },
+    {
+      "epoch": 3.390486725663717,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.5237831858407082e-06,
+      "loss": 0.7873,
+      "step": 3065
+    },
+    {
+      "epoch": 3.3915929203539825,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.5210176991150443e-06,
+      "loss": 0.7977,
+      "step": 3066
+    },
+    {
+      "epoch": 3.392699115044248,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.5182522123893806e-06,
+      "loss": 0.8028,
+      "step": 3067
+    },
+    {
+      "epoch": 3.393805309734513,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.515486725663717e-06,
+      "loss": 0.7551,
+      "step": 3068
+    },
+    {
+      "epoch": 3.394911504424779,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.5127212389380533e-06,
+      "loss": 0.7861,
+      "step": 3069
+    },
+    {
+      "epoch": 3.396017699115044,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.5099557522123894e-06,
+      "loss": 0.8023,
+      "step": 3070
+    },
+    {
+      "epoch": 3.39712389380531,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.5071902654867258e-06,
+      "loss": 0.7769,
+      "step": 3071
+    },
+    {
+      "epoch": 3.398230088495575,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.5044247787610621e-06,
+      "loss": 0.8126,
+      "step": 3072
+    },
+    {
+      "epoch": 3.399336283185841,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.5016592920353984e-06,
+      "loss": 0.6925,
+      "step": 3073
+    },
+    {
+      "epoch": 3.400442477876106,
+      "grad_norm": 0.103515625,
+      "learning_rate": 1.4988938053097346e-06,
+      "loss": 0.7322,
+      "step": 3074
+    },
+    {
+      "epoch": 3.4015486725663715,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.496128318584071e-06,
+      "loss": 0.7938,
+      "step": 3075
+    },
+    {
+      "epoch": 3.4026548672566372,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.4933628318584072e-06,
+      "loss": 0.7724,
+      "step": 3076
+    },
+    {
+      "epoch": 3.4037610619469025,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.4905973451327436e-06,
+      "loss": 0.7471,
+      "step": 3077
+    },
+    {
+      "epoch": 3.4048672566371683,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.4878318584070797e-06,
+      "loss": 0.7635,
+      "step": 3078
+    },
+    {
+      "epoch": 3.4059734513274336,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.4850663716814158e-06,
+      "loss": 0.8145,
+      "step": 3079
+    },
+    {
+      "epoch": 3.4070796460176993,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.4823008849557524e-06,
+      "loss": 0.7641,
+      "step": 3080
+    },
+    {
+      "epoch": 3.4081858407079646,
+      "grad_norm": 0.10546875,
+      "learning_rate": 1.4795353982300887e-06,
+      "loss": 0.7644,
+      "step": 3081
+    },
+    {
+      "epoch": 3.40929203539823,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.4767699115044249e-06,
+      "loss": 0.7885,
+      "step": 3082
+    },
+    {
+      "epoch": 3.4103982300884956,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.474004424778761e-06,
+      "loss": 0.7944,
+      "step": 3083
+    },
+    {
+      "epoch": 3.411504424778761,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.4712389380530975e-06,
+      "loss": 0.7514,
+      "step": 3084
+    },
+    {
+      "epoch": 3.4126106194690267,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.4684734513274337e-06,
+      "loss": 0.787,
+      "step": 3085
+    },
+    {
+      "epoch": 3.413716814159292,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.46570796460177e-06,
+      "loss": 0.7894,
+      "step": 3086
+    },
+    {
+      "epoch": 3.4148230088495577,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 1.4629424778761065e-06,
+      "loss": 0.7489,
+      "step": 3087
+    },
+    {
+      "epoch": 3.415929203539823,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.4601769911504427e-06,
+      "loss": 0.7375,
+      "step": 3088
+    },
+    {
+      "epoch": 3.4170353982300883,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.4574115044247788e-06,
+      "loss": 0.7296,
+      "step": 3089
+    },
+    {
+      "epoch": 3.418141592920354,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.4546460176991151e-06,
+      "loss": 0.7683,
+      "step": 3090
+    },
+    {
+      "epoch": 3.4192477876106193,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.4518805309734515e-06,
+      "loss": 0.7462,
+      "step": 3091
+    },
+    {
+      "epoch": 3.420353982300885,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.4491150442477878e-06,
+      "loss": 0.7746,
+      "step": 3092
+    },
+    {
+      "epoch": 3.4214601769911503,
+      "grad_norm": 0.125,
+      "learning_rate": 1.446349557522124e-06,
+      "loss": 0.7948,
+      "step": 3093
+    },
+    {
+      "epoch": 3.422566371681416,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.4435840707964603e-06,
+      "loss": 0.7759,
+      "step": 3094
+    },
+    {
+      "epoch": 3.4236725663716814,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.4408185840707966e-06,
+      "loss": 0.7402,
+      "step": 3095
+    },
+    {
+      "epoch": 3.4247787610619467,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 1.438053097345133e-06,
+      "loss": 0.7578,
+      "step": 3096
+    },
+    {
+      "epoch": 3.4258849557522124,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.435287610619469e-06,
+      "loss": 0.788,
+      "step": 3097
+    },
+    {
+      "epoch": 3.4269911504424777,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.4325221238938052e-06,
+      "loss": 0.7669,
+      "step": 3098
+    },
+    {
+      "epoch": 3.4280973451327434,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.4297566371681417e-06,
+      "loss": 0.783,
+      "step": 3099
+    },
+    {
+      "epoch": 3.4292035398230087,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.426991150442478e-06,
+      "loss": 0.7729,
+      "step": 3100
+    },
+    {
+      "epoch": 3.4303097345132745,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.4242256637168142e-06,
+      "loss": 0.7259,
+      "step": 3101
+    },
+    {
+      "epoch": 3.4314159292035398,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.4214601769911508e-06,
+      "loss": 0.7694,
+      "step": 3102
+    },
+    {
+      "epoch": 3.432522123893805,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.4186946902654869e-06,
+      "loss": 0.7564,
+      "step": 3103
+    },
+    {
+      "epoch": 3.433628318584071,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.415929203539823e-06,
+      "loss": 0.7785,
+      "step": 3104
+    },
+    {
+      "epoch": 3.434734513274336,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.4131637168141594e-06,
+      "loss": 0.7685,
+      "step": 3105
+    },
+    {
+      "epoch": 3.435840707964602,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.410398230088496e-06,
+      "loss": 0.7482,
+      "step": 3106
+    },
+    {
+      "epoch": 3.436946902654867,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 1.407632743362832e-06,
+      "loss": 0.7776,
+      "step": 3107
+    },
+    {
+      "epoch": 3.438053097345133,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.4048672566371682e-06,
+      "loss": 0.8133,
+      "step": 3108
+    },
+    {
+      "epoch": 3.439159292035398,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.4021017699115045e-06,
+      "loss": 0.7506,
+      "step": 3109
+    },
+    {
+      "epoch": 3.440265486725664,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.3993362831858408e-06,
+      "loss": 0.774,
+      "step": 3110
+    },
+    {
+      "epoch": 3.441371681415929,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.3965707964601772e-06,
+      "loss": 0.8146,
+      "step": 3111
+    },
+    {
+      "epoch": 3.442477876106195,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.3938053097345133e-06,
+      "loss": 0.7578,
+      "step": 3112
+    },
+    {
+      "epoch": 3.4435840707964602,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.3910398230088496e-06,
+      "loss": 0.7311,
+      "step": 3113
+    },
+    {
+      "epoch": 3.4446902654867255,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.388274336283186e-06,
+      "loss": 0.753,
+      "step": 3114
+    },
+    {
+      "epoch": 3.4457964601769913,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 1.3855088495575223e-06,
+      "loss": 0.7302,
+      "step": 3115
+    },
+    {
+      "epoch": 3.4469026548672566,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.3827433628318584e-06,
+      "loss": 0.7934,
+      "step": 3116
+    },
+    {
+      "epoch": 3.4480088495575223,
+      "grad_norm": 0.103515625,
+      "learning_rate": 1.379977876106195e-06,
+      "loss": 0.7162,
+      "step": 3117
+    },
+    {
+      "epoch": 3.4491150442477876,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.377212389380531e-06,
+      "loss": 0.789,
+      "step": 3118
+    },
+    {
+      "epoch": 3.4502212389380533,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.3744469026548674e-06,
+      "loss": 0.796,
+      "step": 3119
+    },
+    {
+      "epoch": 3.4513274336283186,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.3716814159292036e-06,
+      "loss": 0.7788,
+      "step": 3120
+    },
+    {
+      "epoch": 3.452433628318584,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.3689159292035401e-06,
+      "loss": 0.8145,
+      "step": 3121
+    },
+    {
+      "epoch": 3.4535398230088497,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.3661504424778762e-06,
+      "loss": 0.7648,
+      "step": 3122
+    },
+    {
+      "epoch": 3.454646017699115,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 1.3633849557522124e-06,
+      "loss": 0.7182,
+      "step": 3123
+    },
+    {
+      "epoch": 3.4557522123893807,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.3606194690265487e-06,
+      "loss": 0.7394,
+      "step": 3124
+    },
+    {
+      "epoch": 3.456858407079646,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.3578539823008853e-06,
+      "loss": 0.7572,
+      "step": 3125
+    },
+    {
+      "epoch": 3.4579646017699117,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.3550884955752214e-06,
+      "loss": 0.7561,
+      "step": 3126
+    },
+    {
+      "epoch": 3.459070796460177,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.3523230088495575e-06,
+      "loss": 0.7885,
+      "step": 3127
+    },
+    {
+      "epoch": 3.4601769911504423,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.3495575221238938e-06,
+      "loss": 0.7643,
+      "step": 3128
+    },
+    {
+      "epoch": 3.461283185840708,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.3467920353982302e-06,
+      "loss": 0.744,
+      "step": 3129
+    },
+    {
+      "epoch": 3.4623893805309733,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.3440265486725665e-06,
+      "loss": 0.728,
+      "step": 3130
+    },
+    {
+      "epoch": 3.463495575221239,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.3412610619469026e-06,
+      "loss": 0.7464,
+      "step": 3131
+    },
+    {
+      "epoch": 3.4646017699115044,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.3384955752212392e-06,
+      "loss": 0.7714,
+      "step": 3132
+    },
+    {
+      "epoch": 3.46570796460177,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.3357300884955753e-06,
+      "loss": 0.7504,
+      "step": 3133
+    },
+    {
