diff --git "a/checkpoint-1373760/trainer_state.json" "b/checkpoint-1373760/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1373760/trainer_state.json"
@@ -0,0 +1,16267 @@
+{
+  "best_metric": 4.008997440338135,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/passive/lstm/2/checkpoints/checkpoint-1373760",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 1373760,
+  "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": 10.8206,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5517,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0492,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9882,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.943,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8668,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7139,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.615,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5194,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992455923001279e-05,
+      "loss": 6.4393,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991617328250227e-05,
+      "loss": 6.3772,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990778733499175e-05,
+      "loss": 6.3056,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989940138748123e-05,
+      "loss": 6.2402,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989101543997071e-05,
+      "loss": 6.1777,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988262949246018e-05,
+      "loss": 6.1283,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987424354494966e-05,
+      "loss": 6.0656,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986585759743914e-05,
+      "loss": 6.0332,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985747164992862e-05,
+      "loss": 5.9864,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98490857024181e-05,
+      "loss": 5.9467,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984069975490758e-05,
+      "loss": 5.9056,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.8737,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8322,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815558291179756e-05,
+      "loss": 5.8041,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807172343669236e-05,
+      "loss": 5.7614,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798786396158716e-05,
+      "loss": 5.7496,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 5.7106,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 5.6924,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.6523,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6354,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6148,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.5896,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.58,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5538,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5377,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5294,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.506,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9698171404836204e-05,
+      "loss": 5.4977,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9689785457325684e-05,
+      "loss": 5.4519,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9681399509815164e-05,
+      "loss": 5.4578,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9673013562304644e-05,
+      "loss": 5.4188,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9664627614794124e-05,
+      "loss": 5.4256,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4043,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4025,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.3707,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.3789,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962271425604525e-05,
+      "loss": 5.3669,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3595,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3521,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3231,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3145,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.3136,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957241494978587e-05,
+      "loss": 5.3003,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956402900227535e-05,
+      "loss": 5.2938,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955564305476483e-05,
+      "loss": 5.271,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2684,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2476,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2694,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2219,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2387,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.228,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2077,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2068,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.1939,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.1821,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1751,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.189,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1628,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1535,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.144,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1415,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1472,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1426,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1174,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1341,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1269,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.0984,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.0971,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.0869,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.0785,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0734,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0779,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0732,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.0751,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0513,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.929582609117229e-05,
+      "loss": 5.0322,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.928744014366177e-05,
+      "loss": 5.0385,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.0402,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.039,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0211,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0188,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0187,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0121,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922875488989185e-05,
+      "loss": 5.0,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922036894238133e-05,
+      "loss": 4.9985,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 4.9846,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920361342616403e-05,
+      "loss": 4.9976,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919524385745724e-05,
+      "loss": 4.9831,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918685790994672e-05,
+      "loss": 4.9624,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91784719624362e-05,
+      "loss": 4.9672,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9170086014925676e-05,
+      "loss": 4.9667,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9161700067415156e-05,
+      "loss": 4.9722,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9544,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9512,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.937,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128172656176805e-05,
+      "loss": 4.9429,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119786708666285e-05,
+      "loss": 4.9218,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9317,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9198,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9228,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.911,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9157,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9017,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9064,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.8975,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.8909,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9032,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.8873,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.8853,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.874,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.8793,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.8742,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8674,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8735,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8595,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8598,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8395,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.894371456855283e-05,
+      "loss": 4.8452,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8457,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.838,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8288,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8508,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.8343,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889341526229344e-05,
+      "loss": 4.8247,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.8251,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8201,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.822,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859887851055094e-05,
+      "loss": 4.8172,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851501903544574e-05,
+      "loss": 4.8133,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843115956034054e-05,
+      "loss": 4.8159,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834730008523534e-05,
+      "loss": 4.8135,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8021,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8817958113502494e-05,
+      "loss": 4.7927,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8037,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.7904,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879286578618586e-05,
+      "loss": 4.793,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878447983867534e-05,
+      "loss": 4.7851,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877609389116482e-05,
+      "loss": 4.7825,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.787,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7761,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7721,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.7402849197387695,
+      "eval_runtime": 293.2378,
+      "eval_samples_per_second": 1301.302,
+      "eval_steps_per_second": 40.667,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.7586,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.759,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7823,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.7667,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.7598,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.756,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.7551,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.7379,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675462521038576e-05,
+      "loss": 4.7479,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7477,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865870700482127e-05,
+      "loss": 4.7495,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865032105731075e-05,
+      "loss": 4.7478,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7303,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7299,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7187,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7209,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7267,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8600021751051354e-05,
+      "loss": 4.7171,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8591635803540834e-05,
+      "loss": 4.7229,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7366,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7178,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7206,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.855810839230249e-05,
+      "loss": 4.704,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.724,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7004,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7044,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.6906,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.6998,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.850782546484683e-05,
+      "loss": 4.6908,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8499455896140044e-05,
