diff --git "a/checkpoint-992160/trainer_state.json" "b/checkpoint-992160/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-992160/trainer_state.json"
@@ -0,0 +1,11751 @@
+{
+  "best_metric": 4.038602352142334,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl2/lstm/0/checkpoints/checkpoint-992160",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 992160,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8196,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5546,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0588,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9975,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.948,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9267,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7497,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6233,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5276,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.449,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.4031,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.344,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2792,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2144,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1629,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.1059,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0645,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.0259,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.98,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9363,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.9066,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8626,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8368,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.799,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.7804,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7417,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7184,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6894,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6715,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.975684027980238e-05,
+      "loss": 5.6447,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6314,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9740101142388804e-05,
+      "loss": 5.6139,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.5886,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5732,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5575,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5351,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5314,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4978,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4833,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4548,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4532,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4371,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4339,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4069,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.3969,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962271425604525e-05,
+      "loss": 5.3948,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3864,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3737,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3556,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3407,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.357,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3441,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3228,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.3016,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.3115,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953888753854752e-05,
+      "loss": 5.2911,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530501591037e-05,
+      "loss": 5.2943,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952211564352648e-05,
+      "loss": 5.265,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2777,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2603,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2428,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2297,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2418,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.199,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.2142,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.2204,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.204,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.195,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1705,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1663,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1631,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.174,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1529,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1619,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1523,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.1499,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.936284815604152e-05,
+      "loss": 5.1269,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354462208531e-05,
+      "loss": 5.1248,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934607626102048e-05,
+      "loss": 5.1154,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.1125,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1085,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0933,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0954,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0925,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.068,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0763,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0739,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0752,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.055,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0527,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0386,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0347,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0358,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 5.0332,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0241,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.027,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.0176,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9996,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9973,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.9989,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9954,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9989,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9804,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9724,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9768,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9575,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9627,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9508,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9528,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.94,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907784059230996e-05,
+      "loss": 4.9416,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9314,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9353,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9299,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9197,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9306,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9266,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9163,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.9031,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.9036,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.9011,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.9034,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8975,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8861,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8913,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8778,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8808,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893532862104231e-05,
+      "loss": 4.8659,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8734,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.862,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910219914921943e-05,
+      "loss": 4.8718,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8901833967411423e-05,
+      "loss": 4.8703,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.8673,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8442,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8504,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8477,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8494,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8377,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8434,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883476276613099e-05,
+      "loss": 4.832,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8467,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8147,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8336,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8186,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.814,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8138,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.798,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8026,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.8112,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7969,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.763216018676758,
+      "eval_runtime": 291.1721,
+      "eval_samples_per_second": 1310.534,
+      "eval_steps_per_second": 40.955,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7958,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7904,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.8009,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7898,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7963,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7727,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7797,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7651,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7687,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667092952331785e-05,
+      "loss": 4.7717,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7684,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7728,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7506,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7642,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7473,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.861682640367986e-05,
+      "loss": 4.7498,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860845683497307e-05,
+      "loss": 4.753,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860007088746255e-05,
+      "loss": 4.743,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859168493995203e-05,
+      "loss": 4.7451,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.7585,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.735,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.7409,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7343,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7436,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7331,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7313,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7142,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.7229,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.7097,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849947227494377e-05,
+      "loss": 4.7204,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.849108632743325e-05,
+      "loss": 4.7163,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.848270037992273e-05,
+      "loss": 4.714,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847431443241221e-05,
+      "loss": 4.7175,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.7071,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.714,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7061,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6957,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6884,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.6798,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84156291786423e-05,
+      "loss": 4.6873,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6871,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.6923,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.6764,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6781,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8373732198697167e-05,
+      "loss": 4.6816,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8365346251186646e-05,
+      "loss": 4.6744,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8356960303676126e-05,
+      "loss": 4.68,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8348574356165606e-05,
+      "loss": 4.665,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8340188408655086e-05,
+      "loss": 4.6578,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8331802461144566e-05,
+      "loss": 4.6848,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323432892437775e-05,
+      "loss": 4.6778,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315046944927255e-05,
+      "loss": 4.6643,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306660997416735e-05,
+      "loss": 4.6594,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298275049906215e-05,
+      "loss": 4.6667,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.6511,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281519533688904e-05,
+      "loss": 4.6678,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.827314996498212e-05,
+      "loss": 4.6387,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.82647640174716e-05,
+      "loss": 4.6644,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.825637806996108e-05,
+      "loss": 4.6559,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.824799212245056e-05,
+      "loss": 4.6392,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823960617494004e-05,
+      "loss": 4.6358,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823122022742952e-05,
+      "loss": 4.6519,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222834279919e-05,
+      "loss": 4.6143,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821444833240848e-05,
+      "loss": 4.6365,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820606238489796e-05,
+      "loss": 4.6459,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819769281619117e-05,
+      "loss": 4.639,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818930686868065e-05,
+      "loss": 4.6287,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818092092117013e-05,
+      "loss": 4.616,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817253497365961e-05,
+      "loss": 4.6203,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6238,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.6292,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8147393509931785e-05,
+      "loss": 4.6198,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.6269,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.6313,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122235667400224e-05,
+      "loss": 4.6275,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.6112,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810549652998664e-05,
+      "loss": 4.6157,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809711058247612e-05,
+      "loss": 4.6156,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80887246349656e-05,
+      "loss": 4.6082,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808033868745508e-05,
+      "loss": 4.6117,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.807195273994456e-05,
+      "loss": 4.6031,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806356679243404e-05,
+      "loss": 4.6102,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805518084492352e-05,
+      "loss": 4.6101,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.5873,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.602,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.5998,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.6075,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.5976,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800488153866413e-05,
+      "loss": 4.5925,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.5881,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5815,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5876,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.5908,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7962984558718996e-05,
+      "loss": 4.5876,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7954598611208476e-05,
+      "loss": 4.5924,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7946212663697956e-05,
+      "loss": 4.5844,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7937826716187436e-05,
+      "loss": 4.5766,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7929440768676916e-05,
+      "loss": 4.5735,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7921054821166396e-05,
+      "loss": 4.5786,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7912668873655876e-05,
+      "loss": 4.583,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7904282926145356e-05,
+      "loss": 4.5831,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895913357438565e-05,
+      "loss": 4.575,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887527409928045e-05,
+      "loss": 4.5657,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879157841221254e-05,
+      "loss": 4.5766,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870771893710734e-05,
+      "loss": 4.5581,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862385946200214e-05,
+      "loss": 4.5674,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853999998689694e-05,
+      "loss": 4.5611,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.784561405117918e-05,
+      "loss": 4.5627,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.783722810366866e-05,
+      "loss": 4.551,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782884215615814e-05,
+      "loss": 4.5613,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782045620864762e-05,
+      "loss": 4.55,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.5542,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5574,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779531474491979e-05,
+      "loss": 4.547,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.5678,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777854284989875e-05,
+      "loss": 4.5631,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777015690238823e-05,
+      "loss": 4.5561,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.5446,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5418,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5493,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.5441,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.5552,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.5352,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.5428,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.5415,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.5461,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768631380608676e-05,
+      "loss": 4.5264,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.5391,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.5351,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.5364,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5417,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5448,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.5223,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762764493112058e-05,
+      "loss": 4.531,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.5337,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5293,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760250346739275e-05,
+      "loss": 4.5264,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.759411751988223e-05,
+      "loss": 4.5305,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758573157237171e-05,
+      "loss": 4.5235,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757734562486119e-05,
+      "loss": 4.5396,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756895967735067e-05,
+      "loss": 4.5139,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5303,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.517,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5158,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.5161,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.5093,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518676749895016e-05,
+      "loss": 4.5092,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7510290802384496e-05,
+      "loss": 4.5149,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7501904854873976e-05,
+      "loss": 4.5095,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.47607421875,
+      "eval_runtime": 291.0564,
+      "eval_samples_per_second": 1311.055,
+      "eval_steps_per_second": 40.971,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7493518907363456e-05,
+      "loss": 4.5083,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7485132959852936e-05,
+      "loss": 4.5002,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7476747012342416e-05,
+      "loss": 4.5195,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7468361064831896e-05,
+      "loss": 4.5069,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7459975117321376e-05,
+      "loss": 4.5188,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745158916981085e-05,
+      "loss": 4.4955,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744320322230033e-05,
+      "loss": 4.5033,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743481727478981e-05,
+      "loss": 4.4903,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742643132727929e-05,
+      "loss": 4.5007,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741804537976877e-05,