+      "epoch": 3.4668141592920354,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.3329646017699117e-06,
+      "loss": 0.7745,
+      "step": 3134
+    },
+    {
+      "epoch": 3.4679203539823007,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.3301991150442478e-06,
+      "loss": 0.7698,
+      "step": 3135
+    },
+    {
+      "epoch": 3.4690265486725664,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.3274336283185843e-06,
+      "loss": 0.7525,
+      "step": 3136
+    },
+    {
+      "epoch": 3.4701327433628317,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 1.3246681415929205e-06,
+      "loss": 0.7358,
+      "step": 3137
+    },
+    {
+      "epoch": 3.4712389380530975,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.3219026548672568e-06,
+      "loss": 0.7834,
+      "step": 3138
+    },
+    {
+      "epoch": 3.4723451327433628,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.319137168141593e-06,
+      "loss": 0.7735,
+      "step": 3139
+    },
+    {
+      "epoch": 3.4734513274336285,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.3163716814159295e-06,
+      "loss": 0.7444,
+      "step": 3140
+    },
+    {
+      "epoch": 3.474557522123894,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.3136061946902656e-06,
+      "loss": 0.7934,
+      "step": 3141
+    },
+    {
+      "epoch": 3.475663716814159,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.3108407079646017e-06,
+      "loss": 0.8013,
+      "step": 3142
+    },
+    {
+      "epoch": 3.476769911504425,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.308075221238938e-06,
+      "loss": 0.8086,
+      "step": 3143
+    },
+    {
+      "epoch": 3.47787610619469,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.3053097345132746e-06,
+      "loss": 0.8043,
+      "step": 3144
+    },
+    {
+      "epoch": 3.478982300884956,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.3025442477876107e-06,
+      "loss": 0.7868,
+      "step": 3145
+    },
+    {
+      "epoch": 3.480088495575221,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.2997787610619469e-06,
+      "loss": 0.7846,
+      "step": 3146
+    },
+    {
+      "epoch": 3.481194690265487,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.2970132743362832e-06,
+      "loss": 0.7728,
+      "step": 3147
+    },
+    {
+      "epoch": 3.482300884955752,
+      "grad_norm": 0.099609375,
+      "learning_rate": 1.2942477876106195e-06,
+      "loss": 0.7356,
+      "step": 3148
+    },
+    {
+      "epoch": 3.4834070796460175,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.2914823008849559e-06,
+      "loss": 0.7895,
+      "step": 3149
+    },
+    {
+      "epoch": 3.484513274336283,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.288716814159292e-06,
+      "loss": 0.7684,
+      "step": 3150
+    },
+    {
+      "epoch": 3.4856194690265485,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.2859513274336286e-06,
+      "loss": 0.7624,
+      "step": 3151
+    },
+    {
+      "epoch": 3.4867256637168142,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.2831858407079647e-06,
+      "loss": 0.7947,
+      "step": 3152
+    },
+    {
+      "epoch": 3.4878318584070795,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.280420353982301e-06,
+      "loss": 0.7658,
+      "step": 3153
+    },
+    {
+      "epoch": 3.4889380530973453,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.2776548672566371e-06,
+      "loss": 0.7711,
+      "step": 3154
+    },
+    {
+      "epoch": 3.4900442477876106,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.2748893805309737e-06,
+      "loss": 0.7283,
+      "step": 3155
+    },
+    {
+      "epoch": 3.491150442477876,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.2721238938053098e-06,
+      "loss": 0.7852,
+      "step": 3156
+    },
+    {
+      "epoch": 3.4922566371681416,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 1.2693584070796462e-06,
+      "loss": 0.7342,
+      "step": 3157
+    },
+    {
+      "epoch": 3.493362831858407,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.2665929203539823e-06,
+      "loss": 0.7884,
+      "step": 3158
+    },
+    {
+      "epoch": 3.4944690265486726,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.2638274336283188e-06,
+      "loss": 0.7812,
+      "step": 3159
+    },
+    {
+      "epoch": 3.495575221238938,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.261061946902655e-06,
+      "loss": 0.7284,
+      "step": 3160
+    },
+    {
+      "epoch": 3.4966814159292037,
+      "grad_norm": 0.10546875,
+      "learning_rate": 1.2582964601769913e-06,
+      "loss": 0.7506,
+      "step": 3161
+    },
+    {
+      "epoch": 3.497787610619469,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.2555309734513274e-06,
+      "loss": 0.7465,
+      "step": 3162
+    },
+    {
+      "epoch": 3.4988938053097347,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.252765486725664e-06,
+      "loss": 0.7455,
+      "step": 3163
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.25e-06,
+      "loss": 0.7328,
+      "step": 3164
+    },
+    {
+      "epoch": 3.5011061946902657,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.2472345132743364e-06,
+      "loss": 0.7815,
+      "step": 3165
+    },
+    {
+      "epoch": 3.502212389380531,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.2444690265486726e-06,
+      "loss": 0.7559,
+      "step": 3166
+    },
+    {
+      "epoch": 3.5033185840707963,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.241703539823009e-06,
+      "loss": 0.7893,
+      "step": 3167
+    },
+    {
+      "epoch": 3.504424778761062,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.2389380530973452e-06,
+      "loss": 0.8277,
+      "step": 3168
+    },
+    {
+      "epoch": 3.5055309734513274,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.2361725663716816e-06,
+      "loss": 0.784,
+      "step": 3169
+    },
+    {
+      "epoch": 3.5066371681415927,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.2334070796460177e-06,
+      "loss": 0.7988,
+      "step": 3170
+    },
+    {
+      "epoch": 3.5077433628318584,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.230641592920354e-06,
+      "loss": 0.7549,
+      "step": 3171
+    },
+    {
+      "epoch": 3.508849557522124,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.2278761061946904e-06,
+      "loss": 0.7492,
+      "step": 3172
+    },
+    {
+      "epoch": 3.5099557522123894,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.2251106194690267e-06,
+      "loss": 0.8062,
+      "step": 3173
+    },
+    {
+      "epoch": 3.5110619469026547,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 1.2223451327433628e-06,
+      "loss": 0.7559,
+      "step": 3174
+    },
+    {
+      "epoch": 3.5121681415929205,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.2195796460176992e-06,
+      "loss": 0.7651,
+      "step": 3175
+    },
+    {
+      "epoch": 3.5132743362831858,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.2168141592920355e-06,
+      "loss": 0.775,
+      "step": 3176
+    },
+    {
+      "epoch": 3.5143805309734515,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.2140486725663719e-06,
+      "loss": 0.7448,
+      "step": 3177
+    },
+    {
+      "epoch": 3.515486725663717,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.211283185840708e-06,
+      "loss": 0.7734,
+      "step": 3178
+    },
+    {
+      "epoch": 3.5165929203539825,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.2085176991150443e-06,
+      "loss": 0.7844,
+      "step": 3179
+    },
+    {
+      "epoch": 3.517699115044248,
+      "grad_norm": 0.0966796875,
+      "learning_rate": 1.2057522123893807e-06,
+      "loss": 0.7549,
+      "step": 3180
+    },
+    {
+      "epoch": 3.518805309734513,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 1.202986725663717e-06,
+      "loss": 0.7611,
+      "step": 3181
+    },
+    {
+      "epoch": 3.519911504424779,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.2002212389380533e-06,
+      "loss": 0.7597,
+      "step": 3182
+    },
+    {
+      "epoch": 3.521017699115044,
+      "grad_norm": 0.099609375,
+      "learning_rate": 1.1974557522123895e-06,
+      "loss": 0.7502,
+      "step": 3183
+    },
+    {
+      "epoch": 3.52212389380531,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.1946902654867258e-06,
+      "loss": 0.7852,
+      "step": 3184
+    },
+    {
+      "epoch": 3.523230088495575,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.191924778761062e-06,
+      "loss": 0.7705,
+      "step": 3185
+    },
+    {
+      "epoch": 3.524336283185841,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.1891592920353985e-06,
+      "loss": 0.8095,
+      "step": 3186
+    },
+    {
+      "epoch": 3.525442477876106,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.1863938053097346e-06,
+      "loss": 0.7799,
+      "step": 3187
+    },
+    {
+      "epoch": 3.5265486725663715,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.183628318584071e-06,
+      "loss": 0.7736,
+      "step": 3188
+    },
+    {
+      "epoch": 3.5276548672566372,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.180862831858407e-06,
+      "loss": 0.814,
+      "step": 3189
+    },
+    {
+      "epoch": 3.5287610619469025,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 1.1780973451327434e-06,
+      "loss": 0.7429,
+      "step": 3190
+    },
+    {
+      "epoch": 3.5298672566371683,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.1753318584070797e-06,
+      "loss": 0.7459,
+      "step": 3191
+    },
+    {
+      "epoch": 3.5309734513274336,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.172566371681416e-06,
+      "loss": 0.8166,
+      "step": 3192
+    },
+    {
+      "epoch": 3.5320796460176993,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.1698008849557522e-06,
+      "loss": 0.7658,
+      "step": 3193
+    },
+    {
+      "epoch": 3.5331858407079646,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 1.1670353982300885e-06,
+      "loss": 0.7665,
+      "step": 3194
+    },
+    {
+      "epoch": 3.53429203539823,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.1642699115044249e-06,
+      "loss": 0.7663,
+      "step": 3195
+    },
+    {
+      "epoch": 3.5353982300884956,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.1615044247787612e-06,
+      "loss": 0.7568,
+      "step": 3196
+    },
+    {
+      "epoch": 3.536504424778761,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.1587389380530976e-06,
+      "loss": 0.8013,
+      "step": 3197
+    },
+    {
+      "epoch": 3.5376106194690267,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.1559734513274337e-06,
+      "loss": 0.7748,
+      "step": 3198
+    },
+    {
+      "epoch": 3.538716814159292,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 1.15320796460177e-06,
+      "loss": 0.7326,
+      "step": 3199
+    },
+    {
+      "epoch": 3.5398230088495577,
+      "grad_norm": 0.1015625,
+      "learning_rate": 1.1504424778761064e-06,
+      "loss": 0.7426,
+      "step": 3200
+    },
+    {
+      "epoch": 3.540929203539823,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.1476769911504427e-06,
+      "loss": 0.7614,
+      "step": 3201
+    },
+    {
+      "epoch": 3.5420353982300883,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.1449115044247788e-06,
+      "loss": 0.7352,
+      "step": 3202
+    },
+    {
+      "epoch": 3.543141592920354,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.1421460176991152e-06,