+      "loss": 4.6874,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.6928,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.6957,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.6893,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.6908,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.68,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.6897,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6626,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6769,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6532,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6755,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6572,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.6693,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6497,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.6681,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.6645,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.6615,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6675,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6374,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6415,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.655,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6475,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.645,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6372,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.6328,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6271,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.6424,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.6192,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.6311,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.635,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247942986039366e-05,
+      "loss": 4.6181,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8239557038528846e-05,
+      "loss": 4.6194,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823120384862579e-05,
+      "loss": 4.6182,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.6053,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6096,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6276,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.609,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6016,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6028,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.6033,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6169,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.6107,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.5989,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81389748048138e-05,
+      "loss": 4.6144,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6132,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.5955,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811383334108597e-05,
+      "loss": 4.5952,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.5939,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.5882,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.5865,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.5895,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.596,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.5971,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.5855,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804676213980554e-05,
+      "loss": 4.5624,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803837619229502e-05,
+      "loss": 4.5784,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8029990244784504e-05,
+      "loss": 4.583,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802162067607771e-05,
+      "loss": 4.5864,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801323472856719e-05,
+      "loss": 4.5765,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5728,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5787,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5711,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5665,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5731,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79629354223078e-05,
+      "loss": 4.5579,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795454947479728e-05,
+      "loss": 4.576,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5685,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5468,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5608,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5584,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5696,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5567,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.5517,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5448,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.5547,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5367,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5446,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5442,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5443,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5342,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7828809398550675e-05,
+      "loss": 4.5442,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7820423451040155e-05,
+      "loss": 4.5361,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7812037503529635e-05,
+      "loss": 4.5378,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7803651556019115e-05,
+      "loss": 4.5388,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7795281987312324e-05,
+      "loss": 4.5295,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7786896039801804e-05,
+      "loss": 4.5529,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7778510092291284e-05,
+      "loss": 4.5383,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.5356,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776175457607397e-05,
+      "loss": 4.5252,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775336862856345e-05,
+      "loss": 4.5332,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774498268105293e-05,
+      "loss": 4.5307,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773659673354241e-05,
+      "loss": 4.5222,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772821078603189e-05,
+      "loss": 4.5387,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.5258,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5203,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5165,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5162,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5163,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767791147977251e-05,
+      "loss": 4.5189,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766952553226199e-05,
+      "loss": 4.5055,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766113958475147e-05,
+      "loss": 4.5329,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765277001604468e-05,
+      "loss": 4.5228,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.5119,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.511,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5062,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761924260480633e-05,
+      "loss": 4.5181,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7610856657295813e-05,
+      "loss": 4.5148,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602470709785293e-05,
+      "loss": 4.5101,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.5149,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5104,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757734562486119e-05,
+      "loss": 4.5089,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756895967735067e-05,
+      "loss": 4.4995,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756057372984015e-05,
+      "loss": 4.5121,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.501,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.504,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753541588730859e-05,
+      "loss": 4.499,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752702993979807e-05,
+      "loss": 4.5013,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751864399228755e-05,
+      "loss": 4.5006,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.4889,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.4938,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.468594074249268,
+      "eval_runtime": 293.1863,
+      "eval_samples_per_second": 1301.531,
+      "eval_steps_per_second": 40.674,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.4813,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.4803,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747673063353868e-05,
+      "loss": 4.5069,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.4979,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.4866,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745157279100712e-05,
+      "loss": 4.486,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74431868434966e-05,
+      "loss": 4.488,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.4733,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.4907,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.4829,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.4869,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401257105944e-05,
+      "loss": 4.4882,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392887537237216e-05,
+      "loss": 4.4709,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384501589726696e-05,
+      "loss": 4.4747,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376115642216176e-05,
+      "loss": 4.4685,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.4692,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.473,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.4655,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.4732,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.4934,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.4702,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317430388446264e-05,
+      "loss": 4.4742,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730904444093574e-05,
+      "loss": 4.468,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730065849342522e-05,
+      "loss": 4.4812,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292272545914704e-05,
+      "loss": 4.4606,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283886598404184e-05,
+      "loss": 4.4703,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275500650893664e-05,
+      "loss": 4.4543,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4588,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4567,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725035918716583e-05,
+      "loss": 4.4534,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724197323965531e-05,
+      "loss": 4.4594,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723358729214479e-05,
+      "loss": 4.4657,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722520134463427e-05,
+      "loss": 4.4593,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721683177592748e-05,
+      "loss": 4.4649,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4488,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.4673,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4346,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.4572,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717491841717862e-05,
+      "loss": 4.4322,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71665324696681e-05,
+      "loss": 4.4539,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715814652215758e-05,
+      "loss": 4.4405,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714976057464706e-05,
+      "loss": 4.4541,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714139100594027e-05,
+      "loss": 4.4352,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133021437233476e-05,
+      "loss": 4.4473,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124635489722956e-05,
+      "loss": 4.4495,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116249542212436e-05,
+      "loss": 4.4501,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107863594701916e-05,
+      "loss": 4.4512,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099477647191396e-05,
+      "loss": 4.4295,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091091699680876e-05,
+      "loss": 4.4266,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082705752170356e-05,
+      "loss": 4.4488,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074319804659836e-05,
+      "loss": 4.4405,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4344,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.4331,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704917834093201e-05,
+      "loss": 4.429,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.4185,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70324228247147e-05,
+      "loss": 4.4373,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702405325600791e-05,
+      "loss": 4.4161,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701566730849739e-05,
+      "loss": 4.4264,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700728136098687e-05,