+      "loss": 4.4998,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740965943225825e-05,
+      "loss": 4.496,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401273484747736e-05,
+      "loss": 4.5094,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392903916040945e-05,
+      "loss": 4.4829,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384517968530425e-05,
+      "loss": 4.5007,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376132021019905e-05,
+      "loss": 4.4848,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367746073509385e-05,
+      "loss": 4.4844,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359360125998865e-05,
+      "loss": 4.4859,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350974178488345e-05,
+      "loss": 4.4927,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342588230977825e-05,
+      "loss": 4.4833,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334218662271034e-05,
+      "loss": 4.5025,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325832714760514e-05,
+      "loss": 4.485,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731746314605372e-05,
+      "loss": 4.4853,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73090771985432e-05,
+      "loss": 4.4828,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730069125103269e-05,
+      "loss": 4.496,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729230530352217e-05,
+      "loss": 4.4816,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728391935601165e-05,
+      "loss": 4.4846,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727553340850113e-05,
+      "loss": 4.4714,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726714746099061e-05,
+      "loss": 4.4759,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725876151348009e-05,
+      "loss": 4.4652,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725037556596957e-05,
+      "loss": 4.4799,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724198961845904e-05,
+      "loss": 4.4806,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723360367094852e-05,
+      "loss": 4.4751,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225217723438e-05,
+      "loss": 4.4836,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721684815473121e-05,
+      "loss": 4.4715,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4727,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.4748,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4661,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718332074349287e-05,
+      "loss": 4.457,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717493479598235e-05,
+      "loss": 4.4499,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716654884847183e-05,
+      "loss": 4.4585,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715816290096131e-05,
+      "loss": 4.4618,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714977695345079e-05,
+      "loss": 4.4659,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141407384743996e-05,
+      "loss": 4.4562,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133021437233476e-05,
+      "loss": 4.451,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124635489722956e-05,
+      "loss": 4.4636,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116265921016165e-05,
+      "loss": 4.4549,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107879973505645e-05,
+      "loss": 4.4556,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099494025995125e-05,
+      "loss": 4.4496,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091108078484605e-05,
+      "loss": 4.4382,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708272213097409e-05,
+      "loss": 4.4686,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707433618346357e-05,
+      "loss": 4.4619,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4478,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.4475,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704919471973574e-05,
+      "loss": 4.4566,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704080877222522e-05,
+      "loss": 4.4364,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70324228247147e-05,
+      "loss": 4.4571,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702403687720418e-05,
+      "loss": 4.4307,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701565092969366e-05,
+      "loss": 4.4516,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700726498218314e-05,
+      "loss": 4.4535,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699887903467262e-05,
+      "loss": 4.4362,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699052584476956e-05,
+      "loss": 4.4289,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982139897259045e-05,
+      "loss": 4.4472,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973753949748525e-05,
+      "loss": 4.4166,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965368002238005e-05,
+      "loss": 4.4354,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956982054727485e-05,
+      "loss": 4.4405,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948596107216965e-05,
+      "loss": 4.441,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940210159706445e-05,
+      "loss": 4.4337,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931824212195925e-05,
+      "loss": 4.4187,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69234382646854e-05,
+      "loss": 4.4227,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691505231717488e-05,
+      "loss": 4.424,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906682748468094e-05,
+      "loss": 4.4353,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68983131797613e-05,
+      "loss": 4.426,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688992723225078e-05,
+      "loss": 4.4357,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688154128474026e-05,
+      "loss": 4.4393,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687315533722974e-05,
+      "loss": 4.4362,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686476938971923e-05,
+      "loss": 4.4179,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68563834422087e-05,
+      "loss": 4.4248,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684799749469818e-05,
+      "loss": 4.431,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683961154718766e-05,
+      "loss": 4.4182,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683124197848087e-05,
+      "loss": 4.4238,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682285603097035e-05,
+      "loss": 4.4194,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681447008345983e-05,
+      "loss": 4.4189,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680608413594931e-05,
+      "loss": 4.4261,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679769818843879e-05,
+      "loss": 4.4045,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6789328619732e-05,
+      "loss": 4.4217,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678094267222148e-05,
+      "loss": 4.4136,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677255672471097e-05,
+      "loss": 4.4259,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676417077720045e-05,
+      "loss": 4.4151,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675578482968993e-05,
+      "loss": 4.4142,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747415260983136e-05,
+      "loss": 4.4084,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6739029313472616e-05,
+      "loss": 4.4044,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6730643365962096e-05,
+      "loss": 4.405,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6722257418451576e-05,
+      "loss": 4.4123,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6713887849744785e-05,
+      "loss": 4.4207,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705501902234265e-05,
+      "loss": 4.4151,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697115954723745e-05,
+      "loss": 4.4073,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688730007213225e-05,
+      "loss": 4.4021,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680360438506434e-05,
+      "loss": 4.4025,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667197449099592e-05,
+      "loss": 4.4042,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66635885434854e-05,
+      "loss": 4.4135,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665520259597488e-05,
+      "loss": 4.4086,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664683302726809e-05,
+      "loss": 4.4038,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66384634585613e-05,
+      "loss": 4.3942,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6630093889854515e-05,
+      "loss": 4.4103,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621707942343995e-05,
+      "loss": 4.3909,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661332199483347e-05,
+      "loss": 4.4001,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660493604732295e-05,
+      "loss": 4.392,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659655009981243e-05,
+      "loss": 4.3979,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658816415230191e-05,
+      "loss": 4.3878,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657977820479139e-05,
+      "loss": 4.399,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657139225728087e-05,
+      "loss": 4.3888,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6563006309770355e-05,
+      "loss": 4.3855,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554620362259835e-05,
+      "loss": 4.3996,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546234414749315e-05,
+      "loss": 4.3827,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537848467238795e-05,
+      "loss": 4.4053,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529478898532004e-05,
+      "loss": 4.4016,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6521092951021484e-05,
+      "loss": 4.3957,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512707003510964e-05,
+      "loss": 4.3858,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650433743480417e-05,
+      "loss": 4.3856,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649595148729365e-05,
+      "loss": 4.3867,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648756553978313e-05,
+      "loss": 4.3863,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647917959227261e-05,
+      "loss": 4.396,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647079364476209e-05,
+      "loss": 4.3784,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646240769725157e-05,
+      "loss": 4.3884,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645402174974105e-05,
+      "loss": 4.3878,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644563580223053e-05,
+      "loss": 4.3889,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643724985472001e-05,
+      "loss": 4.376,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642886390720949e-05,
+      "loss": 4.3838,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642047795969897e-05,
+      "loss": 4.3818,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641209201218845e-05,
+      "loss": 4.3848,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640370606467793e-05,
+      "loss": 4.3871,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639533649597114e-05,
+      "loss": 4.3935,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638695054846062e-05,
+      "loss": 4.3692,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63785646009501e-05,
+      "loss": 4.3825,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637017865343958e-05,
+      "loss": 4.3858,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636179270592906e-05,
+      "loss": 4.3795,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635340675841854e-05,
+      "loss": 4.3769,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.3858,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633665124220124e-05,
+      "loss": 4.3753,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632826529469072e-05,
+      "loss": 4.3918,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63198793471802e-05,
+      "loss": 4.3664,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6311509778473406e-05,
+      "loss": 4.3813,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6303123830962886e-05,
+      "loss": 4.3743,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294737883452366e-05,
+      "loss": 4.3722,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286351935941846e-05,
+      "loss": 4.3696,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277965988431326e-05,
+      "loss": 4.3664,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269580040920806e-05,
+      "loss": 4.3625,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261194093410286e-05,
+      "loss": 4.3717,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252824524703495e-05,
+      "loss": 4.3676,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.3395867347717285,
+      "eval_runtime": 287.9323,
+      "eval_samples_per_second": 1325.28,
+      "eval_steps_per_second": 41.416,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6244438577192975e-05,
+      "loss": 4.3731,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623605262968246e-05,
+      "loss": 4.3578,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622766668217194e-05,
+      "loss": 4.3766,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621928073466142e-05,
+      "loss": 4.3644,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210894787150895e-05,
+      "loss": 4.3811,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202508839640374e-05,
+      "loss": 4.3579,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3648,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3501,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177367375912543e-05,
+      "loss": 4.3651,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168981428402023e-05,
+      "loss": 4.3624,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61605954808915e-05,
+      "loss": 4.3591,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615220953338098e-05,
+      "loss": 4.3656,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61438399646742e-05,
+      "loss": 4.3518,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613545401716368e-05,
+      "loss": 4.3652,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612706806965316e-05,
+      "loss": 4.3516,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611868212214264e-05,
+      "loss": 4.3458,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611029617463212e-05,
+      "loss": 4.3522,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61019102271216e-05,
+      "loss": 4.3555,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609352427961108e-05,
+      "loss": 4.3538,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608513833210056e-05,
+      "loss": 4.3687,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607676876339377e-05,
+      "loss": 4.3532,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606838281588325e-05,
+      "loss": 4.3591,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605999686837273e-05,
+      "loss": 4.3497,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605162729966594e-05,
+      "loss": 4.3618,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604324135215542e-05,
+      "loss": 4.3503,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60348554046449e-05,
+      "loss": 4.3547,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602646945713438e-05,
+      "loss": 4.3439,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601809988842759e-05,
+      "loss": 4.3407,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600971394091707e-05,
+      "loss": 4.3418,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600132799340655e-05,
+      "loss": 4.3483,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599294204589603e-05,
+      "loss": 4.3513,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598455609838551e-05,
+      "loss": 4.3472,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597617015087499e-05,
+      "loss": 4.3571,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596778420336447e-05,
+      "loss": 4.3467,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595939825585395e-05,
+      "loss": 4.3423,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595102868714716e-05,
+      "loss": 4.3465,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594264273963664e-05,
+      "loss": 4.3467,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3292,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3242,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3322,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590909894959456e-05,
+      "loss": 4.3421,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590072938088778e-05,
+      "loss": 4.3415,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5892359812180986e-05,
+      "loss": 4.3364,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883973864670466e-05,
+      "loss": 4.3265,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5875587917159946e-05,
+      "loss": 4.3418,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867201969649426e-05,
+      "loss": 4.334,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858816022138906e-05,
+      "loss": 4.3338,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585043007462838e-05,
+      "loss": 4.3268,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.3124,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583365817960734e-05,
+      "loss": 4.3546,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825288610900555e-05,
+      "loss": 4.3382,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816902663390035e-05,
+      "loss": 4.331,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.3307,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3332,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3215,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3356,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774989304641164e-05,
+      "loss": 4.3163,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766603357130644e-05,
+      "loss": 4.3291,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758217409620124e-05,
+      "loss": 4.3402,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749831462109604e-05,
+      "loss": 4.3201,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3098,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573307594589229e-05,
+      "loss": 4.3363,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572468999838177e-05,
+      "loss": 4.3019,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571632042967499e-05,
+      "loss": 4.3153,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570793448216447e-05,
+      "loss": 4.3281,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569954853465395e-05,
+      "loss": 4.327,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569116258714343e-05,
+      "loss": 4.3188,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568277663963291e-05,
+      "loss": 4.3115,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567439069212239e-05,
+      "loss": 4.3031,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566600474461187e-05,
+      "loss": 4.3094,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565761879710135e-05,
+      "loss": 4.3256,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564923284959083e-05,
+      "loss": 4.3101,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3255,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.3218,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3296,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561570543835248e-05,
+      "loss": 4.3104,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560731949084196e-05,
+      "loss": 4.3113,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559893354333144e-05,
+      "loss": 4.3214,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559054759582092e-05,
+      "loss": 4.3053,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558219440591786e-05,
+      "loss": 4.3199,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557380845840734e-05,
+      "loss": 4.3082,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556542251089682e-05,
+      "loss": 4.3106,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55570365633863e-05,
+      "loss": 4.3177,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554866699467951e-05,
+      "loss": 4.2995,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554028104716899e-05,
+      "loss": 4.3114,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553189509965847e-05,
+      "loss": 4.3096,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552350915214795e-05,
+      "loss": 4.3163,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551513958344116e-05,
+      "loss": 4.3042,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550675363593064e-05,
+      "loss": 4.3085,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.549836768842013e-05,
+      "loss": 4.3005,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548998174090961e-05,
+      "loss": 4.2977,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548159579339909e-05,
+      "loss": 4.3007,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547320984588857e-05,
+      "loss": 4.3054,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546482389837804e-05,
+      "loss": 4.3161,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545643795086752e-05,
+      "loss": 4.3137,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448052003357e-05,
+      "loss": 4.3038,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439682434650216e-05,
+      "loss": 4.2948,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543129648713969e-05,
+      "loss": 4.2976,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5422926918432905e-05,
+      "loss": 4.3031,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414540970922385e-05,