+      "loss": 0.7816,
+      "step": 3203
+    },
+    {
+      "epoch": 3.5442477876106193,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.1393805309734513e-06,
+      "loss": 0.8013,
+      "step": 3204
+    },
+    {
+      "epoch": 3.545353982300885,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.1366150442477878e-06,
+      "loss": 0.7381,
+      "step": 3205
+    },
+    {
+      "epoch": 3.5464601769911503,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.133849557522124e-06,
+      "loss": 0.8078,
+      "step": 3206
+    },
+    {
+      "epoch": 3.547566371681416,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.1310840707964603e-06,
+      "loss": 0.7874,
+      "step": 3207
+    },
+    {
+      "epoch": 3.5486725663716814,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.1283185840707964e-06,
+      "loss": 0.8247,
+      "step": 3208
+    },
+    {
+      "epoch": 3.5497787610619467,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.1255530973451328e-06,
+      "loss": 0.8038,
+      "step": 3209
+    },
+    {
+      "epoch": 3.5508849557522124,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.122787610619469e-06,
+      "loss": 0.7503,
+      "step": 3210
+    },
+    {
+      "epoch": 3.551991150442478,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.1200221238938054e-06,
+      "loss": 0.8698,
+      "step": 3211
+    },
+    {
+      "epoch": 3.5530973451327434,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.1172566371681416e-06,
+      "loss": 0.7221,
+      "step": 3212
+    },
+    {
+      "epoch": 3.5542035398230087,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.114491150442478e-06,
+      "loss": 0.7487,
+      "step": 3213
+    },
+    {
+      "epoch": 3.5553097345132745,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.1117256637168142e-06,
+      "loss": 0.7465,
+      "step": 3214
+    },
+    {
+      "epoch": 3.5564159292035398,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.1089601769911506e-06,
+      "loss": 0.7584,
+      "step": 3215
+    },
+    {
+      "epoch": 3.557522123893805,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.106194690265487e-06,
+      "loss": 0.797,
+      "step": 3216
+    },
+    {
+      "epoch": 3.558628318584071,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.103429203539823e-06,
+      "loss": 0.772,
+      "step": 3217
+    },
+    {
+      "epoch": 3.5597345132743365,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.1006637168141594e-06,
+      "loss": 0.805,
+      "step": 3218
+    },
+    {
+      "epoch": 3.560840707964602,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.0978982300884957e-06,
+      "loss": 0.7804,
+      "step": 3219
+    },
+    {
+      "epoch": 3.561946902654867,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.095132743362832e-06,
+      "loss": 0.7903,
+      "step": 3220
+    },
+    {
+      "epoch": 3.563053097345133,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.0923672566371682e-06,
+      "loss": 0.7809,
+      "step": 3221
+    },
+    {
+      "epoch": 3.564159292035398,
+      "grad_norm": 0.125,
+      "learning_rate": 1.0896017699115045e-06,
+      "loss": 0.844,
+      "step": 3222
+    },
+    {
+      "epoch": 3.5652654867256635,
+      "grad_norm": 0.1015625,
+      "learning_rate": 1.0868362831858409e-06,
+      "loss": 0.7668,
+      "step": 3223
+    },
+    {
+      "epoch": 3.566371681415929,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.0840707964601772e-06,
+      "loss": 0.7811,
+      "step": 3224
+    },
+    {
+      "epoch": 3.567477876106195,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.0813053097345133e-06,
+      "loss": 0.7853,
+      "step": 3225
+    },
+    {
+      "epoch": 3.5685840707964602,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.0785398230088497e-06,
+      "loss": 0.7572,
+      "step": 3226
+    },
+    {
+      "epoch": 3.5696902654867255,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.0757743362831858e-06,
+      "loss": 0.7584,
+      "step": 3227
+    },
+    {
+      "epoch": 3.5707964601769913,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.0730088495575221e-06,
+      "loss": 0.7527,
+      "step": 3228
+    },
+    {
+      "epoch": 3.5719026548672566,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.0702433628318585e-06,
+      "loss": 0.7381,
+      "step": 3229
+    },
+    {
+      "epoch": 3.573008849557522,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.0674778761061948e-06,
+      "loss": 0.8225,
+      "step": 3230
+    },
+    {
+      "epoch": 3.5741150442477876,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.0647123893805311e-06,
+      "loss": 0.7771,
+      "step": 3231
+    },
+    {
+      "epoch": 3.5752212389380533,
+      "grad_norm": 0.125,
+      "learning_rate": 1.0619469026548673e-06,
+      "loss": 0.7878,
+      "step": 3232
+    },
+    {
+      "epoch": 3.5763274336283186,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.0591814159292036e-06,
+      "loss": 0.7325,
+      "step": 3233
+    },
+    {
+      "epoch": 3.577433628318584,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 1.05641592920354e-06,
+      "loss": 0.7527,
+      "step": 3234
+    },
+    {
+      "epoch": 3.5785398230088497,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.0536504424778763e-06,
+      "loss": 0.7746,
+      "step": 3235
+    },
+    {
+      "epoch": 3.579646017699115,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.0508849557522124e-06,
+      "loss": 0.7669,
+      "step": 3236
+    },
+    {
+      "epoch": 3.5807522123893807,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.0481194690265487e-06,
+      "loss": 0.7817,
+      "step": 3237
+    },
+    {
+      "epoch": 3.581858407079646,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.045353982300885e-06,
+      "loss": 0.7405,
+      "step": 3238
+    },
+    {
+      "epoch": 3.5829646017699117,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.0425884955752214e-06,
+      "loss": 0.7462,
+      "step": 3239
+    },
+    {
+      "epoch": 3.584070796460177,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.0398230088495575e-06,
+      "loss": 0.7862,
+      "step": 3240
+    },
+    {
+      "epoch": 3.5851769911504423,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.0370575221238939e-06,
+      "loss": 0.7976,
+      "step": 3241
+    },
+    {
+      "epoch": 3.586283185840708,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.0342920353982302e-06,
+      "loss": 0.7655,
+      "step": 3242
+    },
+    {
+      "epoch": 3.5873893805309733,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.0315265486725666e-06,
+      "loss": 0.7768,
+      "step": 3243
+    },
+    {
+      "epoch": 3.588495575221239,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.0287610619469027e-06,
+      "loss": 0.7894,
+      "step": 3244
+    },
+    {
+      "epoch": 3.5896017699115044,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.025995575221239e-06,
+      "loss": 0.7665,
+      "step": 3245
+    },
+    {
+      "epoch": 3.59070796460177,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.0232300884955754e-06,
+      "loss": 0.7815,
+      "step": 3246
+    },
+    {
+      "epoch": 3.5918141592920354,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.0204646017699115e-06,
+      "loss": 0.7578,
+      "step": 3247
+    },
+    {
+      "epoch": 3.5929203539823007,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.017699115044248e-06,
+      "loss": 0.7622,
+      "step": 3248
+    },
+    {
+      "epoch": 3.5940265486725664,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.0149336283185842e-06,
+      "loss": 0.741,
+      "step": 3249
+    },
+    {
+      "epoch": 3.5951327433628317,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.0121681415929205e-06,
+      "loss": 0.7711,
+      "step": 3250
+    },
+    {
+      "epoch": 3.5962389380530975,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.0094026548672566e-06,
+      "loss": 0.7345,
+      "step": 3251
+    },
+    {
+      "epoch": 3.5973451327433628,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.006637168141593e-06,
+      "loss": 0.8154,
+      "step": 3252
+    },
+    {
+      "epoch": 3.5984513274336285,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.0038716814159293e-06,
+      "loss": 0.8184,
+      "step": 3253
+    },
+    {
+      "epoch": 3.599557522123894,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.0011061946902656e-06,
+      "loss": 0.7724,
+      "step": 3254
+    },
+    {
+      "epoch": 3.600663716814159,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 9.983407079646018e-07,
+      "loss": 0.7719,
+      "step": 3255
+    },
+    {
+      "epoch": 3.601769911504425,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 9.95575221238938e-07,
+      "loss": 0.7636,
+      "step": 3256
+    },
+    {
+      "epoch": 3.60287610619469,
+      "grad_norm": 0.09765625,
+      "learning_rate": 9.928097345132744e-07,
+      "loss": 0.7308,
+      "step": 3257
+    },
+    {
+      "epoch": 3.603982300884956,
+      "grad_norm": 0.099609375,
+      "learning_rate": 9.900442477876108e-07,
+      "loss": 0.7328,
+      "step": 3258
+    },
+    {
+      "epoch": 3.605088495575221,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 9.87278761061947e-07,
+      "loss": 0.746,
+      "step": 3259
+    },
+    {
+      "epoch": 3.606194690265487,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 9.845132743362832e-07,
+      "loss": 0.7477,
+      "step": 3260
+    },
+    {
+      "epoch": 3.607300884955752,
+      "grad_norm": 0.1171875,
+      "learning_rate": 9.817477876106196e-07,
+      "loss": 0.7759,
+      "step": 3261
+    },
+    {
+      "epoch": 3.6084070796460175,
+      "grad_norm": 0.115234375,
+      "learning_rate": 9.78982300884956e-07,
+      "loss": 0.7518,
+      "step": 3262
+    },
+    {
+      "epoch": 3.609513274336283,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 9.762168141592922e-07,
+      "loss": 0.7493,
+      "step": 3263
+    },
+    {
+      "epoch": 3.6106194690265485,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 9.734513274336284e-07,
+      "loss": 0.7691,
+      "step": 3264
+    },
+    {
+      "epoch": 3.6117256637168142,
+      "grad_norm": 0.134765625,
+      "learning_rate": 9.706858407079647e-07,
+      "loss": 0.7905,
+      "step": 3265
+    },
+    {
+      "epoch": 3.6128318584070795,
+      "grad_norm": 0.10546875,
+      "learning_rate": 9.679203539823008e-07,
+      "loss": 0.7734,
+      "step": 3266
+    },
+    {
+      "epoch": 3.6139380530973453,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 9.651548672566374e-07,
+      "loss": 0.7727,
+      "step": 3267
+    },
+    {
+      "epoch": 3.6150442477876106,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 9.623893805309735e-07,
+      "loss": 0.7749,
+      "step": 3268
+    },
+    {
+      "epoch": 3.616150442477876,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.596238938053099e-07,
+      "loss": 0.8299,
+      "step": 3269
+    },
+    {
+      "epoch": 3.6172566371681416,
+      "grad_norm": 0.14453125,
+      "learning_rate": 9.56858407079646e-07,
+      "loss": 0.8083,
+      "step": 3270
+    },
+    {
+      "epoch": 3.6183628318584073,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 9.540929203539823e-07,
+      "loss": 0.7638,
+      "step": 3271
+    },
+    {
+      "epoch": 3.6194690265486726,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 9.513274336283187e-07,