+      "loss": 4.4377,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699889541347635e-05,
+      "loss": 4.4191,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.699050946596583e-05,
+      "loss": 4.4202,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.698212351845531e-05,
+      "loss": 4.4228,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697373757094479e-05,
+      "loss": 4.4064,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965351623434276e-05,
+      "loss": 4.41,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956982054727485e-05,
+      "loss": 4.4309,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948596107216965e-05,
+      "loss": 4.4157,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940210159706445e-05,
+      "loss": 4.4091,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931824212195925e-05,
+      "loss": 4.4105,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69234382646854e-05,
+      "loss": 4.4147,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691505231717488e-05,
+      "loss": 4.4253,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690666636966436e-05,
+      "loss": 4.4222,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689828042215384e-05,
+      "loss": 4.4104,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4255,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4213,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4096,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4071,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856367063404974e-05,
+      "loss": 4.4138,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.406,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839595168383934e-05,
+      "loss": 4.4043,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831209220873414e-05,
+      "loss": 4.406,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4132,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.4174,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.4032,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.389,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.3971,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.4031,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677254034590723e-05,
+      "loss": 4.4097,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.3995,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.3979,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.4082,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.3933,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673062698715837e-05,
+      "loss": 4.3946,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672224103964785e-05,
+      "loss": 4.3985,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.3899,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670546914462681e-05,
+      "loss": 4.4069,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669708319711629e-05,
+      "loss": 4.3936,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668869724960576e-05,
+      "loss": 4.3809,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.3924,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6671941733388456e-05,
+      "loss": 4.3895,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.4059,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655186217171145e-05,
+      "loss": 4.3932,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664680026966063e-05,
+      "loss": 4.3851,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663841432215011e-05,
+      "loss": 4.3777,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.3908,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.3738,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.3838,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.3826,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.3815,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588098637086985e-05,
+      "loss": 4.3703,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.3896,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571326742065945e-05,
+      "loss": 4.3745,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562940794555425e-05,
+      "loss": 4.3746,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.3826,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546185278338114e-05,
+      "loss": 4.3708,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.3925,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.3814,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.3816,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.3709,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650427191958925e-05,
+      "loss": 4.3764,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649590235088246e-05,
+      "loss": 4.3765,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648751640337194e-05,
+      "loss": 4.3663,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647913045586142e-05,
+      "loss": 4.3814,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64707445083509e-05,
+      "loss": 4.3756,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646235856084038e-05,
+      "loss": 4.3677,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645397261332986e-05,
+      "loss": 4.3699,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.3683,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.3599,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428831149602034e-05,
+      "loss": 4.3733,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.3581,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.3795,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.3767,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.3633,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638693416965689e-05,
+      "loss": 4.3616,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637854822214637e-05,
+      "loss": 4.362,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637016227463585e-05,
+      "loss": 4.3713,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636177632712533e-05,
+      "loss": 4.3675,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635339037961481e-05,
+      "loss": 4.3657,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.3682,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.3681,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632824891588699e-05,
+      "loss": 4.3675,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.3531,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.3731,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630310745215916e-05,
+      "loss": 4.3616,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6294737883452366e-05,
+      "loss": 4.3583,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6286351935941846e-05,
+      "loss": 4.3602,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6277965988431326e-05,
+      "loss": 4.3614,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6269580040920806e-05,
+      "loss": 4.3575,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6261194093410286e-05,
+      "loss": 4.3473,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6252808145899766e-05,
+      "loss": 4.3497,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.337387561798096,
+      "eval_runtime": 294.5321,
+      "eval_samples_per_second": 1295.584,
+      "eval_steps_per_second": 40.488,
+      "step": 228960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6244422198389246e-05,
+      "loss": 4.3456,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6236036250878726e-05,
+      "loss": 4.3415,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6227650303368205e-05,
+      "loss": 4.3677,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6219264355857685e-05,
+      "loss": 4.3592,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6210878408347165e-05,
+      "loss": 4.3566,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3438,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.3543,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6185720565815605e-05,
+      "loss": 4.3369,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177334618305085e-05,
+      "loss": 4.3582,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6168948670794565e-05,
+      "loss": 4.3503,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6160562723284045e-05,
+      "loss": 4.3497,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6152176775773525e-05,
+      "loss": 4.3589,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6143807207066734e-05,
+      "loss": 4.337,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6135421259556214e-05,
+      "loss": 4.3454,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6127035312045694e-05,
+      "loss": 4.334,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6118649364535174e-05,
+      "loss": 4.3405,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6110263417024654e-05,
+      "loss": 4.3385,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.610187746951414e-05,
+      "loss": 4.3373,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093491522003614e-05,
+      "loss": 4.3406,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3606,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3408,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3454,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605996411076526e-05,
+      "loss": 4.3405,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.605159454205848e-05,
+      "loss": 4.3532,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.604320859454796e-05,
+      "loss": 4.3331,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.603482264703743e-05,
+      "loss": 4.3407,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.602643669952691e-05,
+      "loss": 4.3299,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601805075201639e-05,
+      "loss": 4.33,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600966480450588e-05,
+      "loss": 4.3291,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600127885699536e-05,
+      "loss": 4.3346,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599289290948484e-05,
+      "loss": 4.3326,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598450696197432e-05,
+      "loss": 4.3404,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3351,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.337,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.331,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3413,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5942609982029176e-05,
+      "loss": 4.3191,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5934224034518656e-05,
+      "loss": 4.332,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5925838087008136e-05,
+      "loss": 4.3097,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5917452139497616e-05,
+      "loss": 4.329,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5909066191987096e-05,
+      "loss": 4.3225,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5900680244476576e-05,
+      "loss": 4.3361,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589229429696606e-05,
+      "loss": 4.3157,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588390834945554e-05,
+      "loss": 4.3273,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587555515955248e-05,
+      "loss": 4.3298,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586716921204196e-05,
+      "loss": 4.3327,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3274,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3155,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584202774831413e-05,
+      "loss": 4.3055,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.583364180080361e-05,
+      "loss": 4.3294,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.582525585329309e-05,
+      "loss": 4.3249,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.581686990578257e-05,
+      "loss": 4.3175,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580848395827205e-05,
+      "loss": 4.314,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580009801076153e-05,
+      "loss": 4.3152,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5791712063251017e-05,
+      "loss": 4.2995,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783342494544226e-05,
+      "loss": 4.3246,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5774956547033706e-05,
+      "loss": 4.2991,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5766570599523186e-05,
+      "loss": 4.3125,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5758184652012666e-05,
+      "loss": 4.3201,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5749798704502146e-05,
+      "loss": 4.3085,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5741429135795355e-05,
+      "loss": 4.3036,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5733043188284835e-05,
+      "loss": 4.3106,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5724657240774315e-05,
+      "loss": 4.2951,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5716271293263794e-05,
+      "loss": 4.3025,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570788534575327e-05,