+      "loss": 4.3075,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406155023411865e-05,
+      "loss": 4.3073,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397769075901345e-05,
+      "loss": 4.3011,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538939950719456e-05,
+      "loss": 4.2909,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538101355968404e-05,
+      "loss": 4.3112,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372627612173514e-05,
+      "loss": 4.284,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364241664662994e-05,
+      "loss": 4.3026,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355855717152473e-05,
+      "loss": 4.2894,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347469769641953e-05,
+      "loss": 4.2971,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533908382213143e-05,
+      "loss": 4.2884,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533069787462091e-05,
+      "loss": 4.2963,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532232830591412e-05,
+      "loss": 4.2909,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53139423584036e-05,
+      "loss": 4.2848,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530555641089308e-05,
+      "loss": 4.2982,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52971868421863e-05,
+      "loss": 4.2853,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528880089467578e-05,
+      "loss": 4.3044,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528041494716526e-05,
+      "loss": 4.3057,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527202899965474e-05,
+      "loss": 4.3009,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526364305214422e-05,
+      "loss": 4.2879,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52552571046337e-05,
+      "loss": 4.2877,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524687115712318e-05,
+      "loss": 4.2863,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523848520961266e-05,
+      "loss": 4.289,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523011564090587e-05,
+      "loss": 4.3001,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522172969339535e-05,
+      "loss": 4.276,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521334374588483e-05,
+      "loss": 4.2916,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520495779837431e-05,
+      "loss": 4.2884,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519657185086379e-05,
+      "loss": 4.2955,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.518818590335327e-05,
+      "loss": 4.2812,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517981633464648e-05,
+      "loss": 4.2903,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517143038713596e-05,
+      "loss": 4.2814,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516304443962544e-05,
+      "loss": 4.2952,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515465849211492e-05,
+      "loss": 4.2893,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51462725446044e-05,
+      "loss": 4.2983,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513790297589761e-05,
+      "loss": 4.2725,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512951702838709e-05,
+      "loss": 4.2911,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512113108087657e-05,
+      "loss": 4.2917,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5112745133366045e-05,
+      "loss": 4.28,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5104359185855525e-05,
+      "loss": 4.2839,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095973238345005e-05,
+      "loss": 4.294,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508760366963822e-05,
+      "loss": 4.2843,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50792177221277e-05,
+      "loss": 4.3003,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507083177461718e-05,
+      "loss": 4.2721,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506244582710666e-05,
+      "loss": 4.2902,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5054076258399876e-05,
+      "loss": 4.2772,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504569031088935e-05,
+      "loss": 4.2813,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503730436337883e-05,
+      "loss": 4.2804,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502891841586831e-05,
+      "loss": 4.2735,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502053246835779e-05,
+      "loss": 4.2709,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012162899651e-05,
+      "loss": 4.2783,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500377695214048e-05,
+      "loss": 4.2783,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.257771968841553,
+      "eval_runtime": 288.3054,
+      "eval_samples_per_second": 1323.565,
+      "eval_steps_per_second": 41.362,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499539100462996e-05,
+      "loss": 4.2781,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498700505711944e-05,
+      "loss": 4.2686,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978619109608925e-05,
+      "loss": 4.2879,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970233162098405e-05,
+      "loss": 4.275,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961847214587885e-05,
+      "loss": 4.287,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953461267077365e-05,
+      "loss": 4.2754,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945075319566845e-05,
+      "loss": 4.2746,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936689372056325e-05,
+      "loss": 4.2628,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928303424545805e-05,
+      "loss": 4.2729,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919917477035285e-05,
+      "loss": 4.2745,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911531529524765e-05,
+      "loss": 4.2729,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490314558201424e-05,
+      "loss": 4.2748,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894776013307454e-05,
+      "loss": 4.2643,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886390065796934e-05,
+      "loss": 4.2786,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487800411828641e-05,
+      "loss": 4.2631,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486961817077589e-05,
+      "loss": 4.2602,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861232223265373e-05,
+      "loss": 4.2649,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852846275754853e-05,
+      "loss": 4.2712,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484446032824433e-05,
+      "loss": 4.2661,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483607438073381e-05,
+      "loss": 4.2811,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482768843322329e-05,
+      "loss": 4.2697,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481930248571277e-05,
+      "loss": 4.2773,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481091653820225e-05,
+      "loss": 4.2629,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480253059069173e-05,
+      "loss": 4.2752,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479417740078867e-05,
+      "loss": 4.2646,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478579145327815e-05,
+      "loss": 4.2722,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477740550576763e-05,
+      "loss": 4.2587,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476901955825711e-05,
+      "loss": 4.2589,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476063361074659e-05,
+      "loss": 4.2558,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475224766323607e-05,
+      "loss": 4.264,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474386171572556e-05,
+      "loss": 4.2681,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473549214701877e-05,
+      "loss": 4.2659,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472710619950825e-05,
+      "loss": 4.2708,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471872025199773e-05,
+      "loss": 4.2598,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710350683290936e-05,
+      "loss": 4.2617,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701964735780416e-05,
+      "loss": 4.2692,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693578788269896e-05,
+      "loss": 4.266,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685192840759376e-05,
+      "loss": 4.2444,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676806893248856e-05,
+      "loss": 4.2508,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668420945738336e-05,
+      "loss": 4.2482,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660034998227816e-05,
+      "loss": 4.2539,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651649050717296e-05,
+      "loss": 4.2655,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643263103206776e-05,
+      "loss": 4.253,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634877155696256e-05,
+      "loss": 4.2482,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626491208185736e-05,
+      "loss": 4.2588,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618105260675215e-05,
+      "loss": 4.2527,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609735691968425e-05,
+      "loss": 4.2565,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460136612326164e-05,
+      "loss": 4.2495,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459298017575112e-05,
+      "loss": 4.2274,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584594228240594e-05,
+      "loss": 4.2755,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576208280730074e-05,
+      "loss": 4.259,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567822333219554e-05,
+      "loss": 4.2565,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4559436385709033e-05,
+      "loss": 4.2504,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4551050438198513e-05,
+      "loss": 4.2506,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454266449068799e-05,
+      "loss": 4.2442,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4534311300784945e-05,
+      "loss": 4.2532,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4525925353274425e-05,
+      "loss": 4.2419,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45175394057639e-05,
+      "loss": 4.2484,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450915345825338e-05,
+      "loss": 4.2633,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450076751074286e-05,
+      "loss": 4.2411,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449238156323234e-05,
+      "loss": 4.2339,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448399561572182e-05,
+      "loss": 4.2588,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.447562604701503e-05,
+      "loss": 4.2246,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446724009950451e-05,
+      "loss": 4.2405,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445885415199399e-05,
+      "loss": 4.2476,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445046820448347e-05,
+      "loss": 4.249,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444208225697295e-05,
+      "loss": 4.2411,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443369630946243e-05,
+      "loss": 4.2376,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425310361951914e-05,
+      "loss": 4.2222,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416924414441394e-05,
+      "loss": 4.2347,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44085548457346e-05,
+      "loss": 4.2458,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440018527702781e-05,
+      "loss": 4.2365,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439179932951729e-05,
+      "loss": 4.2441,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438341338200677e-05,
+      "loss": 4.2489,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437502743449625e-05,
+      "loss": 4.2569,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2357,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2318,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2482,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.2277,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.2483,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4324744507040597e-05,
+      "loss": 4.2352,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4316358559530076e-05,
+      "loss": 4.2351,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4307972612019556e-05,
+      "loss": 4.2453,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299586664509036e-05,
+      "loss": 4.2253,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291200716998516e-05,
+      "loss": 4.2354,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.2324,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274445200781205e-05,
+      "loss": 4.2432,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266059253270685e-05,
+      "loss": 4.2286,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257673305760165e-05,
+      "loss": 4.2381,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249287358249645e-05,
+      "loss": 4.2259,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240901410739125e-05,
+      "loss": 4.2282,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232531842032334e-05,
+      "loss": 4.2237,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.422414589452182e-05,
+      "loss": 4.235,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42157599470113e-05,
+      "loss": 4.2429,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420737399950078e-05,
+      "loss": 4.2377,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419898805199026e-05,
+      "loss": 4.2376,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419061848328347e-05,
+      "loss": 4.2235,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418223253577295e-05,
+      "loss": 4.2263,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417384658826243e-05,
+      "loss": 4.2294,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.2376,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415707469324138e-05,
+      "loss": 4.2338,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414868874573086e-05,
+      "loss": 4.2281,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414031917702408e-05,
+      "loss": 4.2223,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413193322951356e-05,
+      "loss": 4.2396,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412354728200304e-05,
+      "loss": 4.2134,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411516133449252e-05,
+      "loss": 4.2302,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106775386982e-05,
+      "loss": 4.2196,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409840581827521e-05,
+      "loss": 4.2271,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409001987076469e-05,
+      "loss": 4.2181,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.2267,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407324797574365e-05,
+      "loss": 4.2185,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406486202823313e-05,
+      "loss": 4.2185,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405647608072261e-05,
+      "loss": 4.2264,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404809013321209e-05,
+      "loss": 4.2183,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40397205645053e-05,
+      "loss": 4.2344,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031334616994777e-05,
+      "loss": 4.2361,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022948669484257e-05,
+      "loss": 4.2284,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401456272197374e-05,
+      "loss": 4.2246,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400617677446322e-05,
+      "loss": 4.2153,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39977908269527e-05,
+      "loss": 4.2207,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398940487944218e-05,
+      "loss": 4.22,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.236,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397264936322487e-05,
+      "loss": 4.2052,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396426341571435e-05,
+      "loss": 4.223,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395587746820383e-05,
+      "loss": 4.221,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394750789949704e-05,
+      "loss": 4.2262,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393912195198652e-05,
+      "loss": 4.2172,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930736004476e-05,
+      "loss": 4.2202,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392236643576921e-05,
+      "loss": 4.2141,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39139804882587e-05,
+      "loss": 4.227,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390559454074818e-05,
+      "loss": 4.2244,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389720859323766e-05,
+      "loss": 4.2265,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388882264572714e-05,
+      "loss": 4.209,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3880453077020346e-05,
+      "loss": 4.2242,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2257,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2149,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2165,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846925665781995e-05,
+      "loss": 4.2263,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838539718271475e-05,
+      "loss": 4.2158,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830153770760955e-05,
+      "loss": 4.2387,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.2088,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813381875739915e-05,
+      "loss": 4.2216,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.2091,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796609980718875e-05,
+      "loss": 4.2166,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788224033208355e-05,
+      "loss": 4.2178,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.2088,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.2093,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763082569480524e-05,
+      "loss": 4.2046,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2173,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.20110559463501,
+      "eval_runtime": 297.2997,
+      "eval_samples_per_second": 1283.523,
+      "eval_steps_per_second": 40.111,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.2185,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.2042,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2204,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372116921073165e-05,
+      "loss": 4.2164,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371279964202487e-05,
+      "loss": 4.2198,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370441369451435e-05,
+      "loss": 4.2088,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369602774700383e-05,
+      "loss": 4.2112,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368764179949331e-05,
+      "loss": 4.1984,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367925585198279e-05,
+      "loss": 4.2089,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367086990447227e-05,
+      "loss": 4.2107,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366248395696175e-05,
+      "loss": 4.2139,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365409800945123e-05,
+      "loss": 4.2071,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364572844074444e-05,
+      "loss": 4.2075,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363734249323392e-05,
+      "loss": 4.2091,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36289565457234e-05,
+      "loss": 4.2022,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362057059821288e-05,
+      "loss": 4.1942,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612201029506086e-05,
+      "loss": 4.2019,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3603815081995566e-05,
+      "loss": 4.2083,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359542913448505e-05,
+      "loss": 4.2044,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358704318697453e-05,
+      "loss": 4.2181,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357865723946401e-05,
+      "loss": 4.2095,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357028767075722e-05,
+      "loss": 4.2135,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35619017232467e-05,
+      "loss": 4.2011,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355351577573618e-05,
+      "loss": 4.2113,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354514620702939e-05,
+      "loss": 4.2036,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353676025951887e-05,
+      "loss": 4.2133,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352837431200835e-05,
+      "loss": 4.1984,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351998836449783e-05,
+      "loss": 4.1972,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351160241698731e-05,
+      "loss": 4.1952,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350321646947679e-05,
+      "loss": 4.2047,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349483052196627e-05,
+      "loss": 4.2031,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348644457445575e-05,
+      "loss": 4.2063,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347805862694523e-05,
+      "loss": 4.2079,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346967267943471e-05,
+      "loss": 4.2036,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346128673192419e-05,
+      "loss": 4.2002,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345290078441367e-05,
+      "loss": 4.2079,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344453121570688e-05,
+      "loss": 4.2038,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343614526819636e-05,
+      "loss": 4.1882,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342775932068584e-05,
+      "loss": 4.1897,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341937337317532e-05,