+      "loss": 0.7957,
+      "step": 3272
+    },
+    {
+      "epoch": 3.620575221238938,
+      "grad_norm": 0.107421875,
+      "learning_rate": 9.48561946902655e-07,
+      "loss": 0.7823,
+      "step": 3273
+    },
+    {
+      "epoch": 3.6216814159292037,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 9.457964601769912e-07,
+      "loss": 0.7679,
+      "step": 3274
+    },
+    {
+      "epoch": 3.622787610619469,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 9.430309734513276e-07,
+      "loss": 0.7591,
+      "step": 3275
+    },
+    {
+      "epoch": 3.6238938053097343,
+      "grad_norm": 0.11328125,
+      "learning_rate": 9.402654867256637e-07,
+      "loss": 0.8074,
+      "step": 3276
+    },
+    {
+      "epoch": 3.625,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 9.375000000000001e-07,
+      "loss": 0.7747,
+      "step": 3277
+    },
+    {
+      "epoch": 3.6261061946902657,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 9.347345132743365e-07,
+      "loss": 0.7266,
+      "step": 3278
+    },
+    {
+      "epoch": 3.627212389380531,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.319690265486726e-07,
+      "loss": 0.7633,
+      "step": 3279
+    },
+    {
+      "epoch": 3.6283185840707963,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.292035398230089e-07,
+      "loss": 0.7993,
+      "step": 3280
+    },
+    {
+      "epoch": 3.629424778761062,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 9.264380530973452e-07,
+      "loss": 0.7843,
+      "step": 3281
+    },
+    {
+      "epoch": 3.6305309734513274,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 9.236725663716815e-07,
+      "loss": 0.7541,
+      "step": 3282
+    },
+    {
+      "epoch": 3.6316371681415927,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 9.209070796460177e-07,
+      "loss": 0.7793,
+      "step": 3283
+    },
+    {
+      "epoch": 3.6327433628318584,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 9.181415929203541e-07,
+      "loss": 0.7286,
+      "step": 3284
+    },
+    {
+      "epoch": 3.633849557522124,
+      "grad_norm": 0.10546875,
+      "learning_rate": 9.153761061946903e-07,
+      "loss": 0.7659,
+      "step": 3285
+    },
+    {
+      "epoch": 3.6349557522123894,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 9.126106194690266e-07,
+      "loss": 0.7862,
+      "step": 3286
+    },
+    {
+      "epoch": 3.6360619469026547,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 9.098451327433629e-07,
+      "loss": 0.7268,
+      "step": 3287
+    },
+    {
+      "epoch": 3.6371681415929205,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.070796460176992e-07,
+      "loss": 0.7267,
+      "step": 3288
+    },
+    {
+      "epoch": 3.6382743362831858,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 9.043141592920354e-07,
+      "loss": 0.7843,
+      "step": 3289
+    },
+    {
+      "epoch": 3.6393805309734515,
+      "grad_norm": 0.115234375,
+      "learning_rate": 9.015486725663718e-07,
+      "loss": 0.7737,
+      "step": 3290
+    },
+    {
+      "epoch": 3.640486725663717,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 8.98783185840708e-07,
+      "loss": 0.78,
+      "step": 3291
+    },
+    {
+      "epoch": 3.6415929203539825,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 8.960176991150443e-07,
+      "loss": 0.7591,
+      "step": 3292
+    },
+    {
+      "epoch": 3.642699115044248,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.932522123893806e-07,
+      "loss": 0.7642,
+      "step": 3293
+    },
+    {
+      "epoch": 3.643805309734513,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 8.904867256637169e-07,
+      "loss": 0.7467,
+      "step": 3294
+    },
+    {
+      "epoch": 3.644911504424779,
+      "grad_norm": 0.103515625,
+      "learning_rate": 8.877212389380533e-07,
+      "loss": 0.7281,
+      "step": 3295
+    },
+    {
+      "epoch": 3.646017699115044,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 8.849557522123895e-07,
+      "loss": 0.7957,
+      "step": 3296
+    },
+    {
+      "epoch": 3.64712389380531,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.821902654867258e-07,
+      "loss": 0.8408,
+      "step": 3297
+    },
+    {
+      "epoch": 3.648230088495575,
+      "grad_norm": 0.09765625,
+      "learning_rate": 8.79424778761062e-07,
+      "loss": 0.7069,
+      "step": 3298
+    },
+    {
+      "epoch": 3.649336283185841,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.766592920353984e-07,
+      "loss": 0.7436,
+      "step": 3299
+    },
+    {
+      "epoch": 3.650442477876106,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 8.738938053097345e-07,
+      "loss": 0.7438,
+      "step": 3300
+    },
+    {
+      "epoch": 3.6515486725663715,
+      "grad_norm": 0.099609375,
+      "learning_rate": 8.711283185840709e-07,
+      "loss": 0.7164,
+      "step": 3301
+    },
+    {
+      "epoch": 3.6526548672566372,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.683628318584071e-07,
+      "loss": 0.7624,
+      "step": 3302
+    },
+    {
+      "epoch": 3.6537610619469025,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 8.655973451327434e-07,
+      "loss": 0.765,
+      "step": 3303
+    },
+    {
+      "epoch": 3.6548672566371683,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 8.628318584070797e-07,
+      "loss": 0.7902,
+      "step": 3304
+    },
+    {
+      "epoch": 3.6559734513274336,
+      "grad_norm": 0.107421875,
+      "learning_rate": 8.60066371681416e-07,
+      "loss": 0.7641,
+      "step": 3305
+    },
+    {
+      "epoch": 3.6570796460176993,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.573008849557522e-07,
+      "loss": 0.7784,
+      "step": 3306
+    },
+    {
+      "epoch": 3.6581858407079646,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 8.545353982300886e-07,
+      "loss": 0.7599,
+      "step": 3307
+    },
+    {
+      "epoch": 3.65929203539823,
+      "grad_norm": 0.10546875,
+      "learning_rate": 8.517699115044248e-07,
+      "loss": 0.7526,
+      "step": 3308
+    },
+    {
+      "epoch": 3.6603982300884956,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 8.490044247787611e-07,
+      "loss": 0.8062,
+      "step": 3309
+    },
+    {
+      "epoch": 3.661504424778761,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 8.462389380530975e-07,
+      "loss": 0.7701,
+      "step": 3310
+    },
+    {
+      "epoch": 3.6626106194690267,
+      "grad_norm": 0.126953125,
+      "learning_rate": 8.434734513274337e-07,
+      "loss": 0.8195,
+      "step": 3311
+    },
+    {
+      "epoch": 3.663716814159292,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 8.4070796460177e-07,
+      "loss": 0.7711,
+      "step": 3312
+    },
+    {
+      "epoch": 3.6648230088495577,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 8.379424778761063e-07,
+      "loss": 0.7768,
+      "step": 3313
+    },
+    {
+      "epoch": 3.665929203539823,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 8.351769911504426e-07,
+      "loss": 0.7912,
+      "step": 3314
+    },
+    {
+      "epoch": 3.6670353982300883,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 8.324115044247788e-07,
+      "loss": 0.7458,
+      "step": 3315
+    },
+    {
+      "epoch": 3.668141592920354,
+      "grad_norm": 0.1015625,
+      "learning_rate": 8.296460176991152e-07,
+      "loss": 0.7717,
+      "step": 3316
+    },
+    {
+      "epoch": 3.6692477876106193,
+      "grad_norm": 0.109375,
+      "learning_rate": 8.268805309734513e-07,
+      "loss": 0.7777,
+      "step": 3317
+    },
+    {
+      "epoch": 3.670353982300885,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 8.241150442477878e-07,
+      "loss": 0.7946,
+      "step": 3318
+    },
+    {
+      "epoch": 3.6714601769911503,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 8.213495575221239e-07,
+      "loss": 0.7773,
+      "step": 3319
+    },
+    {
+      "epoch": 3.672566371681416,
+      "grad_norm": 0.111328125,
+      "learning_rate": 8.185840707964602e-07,
+      "loss": 0.801,
+      "step": 3320
+    },
+    {
+      "epoch": 3.6736725663716814,
+      "grad_norm": 0.13671875,
+      "learning_rate": 8.158185840707965e-07,
+      "loss": 0.7738,
+      "step": 3321
+    },
+    {
+      "epoch": 3.6747787610619467,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 8.130530973451328e-07,
+      "loss": 0.7937,
+      "step": 3322
+    },
+    {
+      "epoch": 3.6758849557522124,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.10287610619469e-07,
+      "loss": 0.7997,
+      "step": 3323
+    },
+    {
+      "epoch": 3.676991150442478,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 8.075221238938054e-07,
+      "loss": 0.7684,
+      "step": 3324
+    },
+    {
+      "epoch": 3.6780973451327434,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.047566371681416e-07,
+      "loss": 0.7853,
+      "step": 3325
+    },
+    {
+      "epoch": 3.6792035398230087,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 8.019911504424779e-07,
+      "loss": 0.8065,
+      "step": 3326
+    },
+    {
+      "epoch": 3.6803097345132745,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 7.992256637168143e-07,
+      "loss": 0.7538,
+      "step": 3327
+    },
+    {
+      "epoch": 3.6814159292035398,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.964601769911505e-07,
+      "loss": 0.7674,
+      "step": 3328
+    },
+    {
+      "epoch": 3.682522123893805,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 7.936946902654868e-07,
+      "loss": 0.7552,
+      "step": 3329
+    },
+    {
+      "epoch": 3.683628318584071,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 7.909292035398231e-07,
+      "loss": 0.7853,
+      "step": 3330
+    },
+    {
+      "epoch": 3.6847345132743365,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.881637168141594e-07,
+      "loss": 0.7873,
+      "step": 3331
+    },
+    {
+      "epoch": 3.685840707964602,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.853982300884956e-07,
+      "loss": 0.7665,
+      "step": 3332
+    },
+    {
+      "epoch": 3.686946902654867,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 7.82632743362832e-07,
+      "loss": 0.7901,
+      "step": 3333
+    },
+    {
+      "epoch": 3.688053097345133,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 7.798672566371682e-07,
+      "loss": 0.7743,
+      "step": 3334
+    },
+    {
+      "epoch": 3.689159292035398,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.771017699115045e-07,
+      "loss": 0.7681,
+      "step": 3335
+    },
+    {
+      "epoch": 3.6902654867256635,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.743362831858408e-07,
+      "loss": 0.7793,
+      "step": 3336
+    },
+    {
+      "epoch": 3.691371681415929,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 7.715707964601771e-07,
+      "loss": 0.7195,
+      "step": 3337
+    },
+    {
+      "epoch": 3.692477876106195,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.688053097345132e-07,
+      "loss": 0.7547,
+      "step": 3338
+    },
+    {
+      "epoch": 3.6935840707964602,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 7.660398230088497e-07,
+      "loss": 0.7535,
+      "step": 3339
+    },
+    {
+      "epoch": 3.6946902654867255,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 7.632743362831858e-07,
+      "loss": 0.7995,
+      "step": 3340
+    },
+    {