+      "loss": 4.314,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5699515777046484e-05,
+      "loss": 4.3045,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5691129829535963e-05,
+      "loss": 4.2974,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5682743882025443e-05,
+      "loss": 4.2966,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674357934514923e-05,
+      "loss": 4.3046,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56659719870044e-05,
+      "loss": 4.3094,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.565758603949388e-05,
+      "loss": 4.3111,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.3035,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.564083052327657e-05,
+      "loss": 4.3132,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.563244457576605e-05,
+      "loss": 4.3137,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.562405862825553e-05,
+      "loss": 4.2984,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.2969,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560730311203822e-05,
+      "loss": 4.3066,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559891716452771e-05,
+      "loss": 4.3015,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559053121701719e-05,
+      "loss": 4.2992,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55821616483104e-05,
+      "loss": 4.2977,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.3051,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3084,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.2993,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.2852,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5540248289561526e-05,
+      "loss": 4.287,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5531862342051006e-05,
+      "loss": 4.2956,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5523476394540486e-05,
+      "loss": 4.3041,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5515090447029966e-05,
+      "loss": 4.2966,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5506720878323175e-05,
+      "loss": 4.2875,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5498334930812655e-05,
+      "loss": 4.3015,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548994898330214e-05,
+      "loss": 4.2926,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548156303579162e-05,
+      "loss": 4.2889,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54731770882811e-05,
+      "loss": 4.2903,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.546480751957431e-05,
+      "loss": 4.295,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.545642157206379e-05,
+      "loss": 4.3046,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.544803562455327e-05,
+      "loss": 4.2842,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.2767,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.2932,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.2885,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.3016,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54061222658044e-05,
+      "loss": 4.2949,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.539773631829388e-05,
+      "loss": 4.2814,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538935037078336e-05,
+      "loss": 4.2765,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.538096442327284e-05,
+      "loss": 4.2919,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372578475762326e-05,
+      "loss": 4.2723,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5364192528251806e-05,
+      "loss": 4.2831,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535580658074128e-05,
+      "loss": 4.2785,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5347437012034495e-05,
+      "loss": 4.2865,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5339051064523975e-05,
+      "loss": 4.2703,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5330665117013455e-05,
+      "loss": 4.2896,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5322295548306664e-05,
+      "loss": 4.2745,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5313909600796144e-05,
+      "loss": 4.2729,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5305523653285624e-05,
+      "loss": 4.2861,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5297137705775104e-05,
+      "loss": 4.2715,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528875175826458e-05,
+      "loss": 4.2964,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5280365810754064e-05,
+      "loss": 4.2857,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5271979863243544e-05,
+      "loss": 4.282,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5263593915733024e-05,
+      "loss": 4.2743,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5255207968222504e-05,
+      "loss": 4.2772,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5246822020711984e-05,
+      "loss": 4.2796,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238436073201464e-05,
+      "loss": 4.2711,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5230050125690944e-05,
+      "loss": 4.285,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2775,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.272,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.520492504076684e-05,
+      "loss": 4.2733,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.519653909325632e-05,
+      "loss": 4.2738,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.267,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.2818,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51713976295285e-05,
+      "loss": 4.2616,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516301168201798e-05,
+      "loss": 4.284,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515462573450746e-05,
+      "loss": 4.2771,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514623978699694e-05,
+      "loss": 4.2774,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5137870218290147e-05,
+      "loss": 4.2674,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5129484270779626e-05,
+      "loss": 4.2707,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5121098323269106e-05,
+      "loss": 4.2796,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5112712375758586e-05,
+      "loss": 4.2771,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5104326428248066e-05,
+      "loss": 4.2686,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5095940480737546e-05,
+      "loss": 4.2748,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5087554533227026e-05,
+      "loss": 4.2728,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5079168585716506e-05,
+      "loss": 4.2783,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5070782638205986e-05,
+      "loss": 4.2637,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5062396690695466e-05,
+      "loss": 4.282,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5054010743184946e-05,
+      "loss": 4.2726,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.2688,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503725522696764e-05,
+      "loss": 4.2677,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5028869279457115e-05,
+      "loss": 4.2746,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.502049971075033e-05,
+      "loss": 4.2644,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.501211376323981e-05,
+      "loss": 4.2543,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.500372781572929e-05,
+      "loss": 4.2699,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.256018161773682,
+      "eval_runtime": 291.5284,
+      "eval_samples_per_second": 1308.933,
+      "eval_steps_per_second": 40.905,
+      "step": 305280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4995341868218764e-05,
+      "loss": 4.2624,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4986955920708244e-05,
+      "loss": 4.2517,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978569973197724e-05,
+      "loss": 4.2775,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.497020040449093e-05,
+      "loss": 4.2703,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.496181445698042e-05,
+      "loss": 4.269,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49534285094699e-05,
+      "loss": 4.2614,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2643,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.493667299325259e-05,
+      "loss": 4.245,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492828704574207e-05,
+      "loss": 4.2699,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491990109823155e-05,
+      "loss": 4.265,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491151515072103e-05,
+      "loss": 4.2611,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490312920321051e-05,
+      "loss": 4.2689,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489475963450372e-05,
+      "loss": 4.2513,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.48863736869932e-05,
+      "loss": 4.258,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487798773948268e-05,
+      "loss": 4.2515,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486960179197216e-05,
+      "loss": 4.2492,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486121584446164e-05,
+      "loss": 4.2536,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485282989695112e-05,
+      "loss": 4.252,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4844443949440604e-05,
+      "loss": 4.2547,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4836058001930084e-05,
+      "loss": 4.2733,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.482768843322329e-05,
+      "loss": 4.2594,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481930248571277e-05,
+      "loss": 4.2636,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481091653820225e-05,
+      "loss": 4.2544,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.480254696949546e-05,
+      "loss": 4.2641,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479416102198494e-05,
+      "loss": 4.2509,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.478577507447442e-05,
+      "loss": 4.2562,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.47773891269639e-05,
+      "loss": 4.2499,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476900317945338e-05,
+      "loss": 4.2432,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476063361074659e-05,
+      "loss": 4.2428,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475224766323607e-05,
+      "loss": 4.2518,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.474386171572556e-05,
+      "loss": 4.2502,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473547576821504e-05,
+      "loss": 4.2576,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472708982070452e-05,
+      "loss": 4.2546,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4718703873194e-05,
+      "loss": 4.2558,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471033430448721e-05,
+      "loss": 4.2459,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470194835697669e-05,
+      "loss": 4.2598,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4693578788269896e-05,
+      "loss": 4.2405,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4685192840759376e-05,
+      "loss": 4.2466,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4676806893248856e-05,
+      "loss": 4.2316,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4668420945738336e-05,
+      "loss": 4.2477,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4660034998227816e-05,
+      "loss": 4.2439,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4651649050717296e-05,
+      "loss": 4.2523,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4643263103206776e-05,
+      "loss": 4.2367,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4634877155696256e-05,
+      "loss": 4.2473,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626491208185736e-05,
+      "loss": 4.2493,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618105260675215e-05,
+      "loss": 4.2517,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609719313164695e-05,
+      "loss": 4.2482,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4601349744457905e-05,
+      "loss": 4.2353,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4592963796947384e-05,
+      "loss": 4.2249,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584577849436864e-05,
+      "loss": 4.2531,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576191901926344e-05,
+      "loss": 4.2457,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4567805954415824e-05,
+      "loss": 4.2382,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559420006905304e-05,
+      "loss": 4.2378,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551034059394784e-05,
+      "loss": 4.2371,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4542648111884264e-05,
+      "loss": 4.2224,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453427854317748e-05,