+      "loss": 4.187,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341100380446853e-05,
+      "loss": 4.1891,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.340261785695801e-05,
+      "loss": 4.2056,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.339423190944749e-05,
+      "loss": 4.1959,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385845961936975e-05,
+      "loss": 4.1897,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377460014426455e-05,
+      "loss": 4.2005,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33691068245234e-05,
+      "loss": 4.1912,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336072087701287e-05,
+      "loss": 4.1988,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335233492950235e-05,
+      "loss": 4.1895,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334394898199183e-05,
+      "loss": 4.1677,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333556303448131e-05,
+      "loss": 4.2178,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332717708697079e-05,
+      "loss": 4.1963,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331879113946027e-05,
+      "loss": 4.2019,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331040519194975e-05,
+      "loss": 4.1877,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330203562324296e-05,
+      "loss": 4.198,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329366605453618e-05,
+      "loss": 4.1852,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328528010702566e-05,
+      "loss": 4.1906,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327689415951514e-05,
+      "loss": 4.1903,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326850821200462e-05,
+      "loss": 4.1892,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32601222644941e-05,
+      "loss": 4.208,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325173631698358e-05,
+      "loss": 4.1828,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324335036947306e-05,
+      "loss": 4.1744,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323496442196254e-05,
+      "loss": 4.2015,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322657847445202e-05,
+      "loss": 4.1681,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321820890574523e-05,
+      "loss": 4.1775,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209822958234707e-05,
+      "loss": 4.1937,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201437010724187e-05,
+      "loss": 4.194,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193051063213666e-05,
+      "loss": 4.1812,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184681494506876e-05,
+      "loss": 4.1861,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317629554699636e-05,
+      "loss": 4.165,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31679423570933e-05,
+      "loss": 4.1814,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315955640958278e-05,
+      "loss": 4.1894,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315117046207226e-05,
+      "loss": 4.1783,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314278451456174e-05,
+      "loss": 4.1884,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313439856705122e-05,
+      "loss": 4.1901,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31260126195407e-05,
+      "loss": 4.2021,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311762667203018e-05,
+      "loss": 4.1771,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310924072451966e-05,
+      "loss": 4.1822,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310085477700914e-05,
+      "loss": 4.1924,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309246882949862e-05,
+      "loss": 4.1733,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30840828819881e-05,
+      "loss": 4.1881,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307569693447758e-05,
+      "loss": 4.1829,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306731098696706e-05,
+      "loss": 4.1807,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058941418260276e-05,
+      "loss": 4.1922,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050555470749756e-05,
+      "loss": 4.1715,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042169523239236e-05,
+      "loss": 4.1758,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303378357572871e-05,
+      "loss": 4.1825,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.302539762821819e-05,
+      "loss": 4.1885,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301701168070767e-05,
+      "loss": 4.1742,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008642112000885e-05,
+      "loss": 4.183,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300025616449036e-05,
+      "loss": 4.1662,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299187021697984e-05,
+      "loss": 4.1761,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.298348426946932e-05,
+      "loss": 4.1716,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29750983219588e-05,
+      "loss": 4.179,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966712374448285e-05,
+      "loss": 4.1921,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958326426937765e-05,
+      "loss": 4.1804,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949956858230974e-05,
+      "loss": 4.1811,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941570910720454e-05,
+      "loss": 4.1684,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2933184963209934e-05,
+      "loss": 4.1713,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924799015699413e-05,
+      "loss": 4.1744,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291642944699262e-05,
+      "loss": 4.1852,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29080434994821e-05,
+      "loss": 4.1773,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289967393077531e-05,
+      "loss": 4.1816,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289128798326479e-05,
+      "loss": 4.1698,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288290203575427e-05,
+      "loss": 4.1844,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287451608824375e-05,
+      "loss": 4.1628,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286614651953697e-05,
+      "loss": 4.1776,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285776057202645e-05,
+      "loss": 4.162,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284937462451593e-05,
+      "loss": 4.1793,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284098867700541e-05,
+      "loss": 4.1664,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283260272949489e-05,
+      "loss": 4.173,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282421678198437e-05,
+      "loss": 4.168,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281583083447385e-05,
+      "loss": 4.167,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280744488696333e-05,
+      "loss": 4.1706,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279905893945281e-05,
+      "loss": 4.1683,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279067299194229e-05,
+      "loss": 4.1792,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782303423235496e-05,
+      "loss": 4.1832,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773917475724976e-05,
+      "loss": 4.1774,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2765531528214456e-05,
+      "loss": 4.1761,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2757145580703936e-05,
+      "loss": 4.1606,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2748759633193416e-05,
+      "loss": 4.1689,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274039006448663e-05,
+      "loss": 4.1713,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273200411697611e-05,
+      "loss": 4.182,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272361816946559e-05,
+      "loss": 4.1528,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271523222195507e-05,
+      "loss": 4.172,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2706846274444545e-05,
+      "loss": 4.1732,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269847670573776e-05,
+      "loss": 4.1707,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269010713703097e-05,
+      "loss": 4.165,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268172118952045e-05,
+      "loss": 4.1721,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267333524200993e-05,
+      "loss": 4.1637,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266494929449941e-05,
+      "loss": 4.1763,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265656334698889e-05,
+      "loss": 4.1725,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2648193778282105e-05,
+      "loss": 4.1771,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639807830771585e-05,
+      "loss": 4.1568,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631421883261065e-05,
+      "loss": 4.1735,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623035935750545e-05,
+      "loss": 4.1771,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261464998824002e-05,
+      "loss": 4.1642,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606280419533234e-05,
+      "loss": 4.1681,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597894472022714e-05,
+      "loss": 4.1755,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258950852451219e-05,
+      "loss": 4.169,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258112257700167e-05,
+      "loss": 4.1826,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.1631,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256435068198063e-05,
+      "loss": 4.1698,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255598111327384e-05,
+      "loss": 4.1641,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254759516576332e-05,
+      "loss": 4.168,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25392092182528e-05,
+      "loss": 4.1633,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253082327074228e-05,
+      "loss": 4.1591,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252243732323176e-05,
+      "loss": 4.1638,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251405137572124e-05,
+      "loss": 4.1549,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250568180701445e-05,
+      "loss": 4.1732,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.160346031188965,
+      "eval_runtime": 294.5727,
+      "eval_samples_per_second": 1295.405,
+      "eval_steps_per_second": 40.482,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249729585950393e-05,
+      "loss": 4.1725,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248890991199341e-05,
+      "loss": 4.1574,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248052396448289e-05,
+      "loss": 4.1687,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247213801697237e-05,
+      "loss": 4.1646,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246375206946185e-05,
+      "loss": 4.1741,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245536612195133e-05,
+      "loss": 4.161,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244698017444081e-05,
+      "loss": 4.1627,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243859422693029e-05,
+      "loss": 4.151,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243020827941978e-05,
+      "loss": 4.1606,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242182233190926e-05,
+      "loss": 4.1618,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241343638439873e-05,
+      "loss": 4.1646,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240505043688821e-05,
+      "loss": 4.16,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239668086818143e-05,
+      "loss": 4.1587,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238829492067091e-05,
+      "loss": 4.1604,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237990897316038e-05,
+      "loss": 4.1584,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237152302564986e-05,
+      "loss": 4.1459,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236313707813934e-05,
+      "loss": 4.1572,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235475113062882e-05,
+      "loss": 4.1584,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23463651831183e-05,
+      "loss": 4.157,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233797923560778e-05,
+      "loss": 4.1682,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329609666900996e-05,
+      "loss": 4.1646,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321223719390476e-05,
+      "loss": 4.1685,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312837771879956e-05,
+      "loss": 4.1508,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304451824369436e-05,
+      "loss": 4.1632,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296065876858916e-05,
+      "loss": 4.1572,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2287679929348396e-05,
+      "loss": 4.1623,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279293981837876e-05,
+      "loss": 4.1525,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2270908034327356e-05,
+      "loss": 4.154,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262522086816836e-05,
+      "loss": 4.1515,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254152518110045e-05,
+      "loss": 4.1543,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245766570599525e-05,
+      "loss": 4.1554,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237397001892734e-05,
+      "loss": 4.1625,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2229011054382214e-05,
+      "loss": 4.1609,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22206251068717e-05,
+      "loss": 4.159,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221223915936118e-05,
+      "loss": 4.1531,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220385321185066e-05,
+      "loss": 4.157,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219546726434014e-05,
+      "loss": 4.1603,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218709769563335e-05,
+      "loss": 4.1458,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217871174812283e-05,
+      "loss": 4.1402,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217032580061231e-05,
+      "loss": 4.141,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216193985310179e-05,
+      "loss": 4.1484,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215355390559126e-05,
+      "loss": 4.1541,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214516795808074e-05,
+      "loss": 4.151,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213678201057022e-05,
+      "loss": 4.1454,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21283960630597e-05,
+      "loss": 4.1514,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212002649435292e-05,
+      "loss": 4.1479,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.21116405468424e-05,
+      "loss": 4.152,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.1433,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094885030625094e-05,
+      "loss": 4.1247,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208649908311457e-05,
+      "loss": 4.1681,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207811313560405e-05,
+      "loss": 4.1508,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206972718809353e-05,
+      "loss": 4.1593,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206134124058301e-05,
+      "loss": 4.1469,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205295529307249e-05,
+      "loss": 4.149,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204456934556197e-05,
+      "loss": 4.1416,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036199776855176e-05,
+      "loss": 4.1451,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027813829344656e-05,
+      "loss": 4.1471,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019427881834136e-05,
+      "loss": 4.1416,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2011041934323616e-05,
+      "loss": 4.1642,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20026559868131e-05,
+      "loss": 4.1423,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199428641810631e-05,
+      "loss": 4.1315,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198590047059579e-05,
+      "loss": 4.1574,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197751452308527e-05,
+      "loss": 4.1272,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196912857557475e-05,
+      "loss": 4.1291,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196074262806423e-05,
+      "loss": 4.153,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195235668055371e-05,
+      "loss": 4.152,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194397073304319e-05,
+      "loss": 4.1346,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193558478553267e-05,
+      "loss": 4.1431,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192723159562961e-05,
+      "loss": 4.1215,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191884564811909e-05,
+      "loss": 4.1388,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191045970060857e-05,
+      "loss": 4.1423,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902090131901786e-05,
+      "loss": 4.1388,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893704184391266e-05,
+      "loss": 4.1441,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1457,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876932289370226e-05,
+      "loss": 4.1566,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868546341859706e-05,
+      "loss": 4.1352,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860176773152915e-05,
+      "loss": 4.1392,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1851790825642395e-05,
+      "loss": 4.1478,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1843404878131875e-05,
+      "loss": 4.1337,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1835018930621355e-05,
+      "loss": 4.144,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1826632983110834e-05,
+      "loss": 4.1397,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1818247035600314e-05,
+      "loss": 4.1362,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1809861088089794e-05,
+      "loss": 4.1545,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801475140579274e-05,
+      "loss": 4.1264,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179310557187249e-05,
+      "loss": 4.1306,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178471962436197e-05,
+      "loss": 4.1417,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177635005565518e-05,
+      "loss": 4.144,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176796410814466e-05,
+      "loss": 4.1347,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175957816063414e-05,
+      "loss": 4.1414,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175119221312362e-05,
+      "loss": 4.1224,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17428062656131e-05,
+      "loss": 4.1336,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173442031810257e-05,
+      "loss": 4.1295,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172603437059205e-05,
+      "loss": 4.1366,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171764842308153e-05,
+      "loss": 4.1509,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170926247557101e-05,
+      "loss": 4.1383,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170087652806049e-05,
+      "loss": 4.1397,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169249058054998e-05,
+      "loss": 4.1291,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168410463303946e-05,
+      "loss": 4.1293,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167571868552894e-05,
+      "loss": 4.1293,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166736549562588e-05,
+      "loss": 4.1463,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165897954811536e-05,
+      "loss": 4.1374,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165059360060484e-05,
+      "loss": 4.1411,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164220765309432e-05,
+      "loss": 4.129,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633838084387526e-05,
+      "loss": 4.1417,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625452136877006e-05,
+      "loss": 4.1259,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617066189366486e-05,
+      "loss": 4.1347,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608680241855966e-05,
+      "loss": 4.1195,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1600294294345446e-05,
+      "loss": 4.1372,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159192472563866e-05,
+      "loss": 4.1258,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158353877812814e-05,
+      "loss": 4.1335,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157515283061762e-05,
+      "loss": 4.1238,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15667668831071e-05,
+      "loss": 4.1331,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155838093559658e-05,
+      "loss": 4.1264,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154999498808606e-05,
+      "loss": 4.1261,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154160904057554e-05,
+      "loss": 4.1428,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153323947186875e-05,
+      "loss": 4.1429,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152485352435823e-05,
+      "loss": 4.1387,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151646757684771e-05,
+      "loss": 4.1339,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150808162933719e-05,
+      "loss": 4.1224,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14997120606304e-05,
+      "loss": 4.129,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149132611311988e-05,
+      "loss": 4.133,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482956544413095e-05,
+      "loss": 4.1379,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474570596902575e-05,
+      "loss": 4.1168,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.133,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457798701881535e-05,