+      "epoch": 3.6957964601769913,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.605088495575221e-07,
+      "loss": 0.7566,
+      "step": 3341
+    },
+    {
+      "epoch": 3.6969026548672566,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 7.577433628318585e-07,
+      "loss": 0.7889,
+      "step": 3342
+    },
+    {
+      "epoch": 3.698008849557522,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 7.549778761061947e-07,
+      "loss": 0.7561,
+      "step": 3343
+    },
+    {
+      "epoch": 3.6991150442477876,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 7.522123893805311e-07,
+      "loss": 0.7623,
+      "step": 3344
+    },
+    {
+      "epoch": 3.7002212389380533,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 7.494469026548673e-07,
+      "loss": 0.7952,
+      "step": 3345
+    },
+    {
+      "epoch": 3.7013274336283186,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 7.466814159292036e-07,
+      "loss": 0.7769,
+      "step": 3346
+    },
+    {
+      "epoch": 3.702433628318584,
+      "grad_norm": 0.115234375,
+      "learning_rate": 7.439159292035399e-07,
+      "loss": 0.7138,
+      "step": 3347
+    },
+    {
+      "epoch": 3.7035398230088497,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 7.411504424778762e-07,
+      "loss": 0.8225,
+      "step": 3348
+    },
+    {
+      "epoch": 3.704646017699115,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 7.383849557522124e-07,
+      "loss": 0.775,
+      "step": 3349
+    },
+    {
+      "epoch": 3.7057522123893807,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 7.356194690265488e-07,
+      "loss": 0.7498,
+      "step": 3350
+    },
+    {
+      "epoch": 3.706858407079646,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.32853982300885e-07,
+      "loss": 0.7434,
+      "step": 3351
+    },
+    {
+      "epoch": 3.7079646017699117,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 7.300884955752213e-07,
+      "loss": 0.7705,
+      "step": 3352
+    },
+    {
+      "epoch": 3.709070796460177,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.273230088495576e-07,
+      "loss": 0.8518,
+      "step": 3353
+    },
+    {
+      "epoch": 3.7101769911504423,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.245575221238939e-07,
+      "loss": 0.7707,
+      "step": 3354
+    },
+    {
+      "epoch": 3.711283185840708,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 7.217920353982301e-07,
+      "loss": 0.7706,
+      "step": 3355
+    },
+    {
+      "epoch": 3.7123893805309733,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 7.190265486725665e-07,
+      "loss": 0.7248,
+      "step": 3356
+    },
+    {
+      "epoch": 3.713495575221239,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.162610619469026e-07,
+      "loss": 0.8013,
+      "step": 3357
+    },
+    {
+      "epoch": 3.7146017699115044,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 7.13495575221239e-07,
+      "loss": 0.7465,
+      "step": 3358
+    },
+    {
+      "epoch": 3.71570796460177,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 7.107300884955754e-07,
+      "loss": 0.7879,
+      "step": 3359
+    },
+    {
+      "epoch": 3.7168141592920354,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.079646017699115e-07,
+      "loss": 0.7226,
+      "step": 3360
+    },
+    {
+      "epoch": 3.7179203539823007,
+      "grad_norm": 0.1015625,
+      "learning_rate": 7.05199115044248e-07,
+      "loss": 0.7876,
+      "step": 3361
+    },
+    {
+      "epoch": 3.7190265486725664,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 7.024336283185841e-07,
+      "loss": 0.7764,
+      "step": 3362
+    },
+    {
+      "epoch": 3.7201327433628317,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.996681415929204e-07,
+      "loss": 0.7867,
+      "step": 3363
+    },
+    {
+      "epoch": 3.7212389380530975,
+      "grad_norm": 0.11328125,
+      "learning_rate": 6.969026548672566e-07,
+      "loss": 0.7741,
+      "step": 3364
+    },
+    {
+      "epoch": 3.7223451327433628,
+      "grad_norm": 0.1015625,
+      "learning_rate": 6.94137168141593e-07,
+      "loss": 0.7408,
+      "step": 3365
+    },
+    {
+      "epoch": 3.7234513274336285,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.913716814159292e-07,
+      "loss": 0.7406,
+      "step": 3366
+    },
+    {
+      "epoch": 3.724557522123894,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.886061946902656e-07,
+      "loss": 0.7508,
+      "step": 3367
+    },
+    {
+      "epoch": 3.725663716814159,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 6.858407079646018e-07,
+      "loss": 0.7754,
+      "step": 3368
+    },
+    {
+      "epoch": 3.726769911504425,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.830752212389381e-07,
+      "loss": 0.7929,
+      "step": 3369
+    },
+    {
+      "epoch": 3.72787610619469,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 6.803097345132744e-07,
+      "loss": 0.7341,
+      "step": 3370
+    },
+    {
+      "epoch": 3.728982300884956,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.775442477876107e-07,
+      "loss": 0.7528,
+      "step": 3371
+    },
+    {
+      "epoch": 3.730088495575221,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.747787610619469e-07,
+      "loss": 0.8679,
+      "step": 3372
+    },
+    {
+      "epoch": 3.731194690265487,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.720132743362833e-07,
+      "loss": 0.7369,
+      "step": 3373
+    },
+    {
+      "epoch": 3.732300884955752,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.692477876106196e-07,
+      "loss": 0.7855,
+      "step": 3374
+    },
+    {
+      "epoch": 3.7334070796460175,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.664823008849558e-07,
+      "loss": 0.7358,
+      "step": 3375
+    },
+    {
+      "epoch": 3.734513274336283,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.637168141592922e-07,
+      "loss": 0.7425,
+      "step": 3376
+    },
+    {
+      "epoch": 3.7356194690265485,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.609513274336284e-07,
+      "loss": 0.7889,
+      "step": 3377
+    },
+    {
+      "epoch": 3.7367256637168142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.581858407079647e-07,
+      "loss": 0.8149,
+      "step": 3378
+    },
+    {
+      "epoch": 3.7378318584070795,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 6.554203539823009e-07,
+      "loss": 0.8119,
+      "step": 3379
+    },
+    {
+      "epoch": 3.7389380530973453,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 6.526548672566373e-07,
+      "loss": 0.7998,
+      "step": 3380
+    },
+    {
+      "epoch": 3.7400442477876106,
+      "grad_norm": 0.111328125,
+      "learning_rate": 6.498893805309734e-07,
+      "loss": 0.7325,
+      "step": 3381
+    },
+    {
+      "epoch": 3.741150442477876,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.471238938053098e-07,
+      "loss": 0.7372,
+      "step": 3382
+    },
+    {
+      "epoch": 3.7422566371681416,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.44358407079646e-07,
+      "loss": 0.7462,
+      "step": 3383
+    },
+    {
+      "epoch": 3.7433628318584073,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.415929203539823e-07,
+      "loss": 0.7652,
+      "step": 3384
+    },
+    {
+      "epoch": 3.7444690265486726,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 6.388274336283186e-07,
+      "loss": 0.7462,
+      "step": 3385
+    },
+    {
+      "epoch": 3.745575221238938,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.360619469026549e-07,
+      "loss": 0.7574,
+      "step": 3386
+    },
+    {
+      "epoch": 3.7466814159292037,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 6.332964601769911e-07,
+      "loss": 0.8215,
+      "step": 3387
+    },
+    {
+      "epoch": 3.747787610619469,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 6.305309734513275e-07,
+      "loss": 0.7772,
+      "step": 3388
+    },
+    {
+      "epoch": 3.7488938053097343,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 6.277654867256637e-07,
+      "loss": 0.7491,
+      "step": 3389
+    },
+    {
+      "epoch": 3.75,
+      "grad_norm": 0.109375,
+      "learning_rate": 6.25e-07,
+      "loss": 0.7195,
+      "step": 3390
+    },
+    {
+      "epoch": 3.7511061946902657,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.222345132743363e-07,
+      "loss": 0.7868,
+      "step": 3391
+    },
+    {
+      "epoch": 3.752212389380531,
+      "grad_norm": 0.107421875,
+      "learning_rate": 6.194690265486726e-07,
+      "loss": 0.767,
+      "step": 3392
+    },
+    {
+      "epoch": 3.7533185840707963,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 6.167035398230089e-07,
+      "loss": 0.7578,
+      "step": 3393
+    },
+    {
+      "epoch": 3.754424778761062,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 6.139380530973452e-07,
+      "loss": 0.7519,
+      "step": 3394
+    },
+    {
+      "epoch": 3.7555309734513274,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 6.111725663716814e-07,
+      "loss": 0.786,
+      "step": 3395
+    },
+    {
+      "epoch": 3.7566371681415927,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 6.084070796460178e-07,
+      "loss": 0.7297,
+      "step": 3396
+    },
+    {
+      "epoch": 3.7577433628318584,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 6.05641592920354e-07,
+      "loss": 0.7849,
+      "step": 3397
+    },
+    {
+      "epoch": 3.758849557522124,
+      "grad_norm": 0.109375,
+      "learning_rate": 6.028761061946903e-07,
+      "loss": 0.793,
+      "step": 3398
+    },
+    {
+      "epoch": 3.7599557522123894,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 6.001106194690267e-07,
+      "loss": 0.7566,
+      "step": 3399
+    },
+    {
+      "epoch": 3.7610619469026547,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 5.973451327433629e-07,
+      "loss": 0.7404,
+      "step": 3400
+    },
+    {
+      "epoch": 3.7621681415929205,
+      "grad_norm": 0.123046875,
+      "learning_rate": 5.945796460176992e-07,
+      "loss": 0.7673,
+      "step": 3401
+    },
+    {
+      "epoch": 3.7632743362831858,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 5.918141592920355e-07,
+      "loss": 0.7683,
+      "step": 3402
+    },
+    {
+      "epoch": 3.7643805309734515,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 5.890486725663717e-07,
+      "loss": 0.7572,
+      "step": 3403
+    },
+    {
+      "epoch": 3.765486725663717,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 5.86283185840708e-07,
+      "loss": 0.7719,
+      "step": 3404
+    },
+    {
+      "epoch": 3.7665929203539825,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 5.835176991150443e-07,
+      "loss": 0.7635,
+      "step": 3405
+    },
+    {
+      "epoch": 3.767699115044248,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 5.807522123893806e-07,
+      "loss": 0.8151,
+      "step": 3406
+    },
+    {
+      "epoch": 3.768805309734513,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 5.779867256637168e-07,
+      "loss": 0.7845,
+      "step": 3407
+    },
+    {
+      "epoch": 3.769911504424779,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.752212389380532e-07,
+      "loss": 0.7951,
+      "step": 3408
+    },
+    {
+      "epoch": 3.771017699115044,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.724557522123894e-07,
+      "loss": 0.7838,
+      "step": 3409
+    },
+    {
+      "epoch": 3.77212389380531,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.696902654867256e-07,