+      "loss": 4.2402,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452589259566696e-05,
+      "loss": 4.2252,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451750664815644e-05,
+      "loss": 4.2323,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450912070064592e-05,
+      "loss": 4.2451,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450075113193913e-05,
+      "loss": 4.2303,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449236518442861e-05,
+      "loss": 4.2259,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448397923691809e-05,
+      "loss": 4.2393,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447559328940757e-05,
+      "loss": 4.2169,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446720734189705e-05,
+      "loss": 4.2264,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445882139438653e-05,
+      "loss": 4.2354,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445043544687601e-05,
+      "loss": 4.2309,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.444204949936548e-05,
+      "loss": 4.2222,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.443366355185496e-05,
+      "loss": 4.2231,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.442529398314818e-05,
+      "loss": 4.2249,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441690803563766e-05,
+      "loss": 4.2335,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440852208812714e-05,
+      "loss": 4.2343,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440013614061662e-05,
+      "loss": 4.2284,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4391766571909834e-05,
+      "loss": 4.238,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4383380624399314e-05,
+      "loss": 4.2404,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.437499467688879e-05,
+      "loss": 4.225,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2183,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435825553947521e-05,
+      "loss": 4.237,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434986959196469e-05,
+      "loss": 4.2296,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.434148364445417e-05,
+      "loss": 4.2204,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.433309769694365e-05,
+      "loss": 4.2254,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432471174943313e-05,
+      "loss": 4.2303,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431634218072635e-05,
+      "loss": 4.2391,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430795623321583e-05,
+      "loss": 4.2207,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429957028570531e-05,
+      "loss": 4.2117,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429118433819479e-05,
+      "loss": 4.2216,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.428279839068426e-05,
+      "loss": 4.2176,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.427441244317374e-05,
+      "loss": 4.2306,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.426602649566322e-05,
+      "loss": 4.2267,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.42576405481527e-05,
+      "loss": 4.2138,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424925460064218e-05,
+      "loss": 4.2309,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424086865313166e-05,
+      "loss": 4.2201,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423248270562114e-05,
+      "loss": 4.2223,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422409675811062e-05,
+      "loss": 4.2152,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4215727189403836e-05,
+      "loss": 4.2225,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4207341241893316e-05,
+      "loss": 4.2342,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4198955294382796e-05,
+      "loss": 4.2192,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4190569346872276e-05,
+      "loss": 4.2028,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4182199778165485e-05,
+      "loss": 4.2181,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4173813830654965e-05,
+      "loss": 4.2198,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4165427883144445e-05,
+      "loss": 4.2254,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4157041935633925e-05,
+      "loss": 4.2287,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4148655988123405e-05,
+      "loss": 4.2125,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4140270040612885e-05,
+      "loss": 4.2086,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4131884093102365e-05,
+      "loss": 4.2207,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4123514524395574e-05,
+      "loss": 4.1986,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4115128576885054e-05,
+      "loss": 4.215,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4106742629374534e-05,
+      "loss": 4.2089,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409835668186402e-05,
+      "loss": 4.2173,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40899707343535e-05,
+      "loss": 4.2046,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4081584786842974e-05,
+      "loss": 4.2201,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4073198839332453e-05,
+      "loss": 4.2048,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406482927062567e-05,
+      "loss": 4.2035,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405644332311514e-05,
+      "loss": 4.2185,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404805737560462e-05,
+      "loss": 4.2076,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40396714280941e-05,
+      "loss": 4.2269,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403128548058358e-05,
+      "loss": 4.2169,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.402293229068053e-05,
+      "loss": 4.2101,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401454634317001e-05,
+      "loss": 4.2159,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400616039565949e-05,
+      "loss": 4.2031,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399777444814897e-05,
+      "loss": 4.2133,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398938850063845e-05,
+      "loss": 4.2043,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398100255312793e-05,
+      "loss": 4.2203,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397263298442114e-05,
+      "loss": 4.2063,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3964247036910616e-05,
+      "loss": 4.2041,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3955861089400096e-05,
+      "loss": 4.205,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3947475141889576e-05,
+      "loss": 4.2093,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3939089194379056e-05,
+      "loss": 4.1999,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3930703246868536e-05,
+      "loss": 4.2149,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3922317299358016e-05,
+      "loss": 4.1964,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3913931351847496e-05,
+      "loss": 4.2167,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3905545404336976e-05,
+      "loss": 4.21,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.389717583563019e-05,
+      "loss": 4.2113,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38888062669234e-05,
+      "loss": 4.2027,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388042031941288e-05,
+      "loss": 4.2042,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.387203437190236e-05,
+      "loss": 4.2159,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.386364842439184e-05,
+      "loss": 4.2077,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385526247688132e-05,
+      "loss": 4.2003,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38468765293708e-05,
+      "loss": 4.2116,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383849058186028e-05,
+      "loss": 4.2067,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383010463434976e-05,
+      "loss": 4.2158,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.382171868683924e-05,
+      "loss": 4.2013,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381333273932872e-05,
+      "loss": 4.2138,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.38049467918182e-05,
+      "loss": 4.206,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379657722311141e-05,
+      "loss": 4.2031,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3788207654404625e-05,
+      "loss": 4.2048,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3779821706894105e-05,
+      "loss": 4.2052,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3771435759383585e-05,
+      "loss": 4.2049,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3763066190676794e-05,
+      "loss": 4.1903,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3754680243166274e-05,
+      "loss": 4.2042,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.200494289398193,
+      "eval_runtime": 292.0312,
+      "eval_samples_per_second": 1306.679,
+      "eval_steps_per_second": 40.835,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746294295655754e-05,
+      "loss": 4.1936,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3737908348145234e-05,
+      "loss": 4.189,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3729522400634714e-05,
+      "loss": 4.2106,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3721136453124194e-05,
+      "loss": 4.209,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3712750505613674e-05,
+      "loss": 4.2037,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3704364558103154e-05,
+      "loss": 4.1986,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.369597861059263e-05,
+      "loss": 4.2005,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3687592663082114e-05,
+      "loss": 4.1861,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3679206715571594e-05,
+      "loss": 4.2029,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3670820768061074e-05,
+      "loss": 4.2053,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3662434820550554e-05,
+      "loss": 4.1968,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3654048873040034e-05,
+      "loss": 4.2032,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364567930433324e-05,
+      "loss": 4.1919,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363729335682272e-05,
+      "loss": 4.198,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36289074093122e-05,
+      "loss": 4.188,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362052146180168e-05,
+      "loss": 4.1863,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361213551429116e-05,
+      "loss": 4.1914,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.360374956678064e-05,
+      "loss": 4.1911,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359536361927012e-05,
+      "loss": 4.1931,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35869776717596e-05,
+      "loss": 4.2056,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357860810305281e-05,
+      "loss": 4.204,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35702221555423e-05,
+      "loss": 4.2017,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356183620803178e-05,
+      "loss": 4.1911,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355345026052126e-05,
+      "loss": 4.2046,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.354508069181447e-05,
+      "loss": 4.1876,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353669474430395e-05,
+      "loss": 4.1978,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352830879679343e-05,
+      "loss": 4.1869,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351992284928291e-05,
+      "loss": 4.18,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351153690177239e-05,
+      "loss": 4.1872,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350315095426187e-05,
+      "loss": 4.1872,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349476500675134e-05,
+      "loss": 4.1906,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486395438044556e-05,
+      "loss": 4.1927,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3478009490534036e-05,
+      "loss": 4.197,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3469623543023516e-05,
+      "loss": 4.1953,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461237595512996e-05,
+      "loss": 4.1861,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3452851648002476e-05,
+      "loss": 4.1971,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3444465700491956e-05,
+      "loss": 4.1842,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436079752981436e-05,
+      "loss": 4.1825,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3427693805470916e-05,
+      "loss": 4.177,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419324236764125e-05,
+      "loss": 4.1849,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3410938289253605e-05,
+      "loss": 4.1883,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402552341743085e-05,
+      "loss": 4.1914,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394166394232565e-05,