+      "loss": 4.1324,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449429133174744e-05,
+      "loss": 4.1316,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1441043185664224e-05,
+      "loss": 4.1273,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432657238153704e-05,
+      "loss": 4.1301,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424271290643184e-05,
+      "loss": 4.1242,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415885343132664e-05,
+      "loss": 4.1372,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407499395622144e-05,
+      "loss": 4.1354,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1399113448111624e-05,
+      "loss": 4.134,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390727500601104e-05,
+      "loss": 4.1169,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138237431069805e-05,
+      "loss": 4.1372,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137398836318753e-05,
+      "loss": 4.1355,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136560241567701e-05,
+      "loss": 4.1237,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135721646816649e-05,
+      "loss": 4.1316,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134883052065597e-05,
+      "loss": 4.1323,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134044457314545e-05,
+      "loss": 4.1301,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133205862563493e-05,
+      "loss": 4.1439,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132367267812441e-05,
+      "loss": 4.1231,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131528673061388e-05,
+      "loss": 4.1323,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1281,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129853121439658e-05,
+      "loss": 4.1285,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129014526688606e-05,
+      "loss": 4.1212,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128175931937554e-05,
+      "loss": 4.1186,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127337337186502e-05,
+      "loss": 4.1242,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12649874243545e-05,
+      "loss": 4.1185,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125663423445144e-05,
+      "loss": 4.1335,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.130283355712891,
+      "eval_runtime": 300.7639,
+      "eval_samples_per_second": 1268.739,
+      "eval_steps_per_second": 39.649,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124824828694092e-05,
+      "loss": 4.13,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12398623394304e-05,
+      "loss": 4.1189,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123147639191988e-05,
+      "loss": 4.1292,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1223090444409356e-05,
+      "loss": 4.1283,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1214704496898835e-05,
+      "loss": 4.1348,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206318549388315e-05,
+      "loss": 4.1239,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1197932601877795e-05,
+      "loss": 4.1302,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1189546654367275e-05,
+      "loss": 4.1149,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118117708566049e-05,
+      "loss": 4.117,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117280751695371e-05,
+      "loss": 4.1305,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116442156944318e-05,
+      "loss": 4.1253,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115603562193266e-05,
+      "loss": 4.1197,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114764967442214e-05,
+      "loss": 4.129,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113926372691162e-05,
+      "loss": 4.1172,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11308777794011e-05,
+      "loss": 4.1155,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112249183189058e-05,
+      "loss": 4.1101,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111410588438006e-05,
+      "loss": 4.1183,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110573631567327e-05,
+      "loss": 4.1233,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109735036816275e-05,
+      "loss": 4.1147,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108896442065223e-05,
+      "loss": 4.1334,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108057847314171e-05,
+      "loss": 4.1267,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072192525631196e-05,
+      "loss": 4.1303,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063822956924405e-05,
+      "loss": 4.1164,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055437009413885e-05,
+      "loss": 4.1212,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1047051061903365e-05,
+      "loss": 4.1259,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038665114392845e-05,
+      "loss": 4.1232,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030279166882325e-05,
+      "loss": 4.1147,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021893219371805e-05,
+      "loss": 4.1178,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1013507271861285e-05,
+      "loss": 4.1116,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005137703154494e-05,
+      "loss": 4.117,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0996751755643974e-05,
+      "loss": 4.1192,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988365808133454e-05,
+      "loss": 4.1241,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0979979860622933e-05,
+      "loss": 4.1254,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0971593913112413e-05,
+      "loss": 4.1241,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096320796560189e-05,
+      "loss": 4.1185,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095483839689511e-05,
+      "loss": 4.1177,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094646882818832e-05,
+      "loss": 4.1247,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09380828806778e-05,
+      "loss": 4.1086,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092969693316728e-05,
+      "loss": 4.1074,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092131098565676e-05,
+      "loss": 4.1027,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091292503814624e-05,
+      "loss": 4.1116,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090453909063572e-05,
+      "loss": 4.113,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089615314312519e-05,
+      "loss": 4.1164,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088776719561467e-05,
+      "loss": 4.1122,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087938124810415e-05,
+      "loss": 4.1125,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087099530059363e-05,
+      "loss": 4.1137,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086260935308311e-05,
+      "loss": 4.1185,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085423978437633e-05,
+      "loss": 4.1064,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084585383686581e-05,
+      "loss": 4.0905,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0837484268159016e-05,
+      "loss": 4.1263,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829098320648496e-05,
+      "loss": 4.1171,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0820712373137976e-05,
+      "loss": 4.1248,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0812326425627456e-05,
+      "loss": 4.1098,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0803940478116936e-05,
+      "loss": 4.1136,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0795554530606416e-05,
+      "loss": 4.1068,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0787168583095896e-05,
+      "loss": 4.1085,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0778782635585376e-05,
+      "loss": 4.1159,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770413066878585e-05,
+      "loss": 4.1004,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762027119368065e-05,
+      "loss": 4.131,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075364117185755e-05,
+      "loss": 4.1114,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074525522434703e-05,
+      "loss": 4.0914,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073686927683651e-05,
+      "loss": 4.1203,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072849970812972e-05,
+      "loss": 4.0946,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07201137606192e-05,
+      "loss": 4.0932,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071172781310868e-05,
+      "loss": 4.1186,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070334186559816e-05,
+      "loss": 4.1119,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069495591808764e-05,
+      "loss": 4.0999,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068656997057712e-05,
+      "loss": 4.1076,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067820040187033e-05,
+      "loss": 4.0912,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066981445435981e-05,
+      "loss": 4.1017,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066142850684929e-05,
+      "loss": 4.1073,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065304255933877e-05,
+      "loss": 4.1036,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0644672990631985e-05,
+      "loss": 4.1104,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636287043121465e-05,
+      "loss": 4.1092,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627901095610945e-05,
+      "loss": 4.12,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619515148100425e-05,
+      "loss": 4.1008,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611129200589905e-05,
+      "loss": 4.1054,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602759631883114e-05,
+      "loss": 4.1102,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594373684372594e-05,
+      "loss": 4.0987,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585987736862074e-05,
+      "loss": 4.1112,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577601789351554e-05,
+      "loss": 4.1068,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056921584184103e-05,
+      "loss": 4.0965,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056082989433051e-05,
+      "loss": 4.1192,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055244394681999e-05,
+      "loss": 4.0926,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05440743781132e-05,
+      "loss": 4.0941,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053568843060268e-05,
+      "loss": 4.1098,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052730248309216e-05,
+      "loss": 4.1072,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051891653558164e-05,
+      "loss": 4.0997,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051053058807112e-05,
+      "loss": 4.1106,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050216101936433e-05,
+      "loss": 4.0873,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049377507185381e-05,
+      "loss": 4.0967,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048538912434329e-05,
+      "loss": 4.0995,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047700317683277e-05,
+      "loss": 4.0999,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046861722932225e-05,
+      "loss": 4.1168,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046024766061546e-05,
+      "loss": 4.1052,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045186171310494e-05,
+      "loss": 4.1077,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044347576559443e-05,
+      "loss": 4.0948,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043508981808391e-05,
+      "loss": 4.0957,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042670387057339e-05,
+      "loss": 4.0975,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041831792306287e-05,
+      "loss": 4.1084,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040993197555235e-05,
+      "loss": 4.1035,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040154602804183e-05,
+      "loss": 4.1081,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0393176459335036e-05,
+      "loss": 4.0969,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0384790511824516e-05,
+      "loss": 4.1032,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0376404564313996e-05,
+      "loss": 4.0979,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368034995607205e-05,
+      "loss": 4.0987,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359649048096685e-05,
+      "loss": 4.0856,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351263100586165e-05,
+      "loss": 4.1066,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342877153075645e-05,
+      "loss": 4.0908,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0334491205565125e-05,
+      "loss": 4.0981,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0326105258054605e-05,
+      "loss": 4.089,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031773568934782e-05,
+      "loss": 4.0975,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03093497418373e-05,
+      "loss": 4.0952,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030096379432678e-05,
+      "loss": 4.0932,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029257784681626e-05,
+      "loss": 4.1045,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028420827810947e-05,
+      "loss": 4.1122,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027582233059895e-05,
+      "loss": 4.1097,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026743638308843e-05,
+      "loss": 4.0983,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025905043557791e-05,
+      "loss": 4.0911,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025066448806738e-05,
+      "loss": 4.0936,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024227854055686e-05,
+      "loss": 4.1006,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023390897185008e-05,
+      "loss": 4.1033,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022552302433956e-05,
+      "loss": 4.0838,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021713707682904e-05,
+      "loss": 4.1011,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020875112931852e-05,
+      "loss": 4.1022,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200365181808e-05,
+      "loss": 4.0952,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0191995613101214e-05,
+      "loss": 4.0946,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018360966559069e-05,
+      "loss": 4.101,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017522371808017e-05,
+      "loss": 4.0902,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016683777056965e-05,
+      "loss": 4.1028,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015845182305913e-05,
+      "loss": 4.0999,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015008225435234e-05,
+      "loss": 4.1049,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0141696306841817e-05,
+      "loss": 4.0823,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133310359331297e-05,
+      "loss": 4.1052,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012492441182078e-05,
+      "loss": 4.102,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011653846431026e-05,
+      "loss": 4.094,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010816889560347e-05,
+      "loss": 4.0991,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009978294809295e-05,
+      "loss": 4.0959,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009139700058243e-05,
+      "loss": 4.0988,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008301105307191e-05,
+      "loss": 4.1128,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007462510556139e-05,
+      "loss": 4.0869,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00662555368546e-05,
+      "loss": 4.1011,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005786958934408e-05,
+      "loss": 4.0957,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004948364183356e-05,
+      "loss": 4.097,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004109769432304e-05,
+      "loss": 4.0882,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003271174681252e-05,
+      "loss": 4.0835,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002434217810574e-05,
+      "loss": 4.0934,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001595623059522e-05,
+      "loss": 4.0854,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00075702830847e-05,
+      "loss": 4.1006,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.106624126434326,
+      "eval_runtime": 297.2422,
+      "eval_samples_per_second": 1283.771,
+      "eval_steps_per_second": 40.119,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999918433557418e-05,
+      "loss": 4.0956,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999079838806366e-05,
+      "loss": 4.0894,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998241244055314e-05,
+      "loss": 4.096,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997402649304262e-05,
+      "loss": 4.0955,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99656405455321e-05,
+      "loss": 4.1006,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9957254598021577e-05,
+      "loss": 4.0933,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994886865051105e-05,
+      "loss": 4.0983,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994048270300053e-05,
+      "loss": 4.0825,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.993209675549001e-05,
+      "loss": 4.0852,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.992371080797949e-05,
+      "loss": 4.0994,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.991532486046897e-05,
+      "loss": 4.0927,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.990693891295845e-05,
+      "loss": 4.0866,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989855296544793e-05,
+      "loss": 4.0975,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890167017937416e-05,
+      "loss": 4.0837,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9881781070426896e-05,
+      "loss": 4.0857,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9873395122916376e-05,
+      "loss": 4.0733,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9865009175405856e-05,
+      "loss": 4.0886,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9856623227895336e-05,
+      "loss": 4.0954,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9848253659188545e-05,
+      "loss": 4.0813,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9839867711678025e-05,
+      "loss": 4.1009,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9831481764167505e-05,
+      "loss": 4.0963,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9823095816656985e-05,
+      "loss": 4.0964,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981470986914646e-05,
+      "loss": 4.0885,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9806340300439674e-05,
+      "loss": 4.092,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9797954352929154e-05,
+      "loss": 4.093,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9789568405418634e-05,
+      "loss": 4.0898,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9781182457908114e-05,
+      "loss": 4.0832,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9772796510397594e-05,
+      "loss": 4.0896,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976442694169081e-05,
+      "loss": 4.0802,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975604099418029e-05,
+      "loss": 4.0844,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974765504666976e-05,
+      "loss": 4.0881,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973926909915924e-05,
+      "loss": 4.0907,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973088315164872e-05,
+      "loss": 4.0957,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.97224972041382e-05,
+      "loss": 4.0966,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971412763543141e-05,
+      "loss": 4.0816,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.970574168792089e-05,
+      "loss": 4.0914,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.969735574041037e-05,
+      "loss": 4.0943,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968896979289985e-05,
+      "loss": 4.0754,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.968058384538934e-05,
+      "loss": 4.0789,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967219789787882e-05,
+      "loss": 4.0761,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.96638119503683e-05,
+      "loss": 4.0754,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965544238166151e-05,
+      "loss": 4.084,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964705643415099e-05,
+      "loss": 4.0866,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963867048664047e-05,
+      "loss": 4.0805,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963028453912995e-05,
+      "loss": 4.083,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9621914970423156e-05,
+      "loss": 4.0818,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613529022912636e-05,
+      "loss": 4.0908,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9605143075402116e-05,
+      "loss": 4.0735,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9596757127891596e-05,
+      "loss": 4.0625,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9588371180381076e-05,
+      "loss": 4.094,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9579985232870556e-05,
+      "loss": 4.0869,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9571599285360036e-05,
+      "loss": 4.0927,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956321333784952e-05,
+      "loss": 4.0789,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9554827390339e-05,
+      "loss": 4.0818,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954645782163221e-05,
+      "loss": 4.0808,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953807187412169e-05,
+      "loss": 4.0788,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952968592661117e-05,