+      "loss": 0.6889,
+      "step": 3410
+    },
+    {
+      "epoch": 3.773230088495575,
+      "grad_norm": 0.111328125,
+      "learning_rate": 5.66924778761062e-07,
+      "loss": 0.7545,
+      "step": 3411
+    },
+    {
+      "epoch": 3.774336283185841,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 5.641592920353982e-07,
+      "loss": 0.8127,
+      "step": 3412
+    },
+    {
+      "epoch": 3.775442477876106,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 5.613938053097345e-07,
+      "loss": 0.7462,
+      "step": 3413
+    },
+    {
+      "epoch": 3.7765486725663715,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 5.586283185840708e-07,
+      "loss": 0.7565,
+      "step": 3414
+    },
+    {
+      "epoch": 3.7776548672566372,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 5.558628318584071e-07,
+      "loss": 0.7894,
+      "step": 3415
+    },
+    {
+      "epoch": 3.7787610619469025,
+      "grad_norm": 0.099609375,
+      "learning_rate": 5.530973451327435e-07,
+      "loss": 0.7328,
+      "step": 3416
+    },
+    {
+      "epoch": 3.7798672566371683,
+      "grad_norm": 0.107421875,
+      "learning_rate": 5.503318584070797e-07,
+      "loss": 0.7369,
+      "step": 3417
+    },
+    {
+      "epoch": 3.7809734513274336,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.47566371681416e-07,
+      "loss": 0.7763,
+      "step": 3418
+    },
+    {
+      "epoch": 3.7820796460176993,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 5.448008849557523e-07,
+      "loss": 0.7564,
+      "step": 3419
+    },
+    {
+      "epoch": 3.7831858407079646,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 5.420353982300886e-07,
+      "loss": 0.8017,
+      "step": 3420
+    },
+    {
+      "epoch": 3.78429203539823,
+      "grad_norm": 0.123046875,
+      "learning_rate": 5.392699115044248e-07,
+      "loss": 0.7912,
+      "step": 3421
+    },
+    {
+      "epoch": 3.7853982300884956,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 5.365044247787611e-07,
+      "loss": 0.7862,
+      "step": 3422
+    },
+    {
+      "epoch": 3.786504424778761,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 5.337389380530974e-07,
+      "loss": 0.7979,
+      "step": 3423
+    },
+    {
+      "epoch": 3.7876106194690267,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 5.309734513274336e-07,
+      "loss": 0.7697,
+      "step": 3424
+    },
+    {
+      "epoch": 3.788716814159292,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.2820796460177e-07,
+      "loss": 0.724,
+      "step": 3425
+    },
+    {
+      "epoch": 3.7898230088495577,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.254424778761062e-07,
+      "loss": 0.763,
+      "step": 3426
+    },
+    {
+      "epoch": 3.790929203539823,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 5.226769911504425e-07,
+      "loss": 0.8013,
+      "step": 3427
+    },
+    {
+      "epoch": 3.7920353982300883,
+      "grad_norm": 0.1328125,
+      "learning_rate": 5.199115044247788e-07,
+      "loss": 0.8185,
+      "step": 3428
+    },
+    {
+      "epoch": 3.793141592920354,
+      "grad_norm": 0.10546875,
+      "learning_rate": 5.171460176991151e-07,
+      "loss": 0.7786,
+      "step": 3429
+    },
+    {
+      "epoch": 3.7942477876106193,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.143805309734513e-07,
+      "loss": 0.8212,
+      "step": 3430
+    },
+    {
+      "epoch": 3.795353982300885,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.116150442477877e-07,
+      "loss": 0.7802,
+      "step": 3431
+    },
+    {
+      "epoch": 3.7964601769911503,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.08849557522124e-07,
+      "loss": 0.7415,
+      "step": 3432
+    },
+    {
+      "epoch": 3.797566371681416,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 5.060840707964602e-07,
+      "loss": 0.7847,
+      "step": 3433
+    },
+    {
+      "epoch": 3.7986725663716814,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.033185840707965e-07,
+      "loss": 0.8451,
+      "step": 3434
+    },
+    {
+      "epoch": 3.7997787610619467,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.005530973451328e-07,
+      "loss": 0.7168,
+      "step": 3435
+    },
+    {
+      "epoch": 3.8008849557522124,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 4.97787610619469e-07,
+      "loss": 0.7505,
+      "step": 3436
+    },
+    {
+      "epoch": 3.801991150442478,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.950221238938054e-07,
+      "loss": 0.7266,
+      "step": 3437
+    },
+    {
+      "epoch": 3.8030973451327434,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.922566371681416e-07,
+      "loss": 0.7283,
+      "step": 3438
+    },
+    {
+      "epoch": 3.8042035398230087,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.89491150442478e-07,
+      "loss": 0.7988,
+      "step": 3439
+    },
+    {
+      "epoch": 3.8053097345132745,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.867256637168142e-07,
+      "loss": 0.7461,
+      "step": 3440
+    },
+    {
+      "epoch": 3.8064159292035398,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.839601769911504e-07,
+      "loss": 0.7763,
+      "step": 3441
+    },
+    {
+      "epoch": 3.807522123893805,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 4.811946902654868e-07,
+      "loss": 0.7649,
+      "step": 3442
+    },
+    {
+      "epoch": 3.808628318584071,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.78429203539823e-07,
+      "loss": 0.7712,
+      "step": 3443
+    },
+    {
+      "epoch": 3.8097345132743365,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.756637168141593e-07,
+      "loss": 0.7553,
+      "step": 3444
+    },
+    {
+      "epoch": 3.810840707964602,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 4.728982300884956e-07,
+      "loss": 0.7509,
+      "step": 3445
+    },
+    {
+      "epoch": 3.811946902654867,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.7013274336283184e-07,
+      "loss": 0.7929,
+      "step": 3446
+    },
+    {
+      "epoch": 3.813053097345133,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 4.6736725663716823e-07,
+      "loss": 0.7374,
+      "step": 3447
+    },
+    {
+      "epoch": 3.814159292035398,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 4.6460176991150447e-07,
+      "loss": 0.7663,
+      "step": 3448
+    },
+    {
+      "epoch": 3.8152654867256635,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.6183628318584075e-07,
+      "loss": 0.7745,
+      "step": 3449
+    },
+    {
+      "epoch": 3.816371681415929,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.5907079646017704e-07,
+      "loss": 0.7571,
+      "step": 3450
+    },
+    {
+      "epoch": 3.817477876106195,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.563053097345133e-07,
+      "loss": 0.8059,
+      "step": 3451
+    },
+    {
+      "epoch": 3.8185840707964602,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 4.535398230088496e-07,
+      "loss": 0.785,
+      "step": 3452
+    },
+    {
+      "epoch": 3.8196902654867255,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 4.507743362831859e-07,
+      "loss": 0.8131,
+      "step": 3453
+    },
+    {
+      "epoch": 3.8207964601769913,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 4.480088495575222e-07,
+      "loss": 0.8001,
+      "step": 3454
+    },
+    {
+      "epoch": 3.8219026548672566,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 4.4524336283185846e-07,
+      "loss": 0.7726,
+      "step": 3455
+    },
+    {
+      "epoch": 3.823008849557522,
+      "grad_norm": 0.107421875,
+      "learning_rate": 4.4247787610619474e-07,
+      "loss": 0.7942,
+      "step": 3456
+    },
+    {
+      "epoch": 3.8241150442477876,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.39712389380531e-07,
+      "loss": 0.7572,
+      "step": 3457
+    },
+    {
+      "epoch": 3.8252212389380533,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.3694690265486726e-07,
+      "loss": 0.8153,
+      "step": 3458
+    },
+    {
+      "epoch": 3.8263274336283186,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 4.3418141592920355e-07,
+      "loss": 0.7322,
+      "step": 3459
+    },
+    {
+      "epoch": 3.827433628318584,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 4.3141592920353983e-07,
+      "loss": 0.7983,
+      "step": 3460
+    },
+    {
+      "epoch": 3.8285398230088497,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.286504424778761e-07,
+      "loss": 0.7811,
+      "step": 3461
+    },
+    {
+      "epoch": 3.829646017699115,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.258849557522124e-07,
+      "loss": 0.8147,
+      "step": 3462
+    },
+    {
+      "epoch": 3.8307522123893807,
+      "grad_norm": 0.115234375,
+      "learning_rate": 4.2311946902654874e-07,
+      "loss": 0.8276,
+      "step": 3463
+    },
+    {
+      "epoch": 3.831858407079646,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.20353982300885e-07,
+      "loss": 0.7804,
+      "step": 3464
+    },
+    {
+      "epoch": 3.8329646017699117,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 4.175884955752213e-07,
+      "loss": 0.8356,
+      "step": 3465
+    },
+    {
+      "epoch": 3.834070796460177,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 4.148230088495576e-07,
+      "loss": 0.7462,
+      "step": 3466
+    },
+    {
+      "epoch": 3.8351769911504423,
+      "grad_norm": 0.1328125,
+      "learning_rate": 4.120575221238939e-07,
+      "loss": 0.8482,
+      "step": 3467
+    },
+    {
+      "epoch": 3.836283185840708,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.092920353982301e-07,
+      "loss": 0.7886,
+      "step": 3468
+    },
+    {
+      "epoch": 3.8373893805309733,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 4.065265486725664e-07,
+      "loss": 0.7656,
+      "step": 3469
+    },
+    {
+      "epoch": 3.838495575221239,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 4.037610619469027e-07,
+      "loss": 0.7435,
+      "step": 3470
+    },
+    {
+      "epoch": 3.8396017699115044,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.0099557522123896e-07,
+      "loss": 0.7621,
+      "step": 3471
+    },
+    {
+      "epoch": 3.84070796460177,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.9823008849557525e-07,
+      "loss": 0.7819,
+      "step": 3472
+    },
+    {
+      "epoch": 3.8418141592920354,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 3.9546460176991153e-07,
+      "loss": 0.7828,
+      "step": 3473
+    },
+    {
+      "epoch": 3.8429203539823007,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.926991150442478e-07,
+      "loss": 0.7652,
+      "step": 3474
+    },
+    {
+      "epoch": 3.8440265486725664,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 3.899336283185841e-07,
+      "loss": 0.7957,
+      "step": 3475
+    },
+    {
+      "epoch": 3.8451327433628317,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 3.871681415929204e-07,
+      "loss": 0.7634,
+      "step": 3476
+    },
+    {
+      "epoch": 3.8462389380530975,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 3.844026548672566e-07,
+      "loss": 0.7718,
+      "step": 3477
+    },
+    {
+      "epoch": 3.8473451327433628,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 3.816371681415929e-07,
+      "loss": 0.7728,
+      "step": 3478
+    },
+    {