+      "loss": 4.1791,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385796825525774e-05,
+      "loss": 4.1857,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377410878015254e-05,
+      "loss": 4.1917,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369024930504734e-05,
+      "loss": 4.1903,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360638982994214e-05,
+      "loss": 4.1925,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.335226941428743e-05,
+      "loss": 4.1772,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334388346677691e-05,
+      "loss": 4.1632,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333549751926639e-05,
+      "loss": 4.1947,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.332711157175587e-05,
+      "loss": 4.1869,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331874200304908e-05,
+      "loss": 4.1837,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331037243434229e-05,
+      "loss": 4.1821,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330198648683177e-05,
+      "loss": 4.1736,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329360053932125e-05,
+      "loss": 4.1675,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328521459181073e-05,
+      "loss": 4.1798,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327682864430021e-05,
+      "loss": 4.1681,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326844269678969e-05,
+      "loss": 4.1767,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326005674927917e-05,
+      "loss": 4.1865,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3251670801768654e-05,
+      "loss": 4.1707,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243284854258134e-05,
+      "loss": 4.1673,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3234898906747614e-05,
+      "loss": 4.184,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322652933804082e-05,
+      "loss": 4.1631,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32181433905303e-05,
+      "loss": 4.1676,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320975744301978e-05,
+      "loss": 4.1786,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320137149550926e-05,
+      "loss": 4.1762,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319300192680247e-05,
+      "loss": 4.165,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318461597929195e-05,
+      "loss": 4.1677,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317623003178143e-05,
+      "loss": 4.1656,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.316784408427091e-05,
+      "loss": 4.1739,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315945813676039e-05,
+      "loss": 4.1803,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315107218924987e-05,
+      "loss": 4.1702,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314270262054309e-05,
+      "loss": 4.1801,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313431667303257e-05,
+      "loss": 4.1832,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312593072552205e-05,
+      "loss": 4.1792,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311754477801153e-05,
+      "loss": 4.1589,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109158830501e-05,
+      "loss": 4.179,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310077288299048e-05,
+      "loss": 4.1775,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.309238693547996e-05,
+      "loss": 4.1607,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308400098796944e-05,
+      "loss": 4.1763,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307563141926265e-05,
+      "loss": 4.1712,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3067278229359595e-05,
+      "loss": 4.1831,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3058892281849075e-05,
+      "loss": 4.1698,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305050633433856e-05,
+      "loss": 4.1521,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304212038682804e-05,
+      "loss": 4.1699,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303373443931752e-05,
+      "loss": 4.1646,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025348491807e-05,
+      "loss": 4.1727,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016962544296475e-05,
+      "loss": 4.1715,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008576596785955e-05,
+      "loss": 4.1641,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300020702807917e-05,
+      "loss": 4.1715,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299182108056865e-05,
+      "loss": 4.1682,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.298345151186186e-05,
+      "loss": 4.1613,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.297506556435134e-05,
+      "loss": 4.1643,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296667961684082e-05,
+      "loss": 4.1719,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29582936693303e-05,
+      "loss": 4.1795,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294990772181978e-05,
+      "loss": 4.1625,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294152177430926e-05,
+      "loss": 4.1518,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293313582679874e-05,
+      "loss": 4.1653,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292474987928822e-05,
+      "loss": 4.1653,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29163639317777e-05,
+      "loss": 4.174,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290797798426718e-05,
+      "loss": 4.1753,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289959203675666e-05,
+      "loss": 4.1559,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289120608924614e-05,
+      "loss": 4.1567,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288285289934308e-05,
+      "loss": 4.1684,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287446695183256e-05,
+      "loss": 4.1476,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286608100432204e-05,
+      "loss": 4.1641,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285769505681152e-05,
+      "loss": 4.1508,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284932548810473e-05,
+      "loss": 4.1666,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284093954059421e-05,
+      "loss": 4.1524,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283255359308369e-05,
+      "loss": 4.1665,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282416764557317e-05,
+      "loss": 4.1587,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281578169806265e-05,
+      "loss": 4.1478,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280739575055213e-05,
+      "loss": 4.1675,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279900980304161e-05,
+      "loss": 4.1519,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279062385553109e-05,
+      "loss": 4.175,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27822542868243e-05,
+      "loss": 4.1696,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277386833931378e-05,
+      "loss": 4.1545,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276548239180326e-05,
+      "loss": 4.1679,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275709644429274e-05,
+      "loss": 4.1509,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274871049678222e-05,
+      "loss": 4.1597,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27403245492717e-05,
+      "loss": 4.1556,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.1663,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1554,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271516670674014e-05,
+      "loss": 4.1533,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270679713803336e-05,
+      "loss": 4.1582,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269841119052284e-05,
+      "loss": 4.1568,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269002524301231e-05,
+      "loss": 4.1459,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268163929550179e-05,
+      "loss": 4.164,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267325334799127e-05,
+      "loss": 4.1448,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266486740048075e-05,
+      "loss": 4.173,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265648145297023e-05,
+      "loss": 4.1516,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264811188426344e-05,
+      "loss": 4.1648,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263972593675292e-05,
+      "loss": 4.1496,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26313399892424e-05,
+      "loss": 4.157,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622954041731886e-05,
+      "loss": 4.1642,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614568094221366e-05,
+      "loss": 4.1556,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606182146710846e-05,
+      "loss": 4.1521,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597796199200326e-05,
+      "loss": 4.163,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589410251689806e-05,
+      "loss": 4.1591,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581040682983015e-05,
+      "loss": 4.1643,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2572671114276224e-05,
+      "loss": 4.1482,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564285166765704e-05,
+      "loss": 4.1633,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2555899219255184e-05,
+      "loss": 4.1566,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2547513271744664e-05,
+      "loss": 4.1566,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253914370303787e-05,
+      "loss": 4.1563,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253075775552735e-05,
+      "loss": 4.1552,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252237180801684e-05,
+      "loss": 4.1618,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251400223931005e-05,
+      "loss": 4.1337,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.250561629179953e-05,
+      "loss": 4.1597,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.1602396965026855,
+      "eval_runtime": 292.8831,
+      "eval_samples_per_second": 1302.878,
+      "eval_steps_per_second": 40.716,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249723034428901e-05,
+      "loss": 4.1501,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248884439677849e-05,
+      "loss": 4.1398,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248045844926797e-05,
+      "loss": 4.1579,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.247207250175745e-05,
+      "loss": 4.1619,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246368655424693e-05,
+      "loss": 4.1545,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245530060673641e-05,
+      "loss": 4.1491,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244691465922589e-05,
+      "loss": 4.1521,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243852871171537e-05,
+      "loss": 4.1406,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243014276420485e-05,
+      "loss": 4.1539,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242175681669432e-05,
+      "loss": 4.1544,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241337086918381e-05,
+      "loss": 4.1556,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240498492167329e-05,
+      "loss": 4.153,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23966153529665e-05,
+      "loss": 4.149,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238822940545598e-05,
+      "loss": 4.1482,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237984345794546e-05,
+      "loss": 4.1409,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237145751043494e-05,
+      "loss": 4.1363,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236307156292442e-05,
+      "loss": 4.1463,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23546856154139e-05,
+      "loss": 4.1417,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234629966790338e-05,
+      "loss": 4.1475,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233791372039286e-05,
+      "loss": 4.1589,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232952777288234e-05,
+      "loss": 4.1597,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321158204175546e-05,
+      "loss": 4.1517,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312772256665026e-05,
+      "loss": 4.1449,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304386309154506e-05,
+      "loss": 4.156,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296000361643986e-05,
+      "loss": 4.1407,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22876307929372e-05,
+      "loss": 4.1533,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227924484542668e-05,
+      "loss": 4.142,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227085889791616e-05,
+      "loss": 4.13,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226247295040564e-05,
+      "loss": 4.1432,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225408700289512e-05,