+      "loss": 4.0864,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952131635790438e-05,
+      "loss": 4.068,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951293041039386e-05,
+      "loss": 4.1032,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950454446288334e-05,
+      "loss": 4.0783,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949615851537282e-05,
+      "loss": 4.0604,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9487772567862294e-05,
+      "loss": 4.0923,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479386620351774e-05,
+      "loss": 4.0709,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947100067284126e-05,
+      "loss": 4.0603,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.946261472533074e-05,
+      "loss": 4.0868,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.945424515662395e-05,
+      "loss": 4.0842,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445875587917166e-05,
+      "loss": 4.0724,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437489640406646e-05,
+      "loss": 4.0762,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429103692896126e-05,
+      "loss": 4.0602,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.94207177453856e-05,
+      "loss": 4.0692,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941233179787508e-05,
+      "loss": 4.0816,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940394585036456e-05,
+      "loss": 4.0739,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939555990285404e-05,
+      "loss": 4.0778,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938719033414725e-05,
+      "loss": 4.0763,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937880438663673e-05,
+      "loss": 4.0924,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370418439126214e-05,
+      "loss": 4.0753,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9362032491615694e-05,
+      "loss": 4.0752,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9353646544105174e-05,
+      "loss": 4.0789,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9345260596594654e-05,
+      "loss": 4.0721,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9336874649084134e-05,
+      "loss": 4.0819,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9328488701573614e-05,
+      "loss": 4.0764,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9320102754063094e-05,
+      "loss": 4.0707,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9311716806552574e-05,
+      "loss": 4.0878,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930336361664951e-05,
+      "loss": 4.0631,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.929497766913899e-05,
+      "loss": 4.0694,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928659172162847e-05,
+      "loss": 4.078,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927822215292168e-05,
+      "loss": 4.0786,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926983620541116e-05,
+      "loss": 4.0684,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926145025790065e-05,
+      "loss": 4.0814,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.925306431039013e-05,
+      "loss": 4.0639,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924467836287961e-05,
+      "loss": 4.0691,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.923629241536909e-05,
+      "loss": 4.0654,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.922790646785857e-05,
+      "loss": 4.0716,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921952052034805e-05,
+      "loss": 4.088,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921113457283753e-05,
+      "loss": 4.0752,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920274862532701e-05,
+      "loss": 4.0775,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919436267781648e-05,
+      "loss": 4.0684,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918597673030596e-05,
+      "loss": 4.066,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917759078279544e-05,
+      "loss": 4.0682,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916920483528492e-05,
+      "loss": 4.0807,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91608188877744e-05,
+      "loss": 4.0757,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9152449319067617e-05,
+      "loss": 4.074,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9144063371557096e-05,
+      "loss": 4.0762,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9135677424046576e-05,
+      "loss": 4.0712,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9127291476536056e-05,
+      "loss": 4.0707,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9118905529025536e-05,
+      "loss": 4.0689,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9110519581515016e-05,
+      "loss": 4.0605,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9102133634004496e-05,
+      "loss": 4.0758,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9093747686493976e-05,
+      "loss": 4.0642,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9085394496590914e-05,
+      "loss": 4.069,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9077008549080394e-05,
+      "loss": 4.066,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9068622601569874e-05,
+      "loss": 4.0635,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9060236654059354e-05,
+      "loss": 4.0674,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9051850706548834e-05,
+      "loss": 4.0649,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904348113784205e-05,
+      "loss": 4.0819,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903509519033153e-05,
+      "loss": 4.0794,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.902670924282101e-05,
+      "loss": 4.079,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901832329531049e-05,
+      "loss": 4.0695,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900993734779997e-05,
+      "loss": 4.067,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900155140028945e-05,
+      "loss": 4.0653,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899316545277893e-05,
+      "loss": 4.0727,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898479588407214e-05,
+      "loss": 4.0773,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897640993656162e-05,
+      "loss": 4.0585,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89680239890511e-05,
+      "loss": 4.0691,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895963804154058e-05,
+      "loss": 4.0717,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895126847283379e-05,
+      "loss": 4.0701,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894288252532327e-05,
+      "loss": 4.0658,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8934496577812755e-05,
+      "loss": 4.0712,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8926110630302235e-05,
+      "loss": 4.0616,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8917724682791715e-05,
+      "loss": 4.0768,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8909338735281195e-05,
+      "loss": 4.0673,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8900969166574404e-05,
+      "loss": 4.0818,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8892583219063884e-05,
+      "loss": 4.0572,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8884197271553364e-05,
+      "loss": 4.0747,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875811324042843e-05,
+      "loss": 4.0788,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886744175533605e-05,
+      "loss": 4.0668,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885905580782553e-05,
+      "loss": 4.0684,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885066986031501e-05,
+      "loss": 4.0701,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884228391280449e-05,
+      "loss": 4.069,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883389796529397e-05,
+      "loss": 4.0832,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882552839658719e-05,
+      "loss": 4.0632,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881714244907667e-05,
+      "loss": 4.0736,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880875650156615e-05,
+      "loss": 4.0703,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880037055405562e-05,
+      "loss": 4.0682,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87919846065451e-05,
+      "loss": 4.0628,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878361503783832e-05,
+      "loss": 4.0579,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877522909032779e-05,
+      "loss": 4.0634,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876684314281727e-05,
+      "loss": 4.0642,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875845719530675e-05,
+      "loss": 4.0695,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.087048053741455,
+      "eval_runtime": 290.952,
+      "eval_samples_per_second": 1311.526,
+      "eval_steps_per_second": 40.986,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875007124779623e-05,
+      "loss": 4.0661,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874170167908944e-05,
+      "loss": 4.0625,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8733315731578926e-05,
+      "loss": 4.0649,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8724929784068406e-05,
+      "loss": 4.0724,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8716543836557886e-05,
+      "loss": 4.0773,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8708174267851095e-05,
+      "loss": 4.0668,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8699788320340575e-05,
+      "loss": 4.066,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8691402372830055e-05,
+      "loss": 4.0571,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8683032804123264e-05,
+      "loss": 4.0595,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674646856612744e-05,
+      "loss": 4.0689,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666260909102224e-05,
+      "loss": 4.0688,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657874961591704e-05,
+      "loss": 4.0615,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8649489014081184e-05,
+      "loss": 4.0692,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641103066570664e-05,
+      "loss": 4.0582,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8632717119060144e-05,
+      "loss": 4.0595,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862434755035336e-05,
+      "loss": 4.045,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861596160284284e-05,
+      "loss": 4.0667,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.860757565533232e-05,
+      "loss": 4.0677,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85991897078218e-05,
+      "loss": 4.053,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.859080376031128e-05,
+      "loss": 4.071,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858241781280076e-05,
+      "loss": 4.0691,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857403186529024e-05,
+      "loss": 4.0711,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856564591777972e-05,
+      "loss": 4.0631,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855727634907293e-05,
+      "loss": 4.0651,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854889040156241e-05,
+      "loss": 4.0644,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854050445405189e-05,
+      "loss": 4.0692,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853211850654137e-05,
+      "loss": 4.0506,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852373255903085e-05,
+      "loss": 4.0637,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851534661152033e-05,
+      "loss": 4.0524,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8506977042813544e-05,
+      "loss": 4.0623,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8498591095303024e-05,
+      "loss": 4.0638,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8490205147792504e-05,
+      "loss": 4.0656,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848181920028198e-05,
+      "loss": 4.0615,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847343325277146e-05,
+      "loss": 4.0744,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846504730526094e-05,
+      "loss": 4.0567,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845667773655415e-05,
+      "loss": 4.0626,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448291789043626e-05,
+      "loss": 4.0677,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8439905841533106e-05,
+      "loss": 4.0483,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8431519894022586e-05,
+      "loss": 4.052,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8423133946512066e-05,
+      "loss": 4.0531,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414747999001546e-05,
+      "loss": 4.0483,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840636205149103e-05,
+      "loss": 4.0573,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839797610398051e-05,
+      "loss": 4.064,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838960653527372e-05,
+      "loss": 4.0501,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83812205877632e-05,
+      "loss": 4.0568,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837283464025268e-05,
+      "loss": 4.0539,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836444869274216e-05,
+      "loss": 4.0606,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835607912403537e-05,
+      "loss": 4.0482,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834770955532858e-05,
+      "loss": 4.0388,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833932360781806e-05,
+      "loss": 4.0683,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833093766030754e-05,
+      "loss": 4.0611,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832255171279702e-05,
+      "loss": 4.0718,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83141657652865e-05,
+      "loss": 4.0502,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8305779817775986e-05,
+      "loss": 4.0551,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297410249069195e-05,
+      "loss": 4.0544,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8289024301558675e-05,
+      "loss": 4.0489,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8280638354048155e-05,
+      "loss": 4.0647,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8272252406537635e-05,
+      "loss": 4.0384,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8263866459027115e-05,
+      "loss": 4.0804,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8255496890320324e-05,
+      "loss": 4.0564,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8247127321613534e-05,
+      "loss": 4.0324,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8238741374103013e-05,
+      "loss": 4.0625,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8230355426592493e-05,
+      "loss": 4.0488,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822196947908197e-05,
+      "loss": 4.0358,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821358353157145e-05,
+      "loss": 4.0584,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820519758406093e-05,
+      "loss": 4.0621,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819681163655042e-05,
+      "loss": 4.0458,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81884256890399e-05,
+      "loss": 4.0529,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.818003974152938e-05,
+      "loss": 4.0312,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817167017282259e-05,
+      "loss": 4.0451,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.816328422531207e-05,
+      "loss": 4.0525,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.815489827780155e-05,
+      "loss": 4.0467,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814651233029103e-05,
+      "loss": 4.0588,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813812638278051e-05,
+      "loss": 4.0476,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812974043526999e-05,
+      "loss": 4.0708,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812135448775946e-05,
+      "loss": 4.0492,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811298491905268e-05,
+      "loss": 4.053,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810459897154216e-05,
+      "loss": 4.0546,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809621302403164e-05,
+      "loss": 4.0466,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.808782707652112e-05,
+      "loss": 4.0551,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80794411290106e-05,
+      "loss": 4.0474,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807105518150008e-05,
+      "loss": 4.0516,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806266923398956e-05,
+      "loss": 4.0583,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805428328647904e-05,
+      "loss": 4.0474,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804591371777225e-05,
+      "loss": 4.0361,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037527770261727e-05,
+      "loss": 4.0511,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029158201554936e-05,
+      "loss": 4.0572,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8020772254044416e-05,
+      "loss": 4.0433,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8012386306533896e-05,
+      "loss": 4.0537,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8004000359023376e-05,
+      "loss": 4.0387,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7995614411512856e-05,
+      "loss": 4.045,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798724484280607e-05,
+      "loss": 4.0405,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797885889529555e-05,
+      "loss": 4.0452,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797047294778503e-05,
+      "loss": 4.06,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796208700027451e-05,
+      "loss": 4.0525,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.795370105276399e-05,
+      "loss": 4.0538,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79453314840572e-05,
+      "loss": 4.0462,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.793694553654668e-05,
+      "loss": 4.0408,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792855958903616e-05,
+      "loss": 4.0444,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792017364152564e-05,
+      "loss": 4.0534,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791178769401512e-05,
+      "loss": 4.0518,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79034017465046e-05,
+      "loss": 4.0523,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789501579899408e-05,
+      "loss": 4.0502,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788662985148356e-05,
+      "loss": 4.0451,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787824390397304e-05,
+      "loss": 4.0498,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7869874335266256e-05,
+      "loss": 4.0438,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7861488387755736e-05,
+      "loss": 4.0391,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7853102440245216e-05,
+      "loss": 4.0486,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7844716492734696e-05,
+      "loss": 4.0413,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836346924027905e-05,
+      "loss": 4.043,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7827960976517385e-05,
+      "loss": 4.0384,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7819575029006865e-05,
+      "loss": 4.0387,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7811189081496345e-05,
+      "loss": 4.0464,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7802803133985825e-05,
+      "loss": 4.0394,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794433565279034e-05,
+      "loss": 4.0556,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778606399657225e-05,
+      "loss": 4.0543,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777767804906173e-05,
+      "loss": 4.058,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776930848035494e-05,
+      "loss": 4.0466,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776092253284442e-05,
+      "loss": 4.044,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77525365853339e-05,
+      "loss": 4.0386,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.774415063782338e-05,
+      "loss": 4.0545,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.773576469031286e-05,
+      "loss": 4.0505,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772737874280234e-05,
+      "loss": 4.0341,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771899279529182e-05,
+      "loss": 4.0436,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77106068477813e-05,
+      "loss": 4.0463,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770222090027077e-05,
+      "loss": 4.0494,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769385133156399e-05,
+      "loss": 4.0427,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768546538405347e-05,
+      "loss": 4.0444,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767707943654295e-05,
+      "loss": 4.0388,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766869348903243e-05,
+      "loss": 4.0543,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766030754152191e-05,
+      "loss": 4.0442,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765192159401139e-05,
+      "loss": 4.0551,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764353564650087e-05,
+      "loss": 4.0354,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763514969899035e-05,
+      "loss": 4.0495,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7626780130283556e-05,
+      "loss": 4.055,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7618394182773036e-05,
+      "loss": 4.0406,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7610024614066245e-05,
+      "loss": 4.048,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7601638666555725e-05,
+      "loss": 4.046,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7593252719045205e-05,