+      "epoch": 3.8484513274336285,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 3.7887168141592924e-07,
+      "loss": 0.7569,
+      "step": 3479
+    },
+    {
+      "epoch": 3.849557522123894,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.7610619469026553e-07,
+      "loss": 0.7745,
+      "step": 3480
+    },
+    {
+      "epoch": 3.850663716814159,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.733407079646018e-07,
+      "loss": 0.8341,
+      "step": 3481
+    },
+    {
+      "epoch": 3.851769911504425,
+      "grad_norm": 0.109375,
+      "learning_rate": 3.705752212389381e-07,
+      "loss": 0.7309,
+      "step": 3482
+    },
+    {
+      "epoch": 3.85287610619469,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.678097345132744e-07,
+      "loss": 0.7498,
+      "step": 3483
+    },
+    {
+      "epoch": 3.853982300884956,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 3.6504424778761067e-07,
+      "loss": 0.7414,
+      "step": 3484
+    },
+    {
+      "epoch": 3.855088495575221,
+      "grad_norm": 0.125,
+      "learning_rate": 3.6227876106194695e-07,
+      "loss": 0.7439,
+      "step": 3485
+    },
+    {
+      "epoch": 3.856194690265487,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.5951327433628324e-07,
+      "loss": 0.8091,
+      "step": 3486
+    },
+    {
+      "epoch": 3.857300884955752,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.567477876106195e-07,
+      "loss": 0.7728,
+      "step": 3487
+    },
+    {
+      "epoch": 3.8584070796460175,
+      "grad_norm": 0.103515625,
+      "learning_rate": 3.5398230088495575e-07,
+      "loss": 0.7647,
+      "step": 3488
+    },
+    {
+      "epoch": 3.859513274336283,
+      "grad_norm": 0.119140625,
+      "learning_rate": 3.5121681415929204e-07,
+      "loss": 0.7726,
+      "step": 3489
+    },
+    {
+      "epoch": 3.8606194690265485,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 3.484513274336283e-07,
+      "loss": 0.8064,
+      "step": 3490
+    },
+    {
+      "epoch": 3.8617256637168142,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 3.456858407079646e-07,
+      "loss": 0.7224,
+      "step": 3491
+    },
+    {
+      "epoch": 3.8628318584070795,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 3.429203539823009e-07,
+      "loss": 0.7417,
+      "step": 3492
+    },
+    {
+      "epoch": 3.8639380530973453,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.401548672566372e-07,
+      "loss": 0.7696,
+      "step": 3493
+    },
+    {
+      "epoch": 3.8650442477876106,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 3.3738938053097346e-07,
+      "loss": 0.7677,
+      "step": 3494
+    },
+    {
+      "epoch": 3.866150442477876,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.346238938053098e-07,
+      "loss": 0.7532,
+      "step": 3495
+    },
+    {
+      "epoch": 3.8672566371681416,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.318584070796461e-07,
+      "loss": 0.7974,
+      "step": 3496
+    },
+    {
+      "epoch": 3.8683628318584073,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 3.2909292035398237e-07,
+      "loss": 0.7544,
+      "step": 3497
+    },
+    {
+      "epoch": 3.8694690265486726,
+      "grad_norm": 0.10546875,
+      "learning_rate": 3.2632743362831865e-07,
+      "loss": 0.753,
+      "step": 3498
+    },
+    {
+      "epoch": 3.870575221238938,
+      "grad_norm": 0.1015625,
+      "learning_rate": 3.235619469026549e-07,
+      "loss": 0.7775,
+      "step": 3499
+    },
+    {
+      "epoch": 3.8716814159292037,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 3.2079646017699117e-07,
+      "loss": 0.7652,
+      "step": 3500
+    },
+    {
+      "epoch": 3.872787610619469,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.1803097345132746e-07,
+      "loss": 0.8548,
+      "step": 3501
+    },
+    {
+      "epoch": 3.8738938053097343,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.1526548672566374e-07,
+      "loss": 0.777,
+      "step": 3502
+    },
+    {
+      "epoch": 3.875,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.125e-07,
+      "loss": 0.8376,
+      "step": 3503
+    },
+    {
+      "epoch": 3.8761061946902657,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.097345132743363e-07,
+      "loss": 0.7782,
+      "step": 3504
+    },
+    {
+      "epoch": 3.877212389380531,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 3.069690265486726e-07,
+      "loss": 0.7666,
+      "step": 3505
+    },
+    {
+      "epoch": 3.8783185840707963,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.042035398230089e-07,
+      "loss": 0.7697,
+      "step": 3506
+    },
+    {
+      "epoch": 3.879424778761062,
+      "grad_norm": 0.134765625,
+      "learning_rate": 3.0143805309734516e-07,
+      "loss": 0.8063,
+      "step": 3507
+    },
+    {
+      "epoch": 3.8805309734513274,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.9867256637168145e-07,
+      "loss": 0.7782,
+      "step": 3508
+    },
+    {
+      "epoch": 3.8816371681415927,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 2.9590707964601773e-07,
+      "loss": 0.775,
+      "step": 3509
+    },
+    {
+      "epoch": 3.8827433628318584,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.93141592920354e-07,
+      "loss": 0.7697,
+      "step": 3510
+    },
+    {
+      "epoch": 3.883849557522124,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.903761061946903e-07,
+      "loss": 0.7791,
+      "step": 3511
+    },
+    {
+      "epoch": 3.8849557522123894,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.876106194690266e-07,
+      "loss": 0.7329,
+      "step": 3512
+    },
+    {
+      "epoch": 3.8860619469026547,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.848451327433628e-07,
+      "loss": 0.7207,
+      "step": 3513
+    },
+    {
+      "epoch": 3.8871681415929205,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.820796460176991e-07,
+      "loss": 0.7888,
+      "step": 3514
+    },
+    {
+      "epoch": 3.8882743362831858,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.793141592920354e-07,
+      "loss": 0.7652,
+      "step": 3515
+    },
+    {
+      "epoch": 3.8893805309734515,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.7654867256637173e-07,
+      "loss": 0.7872,
+      "step": 3516
+    },
+    {
+      "epoch": 3.890486725663717,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.73783185840708e-07,
+      "loss": 0.8109,
+      "step": 3517
+    },
+    {
+      "epoch": 3.8915929203539825,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.710176991150443e-07,
+      "loss": 0.8481,
+      "step": 3518
+    },
+    {
+      "epoch": 3.892699115044248,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.6825221238938053e-07,
+      "loss": 0.7889,
+      "step": 3519
+    },
+    {
+      "epoch": 3.893805309734513,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.654867256637168e-07,
+      "loss": 0.7418,
+      "step": 3520
+    },
+    {
+      "epoch": 3.894911504424779,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.627212389380531e-07,
+      "loss": 0.7506,
+      "step": 3521
+    },
+    {
+      "epoch": 3.896017699115044,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.599557522123894e-07,
+      "loss": 0.78,
+      "step": 3522
+    },
+    {
+      "epoch": 3.89712389380531,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.5719026548672567e-07,
+      "loss": 0.7726,
+      "step": 3523
+    },
+    {
+      "epoch": 3.898230088495575,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.54424778761062e-07,
+      "loss": 0.7595,
+      "step": 3524
+    },
+    {
+      "epoch": 3.899336283185841,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.5165929203539824e-07,
+      "loss": 0.8189,
+      "step": 3525
+    },
+    {
+      "epoch": 3.900442477876106,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.488938053097345e-07,
+      "loss": 0.7426,
+      "step": 3526
+    },
+    {
+      "epoch": 3.9015486725663715,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.461283185840708e-07,
+      "loss": 0.7856,
+      "step": 3527
+    },
+    {
+      "epoch": 3.9026548672566372,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.433628318584071e-07,
+      "loss": 0.7655,
+      "step": 3528
+    },
+    {
+      "epoch": 3.9037610619469025,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 2.405973451327434e-07,
+      "loss": 0.7007,
+      "step": 3529
+    },
+    {
+      "epoch": 3.9048672566371683,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.3783185840707966e-07,
+      "loss": 0.7861,
+      "step": 3530
+    },
+    {
+      "epoch": 3.9059734513274336,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.3506637168141592e-07,
+      "loss": 0.816,
+      "step": 3531
+    },
+    {
+      "epoch": 3.9070796460176993,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 2.3230088495575223e-07,
+      "loss": 0.7619,
+      "step": 3532
+    },
+    {
+      "epoch": 3.9081858407079646,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.2953539823008852e-07,
+      "loss": 0.8237,
+      "step": 3533
+    },
+    {
+      "epoch": 3.90929203539823,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.267699115044248e-07,
+      "loss": 0.7601,
+      "step": 3534
+    },
+    {
+      "epoch": 3.9103982300884956,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.240044247787611e-07,
+      "loss": 0.7206,
+      "step": 3535
+    },
+    {
+      "epoch": 3.911504424778761,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.2123893805309737e-07,
+      "loss": 0.7475,
+      "step": 3536
+    },
+    {
+      "epoch": 3.9126106194690267,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.1847345132743363e-07,
+      "loss": 0.7761,
+      "step": 3537
+    },
+    {
+      "epoch": 3.913716814159292,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.1570796460176992e-07,
+      "loss": 0.7445,
+      "step": 3538
+    },
+    {
+      "epoch": 3.9148230088495577,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.129424778761062e-07,
+      "loss": 0.763,
+      "step": 3539
+    },
+    {
+      "epoch": 3.915929203539823,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.101769911504425e-07,
+      "loss": 0.8248,
+      "step": 3540
+    },
+    {
+      "epoch": 3.9170353982300883,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 2.074115044247788e-07,
+      "loss": 0.7363,
+      "step": 3541
+    },
+    {
+      "epoch": 3.918141592920354,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.0464601769911505e-07,
+      "loss": 0.8015,
+      "step": 3542
+    },
+    {
+      "epoch": 3.9192477876106193,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.0188053097345134e-07,
+      "loss": 0.7846,
+      "step": 3543
+    },
+    {
+      "epoch": 3.920353982300885,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.9911504424778762e-07,
+      "loss": 0.7781,
+      "step": 3544
+    },
+    {
+      "epoch": 3.9214601769911503,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.963495575221239e-07,
+      "loss": 0.7601,
+      "step": 3545
+    },
+    {
+      "epoch": 3.922566371681416,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.935840707964602e-07,
+      "loss": 0.7387,
+      "step": 3546
+    },
+    {
+      "epoch": 3.9236725663716814,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.9081858407079645e-07,
+      "loss": 0.743,
+      "step": 3547
+    },
+    {
+      "epoch": 3.9247787610619467,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.8805309734513276e-07,