+      "loss": 4.1404,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22457010553846e-05,
+      "loss": 4.1425,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223731510787408e-05,
+      "loss": 4.1502,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222892916036356e-05,
+      "loss": 4.1491,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22205759704605e-05,
+      "loss": 4.1479,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221220640175371e-05,
+      "loss": 4.1388,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203820454243196e-05,
+      "loss": 4.156,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195434506732676e-05,
+      "loss": 4.1374,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2187048559222156e-05,
+      "loss": 4.1313,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178662611711635e-05,
+      "loss": 4.1365,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170276664201115e-05,
+      "loss": 4.1336,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2161890716690595e-05,
+      "loss": 4.1418,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2153504769180075e-05,
+      "loss": 4.1498,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2145135200473284e-05,
+      "loss": 4.1321,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2136765631766494e-05,
+      "loss": 4.139,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2128379684255973e-05,
+      "loss": 4.1484,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2119993736745453e-05,
+      "loss": 4.1488,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211160778923493e-05,
+      "loss": 4.1472,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210322184172441e-05,
+      "loss": 4.1351,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209483589421389e-05,
+      "loss": 4.1117,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208644994670338e-05,
+      "loss": 4.1539,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207806399919286e-05,
+      "loss": 4.1435,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206967805168233e-05,
+      "loss": 4.1424,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206130848297555e-05,
+      "loss": 4.1342,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.205292253546503e-05,
+      "loss": 4.1295,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204455296675824e-05,
+      "loss": 4.1234,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203616701924772e-05,
+      "loss": 4.1319,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20277810717372e-05,
+      "loss": 4.1271,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201939512422668e-05,
+      "loss": 4.131,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201100917671615e-05,
+      "loss": 4.1437,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200262322920563e-05,
+      "loss": 4.1241,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199423728169512e-05,
+      "loss": 4.1232,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19858513341846e-05,
+      "loss": 4.1392,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197746538667408e-05,
+      "loss": 4.1229,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196909581796729e-05,
+      "loss": 4.1243,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196070987045677e-05,
+      "loss": 4.1325,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195232392294625e-05,
+      "loss": 4.1327,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194393797543573e-05,
+      "loss": 4.1212,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1935568406728936e-05,
+      "loss": 4.1198,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1927182459218416e-05,
+      "loss": 4.1214,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918796511707896e-05,
+      "loss": 4.1353,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910410564197376e-05,
+      "loss": 4.132,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902040995490585e-05,
+      "loss": 4.131,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189365504798007e-05,
+      "loss": 4.132,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188526910046955e-05,
+      "loss": 4.1398,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.187688315295903e-05,
+      "loss": 4.1362,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186849720544851e-05,
+      "loss": 4.1187,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186011125793799e-05,
+      "loss": 4.132,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185172531042747e-05,
+      "loss": 4.1385,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184335574172068e-05,
+      "loss": 4.1145,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183496979421016e-05,
+      "loss": 4.1323,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182658384669964e-05,
+      "loss": 4.131,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181819789918912e-05,
+      "loss": 4.1368,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180982833048233e-05,
+      "loss": 4.1292,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180144238297181e-05,
+      "loss": 4.1115,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179305643546129e-05,
+      "loss": 4.1247,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178467048795077e-05,
+      "loss": 4.1233,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177628454044025e-05,
+      "loss": 4.1335,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767914971733465e-05,
+      "loss": 4.1249,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759529024222945e-05,
+      "loss": 4.1212,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751143076712425e-05,
+      "loss": 4.1269,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1742757129201905e-05,
+      "loss": 4.1323,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734371181691385e-05,
+      "loss": 4.1159,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1726001612984594e-05,
+      "loss": 4.122,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1717615665474074e-05,
+      "loss": 4.1321,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709229717963554e-05,
+      "loss": 4.1347,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1700843770453034e-05,
+      "loss": 4.1216,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692457822942514e-05,
+      "loss": 4.1111,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168407187543199e-05,
+      "loss": 4.1207,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16757023067252e-05,
+      "loss": 4.1196,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166731635921469e-05,
+      "loss": 4.1323,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165893041170417e-05,
+      "loss": 4.134,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165054446419364e-05,
+      "loss": 4.1176,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164217489548686e-05,
+      "loss": 4.1156,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163378894797634e-05,
+      "loss": 4.1231,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162540300046581e-05,
+      "loss": 4.1059,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161701705295529e-05,
+      "loss": 4.1231,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160863110544477e-05,
+      "loss": 4.1046,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160024515793425e-05,
+      "loss": 4.1245,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159185921042373e-05,
+      "loss": 4.1129,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158347326291321e-05,
+      "loss": 4.1226,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157510369420643e-05,
+      "loss": 4.1179,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156671774669591e-05,
+      "loss": 4.1113,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558348177989116e-05,
+      "loss": 4.1234,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1549962230478596e-05,
+      "loss": 4.1105,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1541576282968076e-05,
+      "loss": 4.1349,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1533190335457556e-05,
+      "loss": 4.1273,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1524804387947036e-05,
+      "loss": 4.1157,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1516434819240245e-05,
+      "loss": 4.126,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1508048871729725e-05,
+      "loss": 4.1095,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1499662924219205e-05,
+      "loss": 4.1194,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1491276976708685e-05,
+      "loss": 4.1179,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482891029198165e-05,
+      "loss": 4.1214,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147452146049138e-05,
+      "loss": 4.116,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146613551298086e-05,
+      "loss": 4.115,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145774956547034e-05,
+      "loss": 4.1173,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144936361795982e-05,
+      "loss": 4.1146,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14409776704493e-05,
+      "loss": 4.11,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143260810174251e-05,
+      "loss": 4.121,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.142422215423199e-05,
+      "loss": 4.1023,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141583620672147e-05,
+      "loss": 4.1308,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140745025921095e-05,
+      "loss": 4.1162,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139906431170043e-05,
+      "loss": 4.1242,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139069474299364e-05,
+      "loss": 4.105,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138232517428685e-05,
+      "loss": 4.1206,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373939226776335e-05,
+      "loss": 4.1231,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1365553279265815e-05,
+      "loss": 4.119,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357167331755295e-05,
+      "loss": 4.1094,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1348781384244775e-05,
+      "loss": 4.1226,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1340395436734254e-05,
+      "loss": 4.1176,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1332009489223734e-05,
+      "loss": 4.1268,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1323623541713214e-05,
+      "loss": 4.1076,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315253973006424e-05,
+      "loss": 4.1251,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306868025495903e-05,
+      "loss": 4.1171,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298482077985383e-05,
+      "loss": 4.1145,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129009613047486e-05,
+      "loss": 4.1165,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128171018296434e-05,
+      "loss": 4.1156,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127332423545382e-05,
+      "loss": 4.1178,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12649382879433e-05,
+      "loss": 4.1004,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125655234043278e-05,
+      "loss": 4.1213,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.128416061401367,
+      "eval_runtime": 288.7842,
+      "eval_samples_per_second": 1321.371,
+      "eval_steps_per_second": 41.294,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124816639292226e-05,
+      "loss": 4.1117,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123978044541174e-05,
+      "loss": 4.1021,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123141087670495e-05,
+      "loss": 4.1178,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122302492919443e-05,
+      "loss": 4.1207,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121463898168391e-05,
+      "loss": 4.1135,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120625303417339e-05,
+      "loss": 4.1123,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119786708666287e-05,
+      "loss": 4.1178,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118948113915235e-05,
+      "loss": 4.101,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118109519164183e-05,
+      "loss": 4.1143,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117272562293504e-05,
+      "loss": 4.1155,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116433967542452e-05,
+      "loss": 4.1138,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1155953727914e-05,
+      "loss": 4.117,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114758415920722e-05,
+      "loss": 4.1072,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11391982116967e-05,
+      "loss": 4.1079,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113081226418618e-05,
+      "loss": 4.1042,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112242631667566e-05,
+      "loss": 4.0962,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111404036916514e-05,