+      "loss": 4.0483,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7584866771534685e-05,
+      "loss": 4.0552,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7576480824024165e-05,
+      "loss": 4.0449,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756811125531738e-05,
+      "loss": 4.0464,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755972530780686e-05,
+      "loss": 4.0485,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755133936029634e-05,
+      "loss": 4.0451,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.754295341278582e-05,
+      "loss": 4.0391,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75345674652753e-05,
+      "loss": 4.0351,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752619789656851e-05,
+      "loss": 4.0402,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751782832786172e-05,
+      "loss": 4.041,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.75094423803512e-05,
+      "loss": 4.0465,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.071537017822266,
+      "eval_runtime": 290.8862,
+      "eval_samples_per_second": 1311.822,
+      "eval_steps_per_second": 40.995,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750105643284068e-05,
+      "loss": 4.047,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749267048533016e-05,
+      "loss": 4.0377,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748428453781964e-05,
+      "loss": 4.0415,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747589859030912e-05,
+      "loss": 4.0506,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467512642798605e-05,
+      "loss": 4.0532,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459126695288085e-05,
+      "loss": 4.0437,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7450740747777565e-05,
+      "loss": 4.0465,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7442354800267045e-05,
+      "loss": 4.0367,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433968852756525e-05,
+      "loss": 4.0352,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425582905246005e-05,
+      "loss": 4.0426,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417196957735485e-05,
+      "loss": 4.0501,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740881101022496e-05,
+      "loss": 4.0398,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7400441441518174e-05,
+      "loss": 4.0458,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7392055494007654e-05,
+      "loss": 4.0345,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738366954649713e-05,
+      "loss": 4.0389,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737528359898661e-05,
+      "loss": 4.0201,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736691403027982e-05,
+      "loss": 4.045,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73585280827693e-05,
+      "loss": 4.043,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735014213525878e-05,
+      "loss": 4.0307,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7341772566552e-05,
+      "loss": 4.0492,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733338661904148e-05,
+      "loss": 4.0493,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732500067153096e-05,
+      "loss": 4.0478,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731661472402043e-05,
+      "loss": 4.0388,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730822877650991e-05,
+      "loss": 4.0423,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729984282899939e-05,
+      "loss": 4.0417,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729145688148887e-05,
+      "loss": 4.0451,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.728307093397835e-05,
+      "loss": 4.0293,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.727468498646783e-05,
+      "loss": 4.041,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726629903895731e-05,
+      "loss": 4.0342,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725791309144679e-05,
+      "loss": 4.0342,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724952714393627e-05,
+      "loss": 4.0418,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724114119642576e-05,
+      "loss": 4.0436,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723275524891524e-05,
+      "loss": 4.0411,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722436930140472e-05,
+      "loss": 4.0487,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.72159833538942e-05,
+      "loss": 4.0355,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720761378518741e-05,
+      "loss": 4.0409,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719922783767689e-05,
+      "loss": 4.0469,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719084189016637e-05,
+      "loss": 4.0249,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718245594265585e-05,
+      "loss": 4.0279,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717406999514532e-05,
+      "loss": 4.0318,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71656840476348e-05,
+      "loss": 4.0218,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7157314478928016e-05,
+      "loss": 4.0366,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7148928531417496e-05,
+      "loss": 4.0447,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140542583906976e-05,
+      "loss": 4.0285,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7132156636396456e-05,
+      "loss": 4.0364,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123770688885936e-05,
+      "loss": 4.0364,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7115384741375416e-05,
+      "loss": 4.0375,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7106998793864896e-05,
+      "loss": 4.0259,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7098629225158105e-05,
+      "loss": 4.0157,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7090243277647585e-05,
+      "loss": 4.0432,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7081857330137065e-05,
+      "loss": 4.0433,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7073471382626545e-05,
+      "loss": 4.0522,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7065085435116025e-05,
+      "loss": 4.0292,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7056715866409234e-05,
+      "loss": 4.0321,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7048329918898714e-05,
+      "loss": 4.0328,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7039943971388194e-05,
+      "loss": 4.0264,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7031558023877674e-05,
+      "loss": 4.0453,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702317207636716e-05,
+      "loss": 4.0143,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701478612885664e-05,
+      "loss": 4.0596,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700641656014985e-05,
+      "loss": 4.0367,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699803061263933e-05,
+      "loss": 4.014,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698964466512881e-05,
+      "loss": 4.04,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.698125871761829e-05,
+      "loss": 4.0276,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.697287277010777e-05,
+      "loss": 4.0132,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696448682259725e-05,
+      "loss": 4.0385,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695611725389046e-05,
+      "loss": 4.0389,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694773130637994e-05,
+      "loss": 4.0231,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693934535886942e-05,
+      "loss": 4.0369,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693097579016263e-05,
+      "loss": 4.0096,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6922589842652114e-05,
+      "loss": 4.0254,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6914203895141594e-05,
+      "loss": 4.027,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6905817947631074e-05,
+      "loss": 4.031,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6897432000120554e-05,
+      "loss": 4.0335,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6889046052610034e-05,
+      "loss": 4.0285,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.688066010509951e-05,
+      "loss": 4.0503,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.687227415758899e-05,
+      "loss": 4.0299,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686388821007847e-05,
+      "loss": 4.0334,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685553502017541e-05,
+      "loss": 4.0289,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684714907266489e-05,
+      "loss": 4.0255,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683876312515437e-05,
+      "loss": 4.0346,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683037717764385e-05,
+      "loss": 4.0257,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682199123013333e-05,
+      "loss": 4.0304,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681360528262281e-05,
+      "loss": 4.0425,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680521933511229e-05,
+      "loss": 4.0265,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679683338760177e-05,
+      "loss": 4.0148,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678846381889498e-05,
+      "loss": 4.0284,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.67800942501882e-05,
+      "loss": 4.0379,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677170830267768e-05,
+      "loss": 4.0184,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676332235516715e-05,
+      "loss": 4.0361,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675493640765663e-05,
+      "loss": 4.0237,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674655046014611e-05,
+      "loss": 4.0171,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6738180891439326e-05,
+      "loss": 4.0222,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6729794943928806e-05,
+      "loss": 4.025,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6721408996418286e-05,
+      "loss": 4.044,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6713023048907766e-05,
+      "loss": 4.0285,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704637101397246e-05,
+      "loss": 4.0317,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696251153886726e-05,
+      "loss": 4.0234,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6687865206376206e-05,
+      "loss": 4.0193,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6679479258865685e-05,
+      "loss": 4.0221,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6671093311355165e-05,
+      "loss": 4.0347,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6662707363844645e-05,
+      "loss": 4.0341,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6654321416334125e-05,
+      "loss": 4.0299,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6645935468823605e-05,
+      "loss": 4.0307,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6637565900116814e-05,
+      "loss": 4.0196,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6629179952606294e-05,
+      "loss": 4.0307,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6620794005095774e-05,
+      "loss": 4.0247,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.661244081519272e-05,
+      "loss": 4.0182,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66040548676822e-05,
+      "loss": 4.0245,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.659566892017168e-05,
+      "loss": 4.0229,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.658728297266116e-05,
+      "loss": 4.0242,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657889702515064e-05,
+      "loss": 4.0175,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657051107764012e-05,
+      "loss": 4.018,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65621251301296e-05,
+      "loss": 4.0236,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655373918261908e-05,
+      "loss": 4.0263,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.654535323510856e-05,
+      "loss": 4.0276,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653696728759804e-05,
+      "loss": 4.034,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652858134008751e-05,
+      "loss": 4.0401,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652019539257699e-05,
+      "loss": 4.0256,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.651182582387021e-05,
+      "loss": 4.0252,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6503456255163424e-05,
+      "loss": 4.0181,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6495070307652904e-05,
+      "loss": 4.0308,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6486684360142384e-05,
+      "loss": 4.0292,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6478298412631864e-05,
+      "loss": 4.0186,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6469912465121344e-05,
+      "loss": 4.0197,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.646152651761082e-05,
+      "loss": 4.0309,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64531405701003e-05,
+      "loss": 4.0286,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644475462258978e-05,
+      "loss": 4.0204,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6436385053882986e-05,
+      "loss": 4.023,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6427999106372466e-05,
+      "loss": 4.0224,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6419613158861946e-05,
+      "loss": 4.0309,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6411227211351426e-05,
+      "loss": 4.0286,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640285764264464e-05,
+      "loss": 4.0347,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639447169513412e-05,
+      "loss": 4.0156,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63860857476236e-05,
+      "loss": 4.0303,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637769980011308e-05,
+      "loss": 4.034,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636933023140629e-05,
+      "loss": 4.0168,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636094428389577e-05,
+      "loss": 4.0253,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635255833638525e-05,
+      "loss": 4.028,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634417238887473e-05,
+      "loss": 4.031,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633578644136421e-05,
+      "loss": 4.0349,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632740049385369e-05,
+      "loss": 4.0229,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6319047303950635e-05,
+      "loss": 4.0263,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310661356440115e-05,
+      "loss": 4.0259,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6302275408929595e-05,
+      "loss": 4.0216,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.629390584022281e-05,
+      "loss": 4.0188,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6285519892712284e-05,
+      "loss": 4.0175,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6277133945201764e-05,
+      "loss": 4.0149,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6268747997691244e-05,
+      "loss": 4.0236,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6260362050180724e-05,
+      "loss": 4.0224,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.058211326599121,
+      "eval_runtime": 297.8122,
+      "eval_samples_per_second": 1281.314,
+      "eval_steps_per_second": 40.042,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251976102670204e-05,
+      "loss": 4.0236,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243590155159684e-05,
+      "loss": 4.0202,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623522058645289e-05,
+      "loss": 4.024,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622683463894237e-05,
+      "loss": 4.0274,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621844869143185e-05,
+      "loss": 4.0324,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621006274392133e-05,
+      "loss": 4.024,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620167679641081e-05,
+      "loss": 4.0268,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61932908489003e-05,
+      "loss": 4.0182,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618490490138978e-05,
+      "loss": 4.0131,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617651895387926e-05,
+      "loss": 4.0197,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616813300636874e-05,
+      "loss": 4.0334,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615974705885822e-05,
+      "loss": 4.0154,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615137749015143e-05,
+      "loss": 4.028,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614299154264091e-05,
+      "loss": 4.0123,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613460559513039e-05,
+      "loss": 4.0156,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612621964761987e-05,
+      "loss": 4.003,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.611783370010935e-05,
+      "loss": 4.0248,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.610944775259882e-05,
+      "loss": 4.0226,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.61010618050883e-05,
+      "loss": 4.0151,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609267585757778e-05,
+      "loss": 4.0183,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608428991006727e-05,
+      "loss": 4.0338,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607590396255675e-05,
+      "loss": 4.0249,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606753439384996e-05,
+      "loss": 4.0227,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605914844633944e-05,
+      "loss": 4.0205,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605076249882892e-05,
+      "loss": 4.0282,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60423765513184e-05,
+      "loss": 4.0215,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603399060380788e-05,
+      "loss": 4.014,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602560465629736e-05,
+      "loss": 4.0186,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601721870878684e-05,
+      "loss": 4.0138,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600883276127632e-05,
+      "loss": 4.0106,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.60004468137658e-05,
+      "loss": 4.0241,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5992077245059006e-05,
+      "loss": 4.0173,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5983691297548486e-05,
+      "loss": 4.027,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5975305350037966e-05,
+      "loss": 4.0271,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5966919402527446e-05,
+      "loss": 4.0136,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595854983382066e-05,
+      "loss": 4.0223,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595016388631014e-05,
+      "loss": 4.0248,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594177793879962e-05,
+      "loss": 4.009,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593340837009283e-05,
+      "loss": 4.0133,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592502242258231e-05,
+      "loss": 4.007,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591663647507179e-05,
+      "loss": 4.0055,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590825052756127e-05,
+      "loss": 4.0175,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589986458005075e-05,
+      "loss": 4.0261,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589147863254023e-05,
+      "loss": 4.0093,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588309268502971e-05,
+      "loss": 4.0118,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874706737519184e-05,
+      "loss": 4.0177,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586632079000867e-05,
+      "loss": 4.0137,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585793484249815e-05,
+      "loss": 4.0095,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5849565273791366e-05,
+      "loss": 3.9976,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584117932628084e-05,
+      "loss": 4.0187,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583279337877032e-05,
+      "loss": 4.0258,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58244074312598e-05,
+      "loss": 4.0274,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581602148374928e-05,
+      "loss": 4.0111,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580765191504249e-05,
+      "loss": 4.0112,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5799282346335704e-05,
+      "loss": 4.0133,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5790896398825184e-05,
+      "loss": 4.0052,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578251045131466e-05,
+      "loss": 4.0247,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577412450380414e-05,
+      "loss": 3.9971,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5765738556293624e-05,
+      "loss": 4.0347,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5757352608783104e-05,
+      "loss": 4.0156,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5748966661272584e-05,
+      "loss": 3.9971,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5740580713762064e-05,
+      "loss": 4.0161,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573221114505527e-05,
+      "loss": 4.0093,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572382519754475e-05,
+      "loss": 3.9948,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571543925003423e-05,
+      "loss": 4.0151,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570705330252371e-05,
+      "loss": 4.0215,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569866735501319e-05,
+      "loss": 4.0036,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569028140750267e-05,
+      "loss": 4.0173,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.568191183879588e-05,
+      "loss": 3.9905,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.567352589128536e-05,