+      "loss": 0.7477,
+      "step": 3548
+    },
+    {
+      "epoch": 3.9258849557522124,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.8528761061946905e-07,
+      "loss": 0.7312,
+      "step": 3549
+    },
+    {
+      "epoch": 3.926991150442478,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.8252212389380533e-07,
+      "loss": 0.7648,
+      "step": 3550
+    },
+    {
+      "epoch": 3.9280973451327434,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.7975663716814162e-07,
+      "loss": 0.7756,
+      "step": 3551
+    },
+    {
+      "epoch": 3.9292035398230087,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.7699115044247788e-07,
+      "loss": 0.7435,
+      "step": 3552
+    },
+    {
+      "epoch": 3.9303097345132745,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.7422566371681416e-07,
+      "loss": 0.7355,
+      "step": 3553
+    },
+    {
+      "epoch": 3.9314159292035398,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.7146017699115045e-07,
+      "loss": 0.7788,
+      "step": 3554
+    },
+    {
+      "epoch": 3.932522123893805,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.6869469026548673e-07,
+      "loss": 0.7776,
+      "step": 3555
+    },
+    {
+      "epoch": 3.933628318584071,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 1.6592920353982304e-07,
+      "loss": 0.7268,
+      "step": 3556
+    },
+    {
+      "epoch": 3.9347345132743365,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.6316371681415933e-07,
+      "loss": 0.7926,
+      "step": 3557
+    },
+    {
+      "epoch": 3.935840707964602,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.6039823008849559e-07,
+      "loss": 0.7526,
+      "step": 3558
+    },
+    {
+      "epoch": 3.936946902654867,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.5763274336283187e-07,
+      "loss": 0.787,
+      "step": 3559
+    },
+    {
+      "epoch": 3.938053097345133,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.5486725663716816e-07,
+      "loss": 0.7571,
+      "step": 3560
+    },
+    {
+      "epoch": 3.939159292035398,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.5210176991150444e-07,
+      "loss": 0.7465,
+      "step": 3561
+    },
+    {
+      "epoch": 3.9402654867256635,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.4933628318584072e-07,
+      "loss": 0.801,
+      "step": 3562
+    },
+    {
+      "epoch": 3.941371681415929,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.46570796460177e-07,
+      "loss": 0.7511,
+      "step": 3563
+    },
+    {
+      "epoch": 3.942477876106195,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.438053097345133e-07,
+      "loss": 0.7446,
+      "step": 3564
+    },
+    {
+      "epoch": 3.9435840707964602,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.4103982300884955e-07,
+      "loss": 0.7334,
+      "step": 3565
+    },
+    {
+      "epoch": 3.9446902654867255,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.3827433628318586e-07,
+      "loss": 0.7963,
+      "step": 3566
+    },
+    {
+      "epoch": 3.9457964601769913,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.3550884955752215e-07,
+      "loss": 0.7607,
+      "step": 3567
+    },
+    {
+      "epoch": 3.9469026548672566,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.327433628318584e-07,
+      "loss": 0.7825,
+      "step": 3568
+    },
+    {
+      "epoch": 3.948008849557522,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.299778761061947e-07,
+      "loss": 0.7628,
+      "step": 3569
+    },
+    {
+      "epoch": 3.9491150442477876,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.27212389380531e-07,
+      "loss": 0.7761,
+      "step": 3570
+    },
+    {
+      "epoch": 3.9502212389380533,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.2444690265486726e-07,
+      "loss": 0.75,
+      "step": 3571
+    },
+    {
+      "epoch": 3.9513274336283186,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.2168141592920355e-07,
+      "loss": 0.7487,
+      "step": 3572
+    },
+    {
+      "epoch": 3.952433628318584,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.1891592920353983e-07,
+      "loss": 0.7636,
+      "step": 3573
+    },
+    {
+      "epoch": 3.9535398230088497,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.1615044247787612e-07,
+      "loss": 0.7855,
+      "step": 3574
+    },
+    {
+      "epoch": 3.954646017699115,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.133849557522124e-07,
+      "loss": 0.7893,
+      "step": 3575
+    },
+    {
+      "epoch": 3.9557522123893807,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.1061946902654869e-07,
+      "loss": 0.7756,
+      "step": 3576
+    },
+    {
+      "epoch": 3.956858407079646,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.0785398230088496e-07,
+      "loss": 0.8081,
+      "step": 3577
+    },
+    {
+      "epoch": 3.9579646017699117,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.0508849557522126e-07,
+      "loss": 0.7685,
+      "step": 3578
+    },
+    {
+      "epoch": 3.959070796460177,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.0232300884955753e-07,
+      "loss": 0.7497,
+      "step": 3579
+    },
+    {
+      "epoch": 3.9601769911504423,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 9.955752212389381e-08,
+      "loss": 0.7398,
+      "step": 3580
+    },
+    {
+      "epoch": 3.961283185840708,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 9.67920353982301e-08,
+      "loss": 0.7704,
+      "step": 3581
+    },
+    {
+      "epoch": 3.9623893805309733,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.402654867256638e-08,
+      "loss": 0.7609,
+      "step": 3582
+    },
+    {
+      "epoch": 3.963495575221239,
+      "grad_norm": 0.11328125,
+      "learning_rate": 9.126106194690267e-08,
+      "loss": 0.7699,
+      "step": 3583
+    },
+    {
+      "epoch": 3.9646017699115044,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 8.849557522123894e-08,
+      "loss": 0.7692,
+      "step": 3584
+    },
+    {
+      "epoch": 3.96570796460177,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 8.573008849557522e-08,
+      "loss": 0.731,
+      "step": 3585
+    },
+    {
+      "epoch": 3.9668141592920354,
+      "grad_norm": 0.109375,
+      "learning_rate": 8.296460176991152e-08,
+      "loss": 0.7784,
+      "step": 3586
+    },
+    {
+      "epoch": 3.9679203539823007,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 8.019911504424779e-08,
+      "loss": 0.741,
+      "step": 3587
+    },
+    {
+      "epoch": 3.9690265486725664,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 7.743362831858408e-08,
+      "loss": 0.7622,
+      "step": 3588
+    },
+    {
+      "epoch": 3.9701327433628317,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 7.466814159292036e-08,
+      "loss": 0.7787,
+      "step": 3589
+    },
+    {
+      "epoch": 3.9712389380530975,
+      "grad_norm": 0.119140625,
+      "learning_rate": 7.190265486725665e-08,
+      "loss": 0.7842,
+      "step": 3590
+    },
+    {
+      "epoch": 3.9723451327433628,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.913716814159293e-08,
+      "loss": 0.7662,
+      "step": 3591
+    },
+    {
+      "epoch": 3.9734513274336285,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.63716814159292e-08,
+      "loss": 0.7615,
+      "step": 3592
+    },
+    {
+      "epoch": 3.974557522123894,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.36061946902655e-08,
+      "loss": 0.7459,
+      "step": 3593
+    },
+    {
+      "epoch": 3.975663716814159,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 6.084070796460177e-08,
+      "loss": 0.7654,
+      "step": 3594
+    },
+    {
+      "epoch": 3.976769911504425,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 5.807522123893806e-08,
+      "loss": 0.7868,
+      "step": 3595
+    },
+    {
+      "epoch": 3.97787610619469,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 5.530973451327434e-08,
+      "loss": 0.7866,
+      "step": 3596
+    },
+    {
+      "epoch": 3.978982300884956,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.254424778761063e-08,
+      "loss": 0.8269,
+      "step": 3597
+    },
+    {
+      "epoch": 3.980088495575221,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 4.9778761061946906e-08,
+      "loss": 0.7598,
+      "step": 3598
+    },
+    {
+      "epoch": 3.981194690265487,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 4.701327433628319e-08,
+      "loss": 0.7803,
+      "step": 3599
+    },
+    {
+      "epoch": 3.982300884955752,
+      "grad_norm": 0.115234375,
+      "learning_rate": 4.424778761061947e-08,
+      "loss": 0.783,
+      "step": 3600
+    },
+    {
+      "epoch": 3.9834070796460175,
+      "grad_norm": 0.111328125,
+      "learning_rate": 4.148230088495576e-08,
+      "loss": 0.7567,
+      "step": 3601
+    },
+    {
+      "epoch": 3.984513274336283,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.871681415929204e-08,
+      "loss": 0.786,
+      "step": 3602
+    },
+    {
+      "epoch": 3.9856194690265485,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 3.5951327433628324e-08,
+      "loss": 0.7732,
+      "step": 3603
+    },
+    {
+      "epoch": 3.9867256637168142,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.31858407079646e-08,
+      "loss": 0.7544,
+      "step": 3604
+    },
+    {
+      "epoch": 3.9878318584070795,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 3.0420353982300887e-08,
+      "loss": 0.7933,
+      "step": 3605
+    },
+    {
+      "epoch": 3.9889380530973453,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.765486725663717e-08,
+      "loss": 0.7583,
+      "step": 3606
+    },
+    {
+      "epoch": 3.9900442477876106,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.4889380530973453e-08,
+      "loss": 0.8267,
+      "step": 3607
+    },
+    {
+      "epoch": 3.991150442477876,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.2123893805309735e-08,
+      "loss": 0.7518,
+      "step": 3608
+    },
+    {
+      "epoch": 3.9922566371681416,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.935840707964602e-08,
+      "loss": 0.7463,
+      "step": 3609
+    },
+    {
+      "epoch": 3.9933628318584073,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.65929203539823e-08,
+      "loss": 0.7644,
+      "step": 3610
+    },
+    {
+      "epoch": 3.9944690265486726,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 1.3827433628318586e-08,
+      "loss": 0.7414,
+      "step": 3611
+    },
+    {
+      "epoch": 3.995575221238938,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.1061946902654867e-08,
+      "loss": 0.762,
+      "step": 3612
+    },
+    {
+      "epoch": 3.9966814159292037,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 8.29646017699115e-09,
+      "loss": 0.7632,
+      "step": 3613
+    },
+    {
+      "epoch": 3.997787610619469,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.530973451327434e-09,
+      "loss": 0.7431,
+      "step": 3614
+    },
+    {
+      "epoch": 3.9988938053097343,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.765486725663717e-09,
+      "loss": 0.7647,
+      "step": 3615
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 0.0,
+      "loss": 0.7439,
+      "step": 3616
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3616,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1647918971822277e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}