+      "loss": 4.0078,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566513994377484e-05,
+      "loss": 4.0079,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565675399626432e-05,
+      "loss": 4.0132,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564836804875381e-05,
+      "loss": 4.0093,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563999848004702e-05,
+      "loss": 4.0118,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56316125325365e-05,
+      "loss": 4.0292,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562324296382971e-05,
+      "loss": 4.0135,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561485701631919e-05,
+      "loss": 4.013,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5606471068808667e-05,
+      "loss": 4.0093,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5598085121298147e-05,
+      "loss": 4.0064,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5589699173787627e-05,
+      "loss": 4.0169,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5581313226277106e-05,
+      "loss": 4.0085,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5572943657570316e-05,
+      "loss": 4.0071,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5564557710059796e-05,
+      "loss": 4.0201,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5556171762549275e-05,
+      "loss": 4.0095,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5547785815038755e-05,
+      "loss": 3.9918,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553939986752824e-05,
+      "loss": 4.0106,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553101392001772e-05,
+      "loss": 4.0147,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552264435131093e-05,
+      "loss": 4.002,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551425840380041e-05,
+      "loss": 4.0139,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550587245628989e-05,
+      "loss": 4.0066,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549748650877937e-05,
+      "loss": 3.9978,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5489100561268844e-05,
+      "loss": 4.0081,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5480714613758324e-05,
+      "loss": 4.0019,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5472328666247804e-05,
+      "loss": 4.0226,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546395909754102e-05,
+      "loss": 4.0124,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545558952883423e-05,
+      "loss": 4.0164,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544720358132371e-05,
+      "loss": 4.0039,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5438817633813196e-05,
+      "loss": 4.0003,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543043168630267e-05,
+      "loss": 4.0077,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542204573879215e-05,
+      "loss": 4.0148,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541365979128163e-05,
+      "loss": 4.0143,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540527384377111e-05,
+      "loss": 4.0091,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539688789626059e-05,
+      "loss": 4.0094,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53885183275538e-05,
+      "loss": 4.0006,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538013238004328e-05,
+      "loss": 4.015,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537174643253276e-05,
+      "loss": 4.0033,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536336048502224e-05,
+      "loss": 4.0066,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535497453751172e-05,
+      "loss": 4.0037,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.53465885900012e-05,
+      "loss": 4.0032,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533820264249068e-05,
+      "loss": 4.0078,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5329816694980164e-05,
+      "loss": 3.9951,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5321430747469644e-05,
+      "loss": 3.9976,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313044799959124e-05,
+      "loss": 4.008,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304658852448604e-05,
+      "loss": 4.0059,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296272904938084e-05,
+      "loss": 4.01,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528790333623129e-05,
+      "loss": 4.0145,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527951738872077e-05,
+      "loss": 4.0226,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527114782001398e-05,
+      "loss": 4.0049,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526276187250346e-05,
+      "loss": 4.0033,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525437592499294e-05,
+      "loss": 4.0036,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524598997748242e-05,
+      "loss": 4.0116,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52376040299719e-05,
+      "loss": 4.0076,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522923446126512e-05,
+      "loss": 4.0028,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52208485137546e-05,
+      "loss": 4.0014,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521246256624408e-05,
+      "loss": 4.0121,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520407661873356e-05,
+      "loss": 4.0032,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519570705002677e-05,
+      "loss": 4.0052,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518732110251625e-05,
+      "loss": 4.0076,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517893515500573e-05,
+      "loss": 4.0051,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517054920749521e-05,
+      "loss": 4.0065,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516216325998468e-05,
+      "loss": 4.0115,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153793691277896e-05,
+      "loss": 4.0139,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145407743767376e-05,
+      "loss": 4.0005,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137021796256856e-05,
+      "loss": 4.0124,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128635848746336e-05,
+      "loss": 4.0132,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512026628003955e-05,
+      "loss": 4.0022,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511188033252903e-05,
+      "loss": 4.0054,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510351076382224e-05,
+      "loss": 4.0081,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509512481631172e-05,
+      "loss": 4.0166,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50867388688012e-05,
+      "loss": 4.016,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507835292129068e-05,
+      "loss": 4.0078,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5069966973780154e-05,
+      "loss": 4.0079,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5061581026269634e-05,
+      "loss": 4.0112,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5053195078759114e-05,
+      "loss": 4.0092,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5044809131248594e-05,
+      "loss": 3.9976,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5036423183738074e-05,
+      "loss": 4.0074,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5028037236227554e-05,
+      "loss": 3.9939,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501965128871704e-05,
+      "loss": 4.0064,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501126534120652e-05,
+      "loss": 4.0057,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.047008991241455,
+      "eval_runtime": 298.4825,
+      "eval_samples_per_second": 1278.437,
+      "eval_steps_per_second": 39.952,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002879393696e-05,
+      "loss": 4.0164,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499449344618548e-05,
+      "loss": 3.9985,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498610749867496e-05,
+      "loss": 4.0043,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497772155116444e-05,
+      "loss": 4.0151,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496933560365392e-05,
+      "loss": 4.0115,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496094965614339e-05,
+      "loss": 4.0135,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495256370863287e-05,
+      "loss": 4.0087,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494417776112235e-05,
+      "loss": 3.9995,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493579181361183e-05,
+      "loss": 3.9941,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492740586610131e-05,
+      "loss": 4.0077,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491901991859079e-05,
+      "loss": 4.0056,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491063397108027e-05,
+      "loss": 4.0046,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490226440237349e-05,
+      "loss": 4.012,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489387845486297e-05,
+      "loss": 3.9931,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488549250735245e-05,
+      "loss": 4.0028,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487710655984193e-05,
+      "loss": 3.9839,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486872061233141e-05,
+      "loss": 4.0094,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486033466482089e-05,
+      "loss": 4.0027,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48519650961141e-05,
+      "loss": 4.0027,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484357914860358e-05,
+      "loss": 4.0047,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483519320109306e-05,
+      "loss": 4.013,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482680725358254e-05,
+      "loss": 4.0095,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481842130607202e-05,
+      "loss": 4.0034,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.48100353585615e-05,
+      "loss": 4.0025,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480164941105098e-05,
+      "loss": 4.0154,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479326346354046e-05,
+      "loss": 4.0029,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478487751602994e-05,
+      "loss": 3.9961,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477650794732315e-05,
+      "loss": 4.0063,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476812199981263e-05,
+      "loss": 3.9938,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4759736052302106e-05,
+      "loss": 4.0008,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4751350104791586e-05,
+      "loss": 4.0068,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47429805360848e-05,
+      "loss": 3.9994,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473459458857428e-05,
+      "loss": 4.0058,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726208641063755e-05,
+      "loss": 4.013,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717822693553235e-05,
+      "loss": 3.9999,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470945312484645e-05,
+      "loss": 4.0082,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470106717733593e-05,
+      "loss": 4.0097,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469268122982541e-05,
+      "loss": 3.9916,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468429528231489e-05,
+      "loss": 3.9998,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467590933480437e-05,
+      "loss": 3.9901,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466752338729385e-05,
+      "loss": 3.9901,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465913743978333e-05,
+      "loss": 4.0006,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465075149227281e-05,
+      "loss": 4.008,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464236554476229e-05,
+      "loss": 3.9908,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.463397959725177e-05,
+      "loss": 3.9989,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462561002854498e-05,
+      "loss": 4.0018,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461722408103446e-05,
+      "loss": 4.0009,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460885451232767e-05,
+      "loss": 3.9936,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460046856481715e-05,
+      "loss": 3.9826,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459208261730663e-05,
+      "loss": 4.0009,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458369666979611e-05,
+      "loss": 4.0135,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575310722285595e-05,
+      "loss": 4.011,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566924774775075e-05,
+      "loss": 3.9909,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558538827264555e-05,
+      "loss": 3.9967,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550152879754035e-05,
+      "loss": 4.0011,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541783311047244e-05,
+      "loss": 3.985,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533397363536724e-05,
+      "loss": 4.0168,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525011416026204e-05,
+      "loss": 3.9791,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516625468515684e-05,
+      "loss": 4.0179,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508239521005164e-05,
+      "loss": 4.0004,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4499853573494644e-05,
+      "loss": 3.9828,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449146762598412e-05,
+      "loss": 3.9991,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44830816784736e-05,
+      "loss": 3.9963,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447471210976681e-05,
+      "loss": 3.9775,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446634254106003e-05,
+      "loss": 4.0002,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445795659354951e-05,
+      "loss": 4.0047,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444957064603899e-05,
+      "loss": 3.9897,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444118469852847e-05,
+      "loss": 3.9995,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443281512982168e-05,
+      "loss": 3.9765,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442442918231116e-05,
+      "loss": 3.9913,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441604323480064e-05,
+      "loss": 3.99,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440765728729012e-05,
+      "loss": 3.9975,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439927133977959e-05,
+      "loss": 3.9912,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439088539226907e-05,
+      "loss": 3.997,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438249944475855e-05,
+      "loss": 4.0099,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437412987605177e-05,
+      "loss": 3.9968,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436574392854125e-05,
+      "loss": 3.9976,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435735798103073e-05,
+      "loss": 3.9947,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434897203352021e-05,
+      "loss": 3.9907,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434058608600969e-05,
+      "loss": 4.0015,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433220013849917e-05,
+      "loss": 3.9894,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432381419098865e-05,
+      "loss": 3.9969,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315444622281856e-05,
+      "loss": 4.001,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307058674771336e-05,
+      "loss": 3.9987,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4298672727260816e-05,
+      "loss": 3.9746,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4290286779750296e-05,
+      "loss": 3.9975,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4281917211043505e-05,
+      "loss": 3.9975,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4273531263532985e-05,
+      "loss": 3.9875,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265145316022465e-05,
+      "loss": 3.9989,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425677574731568e-05,
+      "loss": 3.9891,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424838979980516e-05,
+      "loss": 3.9867,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424000385229464e-05,
+      "loss": 3.9894,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423161790478412e-05,
+      "loss": 3.9914,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42232319572736e-05,
+      "loss": 4.0001,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421484600976308e-05,
+      "loss": 3.998,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420646006225256e-05,
+      "loss": 3.9994,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419807411474204e-05,
+      "loss": 3.9897,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418968816723152e-05,
+      "loss": 3.9855,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418131859852473e-05,
+      "loss": 3.9915,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417293265101421e-05,
+      "loss": 4.0001,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416454670350369e-05,
+      "loss": 3.9994,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.415616075599317e-05,
+      "loss": 3.9949,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.414777480848265e-05,
+      "loss": 3.9946,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139405239775865e-05,
+      "loss": 3.9849,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131019292265345e-05,
+      "loss": 3.9984,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122633344754825e-05,
+      "loss": 3.9896,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114247397244305e-05,
+      "loss": 3.9906,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410586144973378e-05,
+      "loss": 3.9861,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409747550222326e-05,
+      "loss": 3.9919,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408908955471274e-05,
+      "loss": 3.9899,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408070360720222e-05,
+      "loss": 3.986,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40723176596917e-05,
+      "loss": 3.9801,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406394809098491e-05,
+      "loss": 3.9929,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405556214347439e-05,
+      "loss": 3.9977,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047176195963874e-05,
+      "loss": 3.9887,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4038790248453354e-05,
+      "loss": 3.9975,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403042067974656e-05,
+      "loss": 4.0081,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402205111103978e-05,
+      "loss": 3.9921,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401366516352925e-05,
+      "loss": 3.9899,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400527921601873e-05,
+      "loss": 3.9906,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399689326850821e-05,
+      "loss": 3.9947,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398852369980142e-05,
+      "loss": 3.9922,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39801377522909e-05,
+      "loss": 3.9901,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397175180478038e-05,
+      "loss": 3.9868,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396336585726986e-05,
+      "loss": 3.9952,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395497990975934e-05,
+      "loss": 3.9884,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3946610341052556e-05,
+      "loss": 3.995,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3938224393542036e-05,
+      "loss": 3.9921,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929838446031516e-05,
+      "loss": 3.9892,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921452498520996e-05,
+      "loss": 3.9872,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913066551010476e-05,
+      "loss": 4.0021,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904680603499956e-05,
+      "loss": 3.9966,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896311034793165e-05,
+      "loss": 3.9871,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3887925087282645e-05,
+      "loss": 3.9916,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879539139772125e-05,
+      "loss": 3.9997,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871153192261605e-05,
+      "loss": 3.9854,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3862783623554814e-05,
+      "loss": 3.9952,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854397676044294e-05,
+      "loss": 3.9932,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384602810733751e-05,
+      "loss": 4.0011,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383764215982699e-05,
+      "loss": 3.9969,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382925621231647e-05,
+      "loss": 3.9973,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382087026480595e-05,
+      "loss": 3.9898,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.381248431729543e-05,
+      "loss": 4.0007,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380409836978491e-05,
+      "loss": 3.99,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379572880107812e-05,
+      "loss": 3.9828,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.37873428535676e-05,
+      "loss": 3.991,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377895690605708e-05,
+      "loss": 3.9782,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377057095854656e-05,
+      "loss": 3.993,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376218501103604e-05,
+      "loss": 3.9883,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.038602352142334,
+      "eval_runtime": 307.4727,
+      "eval_samples_per_second": 1241.057,
+      "eval_steps_per_second": 38.784,
+      "step": 992160
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 4.0218938426440384e+17,
+  "trial_name": null,
+  "trial_params": null
+}