+      "loss": 4.1087,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1105654421654617e-05,
+      "loss": 4.1033,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1097268474144097e-05,
+      "loss": 4.1086,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1088882526633577e-05,
+      "loss": 4.1183,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1080496579123056e-05,
+      "loss": 4.1232,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1072110631612536e-05,
+      "loss": 4.1195,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1063741062905746e-05,
+      "loss": 4.104,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1055355115395225e-05,
+      "loss": 4.1136,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1046969167884705e-05,
+      "loss": 4.1054,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.103859959917792e-05,
+      "loss": 4.1128,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10302136516674e-05,
+      "loss": 4.1069,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.102182770415688e-05,
+      "loss": 4.0942,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.101344175664636e-05,
+      "loss": 4.109,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005055809135834e-05,
+      "loss": 4.099,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0996669861625314e-05,
+      "loss": 4.1056,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988283914114794e-05,
+      "loss": 4.1128,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0979897966604274e-05,
+      "loss": 4.1087,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0971512019093754e-05,
+      "loss": 4.1086,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096314245038696e-05,
+      "loss": 4.1007,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095475650287644e-05,
+      "loss": 4.1197,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094638693416966e-05,
+      "loss": 4.1061,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093800098665914e-05,
+      "loss": 4.092,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092961503914862e-05,
+      "loss": 4.1034,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09212290916381e-05,
+      "loss": 4.0941,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091284314412758e-05,
+      "loss": 4.1024,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090445719661706e-05,
+      "loss": 4.1081,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089607124910654e-05,
+      "loss": 4.1015,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088768530159602e-05,
+      "loss": 4.0979,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08792993540855e-05,
+      "loss": 4.1105,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087092978537871e-05,
+      "loss": 4.111,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086254383786819e-05,
+      "loss": 4.1116,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.085415789035767e-05,
+      "loss": 4.0957,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084577194284715e-05,
+      "loss": 4.0725,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083738599533663e-05,
+      "loss": 4.1203,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0829016426629844e-05,
+      "loss": 4.1014,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0820630479119323e-05,
+      "loss": 4.1069,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081226091041253e-05,
+      "loss": 4.0985,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080387496290201e-05,
+      "loss": 4.0911,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079548901539149e-05,
+      "loss": 4.0896,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078710306788097e-05,
+      "loss": 4.0955,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077871712037045e-05,
+      "loss": 4.0892,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077033117285993e-05,
+      "loss": 4.0923,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076194522534941e-05,
+      "loss": 4.1124,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075355927783889e-05,
+      "loss": 4.0869,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074517333032837e-05,
+      "loss": 4.0858,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0736787382817845e-05,
+      "loss": 4.1077,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0728401435307325e-05,
+      "loss": 4.0826,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072003186660054e-05,
+      "loss": 4.087,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071166229789376e-05,
+      "loss": 4.0967,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070327635038324e-05,
+      "loss": 4.0974,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069489040287272e-05,
+      "loss": 4.0854,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06865044553622e-05,
+      "loss": 4.0937,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0678134886655406e-05,
+      "loss": 4.0809,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0669748939144886e-05,
+      "loss": 4.0991,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0661362991634366e-05,
+      "loss": 4.0974,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0652977044123846e-05,
+      "loss": 4.0937,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064459109661332e-05,
+      "loss": 4.0988,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636221527906535e-05,
+      "loss": 4.1009,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627835580396015e-05,
+      "loss": 4.1065,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619449632885495e-05,
+      "loss": 4.0833,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611063685374975e-05,
+      "loss": 4.0974,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602677737864455e-05,
+      "loss": 4.1036,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594291790353935e-05,
+      "loss": 4.0772,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585905842843415e-05,
+      "loss": 4.1018,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577519895332895e-05,
+      "loss": 4.0929,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056916670542984e-05,
+      "loss": 4.1033,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056078075791932e-05,
+      "loss": 4.0972,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055239481040879e-05,
+      "loss": 4.0796,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.054400886289827e-05,
+      "loss": 4.0823,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.053562291538775e-05,
+      "loss": 4.0897,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052723696787723e-05,
+      "loss": 4.1001,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051886739917045e-05,
+      "loss": 4.0868,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051048145165993e-05,
+      "loss": 4.0887,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050209550414941e-05,
+      "loss": 4.0921,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049370955663889e-05,
+      "loss": 4.0966,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04853399879321e-05,
+      "loss": 4.0858,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047695404042158e-05,
+      "loss": 4.0869,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046858447171479e-05,
+      "loss": 4.0966,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0460198524204267e-05,
+      "loss": 4.099,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0451812576693746e-05,
+      "loss": 4.0891,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0443426629183226e-05,
+      "loss": 4.0776,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0435040681672706e-05,
+      "loss": 4.0872,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0426654734162186e-05,
+      "loss": 4.0848,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418268786651666e-05,
+      "loss": 4.1002,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040988283914115e-05,
+      "loss": 4.0929,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040151327043436e-05,
+      "loss": 4.0893,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039312732292384e-05,
+      "loss": 4.0849,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.038474137541332e-05,
+      "loss": 4.0847,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03763554279028e-05,
+      "loss": 4.0768,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.036796948039228e-05,
+      "loss": 4.0859,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035958353288176e-05,
+      "loss": 4.0752,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.035119758537124e-05,
+      "loss": 4.0914,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.034282801666445e-05,
+      "loss": 4.0821,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033444206915393e-05,
+      "loss": 4.0849,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032605612164341e-05,
+      "loss": 4.082,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031767017413289e-05,
+      "loss": 4.0764,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030928422662237e-05,
+      "loss": 4.0919,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030089827911185e-05,
+      "loss": 4.082,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029251233160133e-05,
+      "loss": 4.1032,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028412638409081e-05,
+      "loss": 4.0936,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0275756815384027e-05,
+      "loss": 4.0848,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0267387246677236e-05,
+      "loss": 4.0901,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0259001299166716e-05,
+      "loss": 4.0781,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0250615351656196e-05,
+      "loss": 4.0841,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0242229404145675e-05,
+      "loss": 4.09,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0233859835438885e-05,
+      "loss": 4.0835,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0225473887928365e-05,
+      "loss": 4.083,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217087940417845e-05,
+      "loss": 4.0838,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020871837171106e-05,
+      "loss": 4.086,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020033242420054e-05,
+      "loss": 4.0854,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.019194647669002e-05,
+      "loss": 4.0722,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01835605291795e-05,
+      "loss": 4.0885,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017517458166898e-05,
+      "loss": 4.0739,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016678863415845e-05,
+      "loss": 4.0967,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015840268664793e-05,
+      "loss": 4.0844,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015001673913741e-05,
+      "loss": 4.0866,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014163079162689e-05,
+      "loss": 4.0749,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.01332612229201e-05,
+      "loss": 4.0901,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012487527540958e-05,
+      "loss": 4.0886,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011648932789906e-05,
+      "loss": 4.0882,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010810338038854e-05,
+      "loss": 4.0814,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009971743287803e-05,
+      "loss": 4.0892,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009133148536751e-05,
+      "loss": 4.0849,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008294553785699e-05,
+      "loss": 4.0994,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00745759691502e-05,
+      "loss": 4.0749,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006619002163968e-05,
+      "loss": 4.093,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005780407412916e-05,
+      "loss": 4.0845,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004941812661864e-05,
+      "loss": 4.0838,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004104855791185e-05,
+      "loss": 4.0829,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003266261040133e-05,
+      "loss": 4.0861,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002427666289081e-05,
+      "loss": 4.0848,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0015907094184016e-05,
+      "loss": 4.0691,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0007521146673496e-05,
+      "loss": 4.0881,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.1050825119018555,
+      "eval_runtime": 296.821,
+      "eval_samples_per_second": 1285.593,
+      "eval_steps_per_second": 40.176,
+      "step": 610560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999913519916298e-05,
+      "loss": 4.0751,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999074925165246e-05,
+      "loss": 4.0707,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998236330414194e-05,
+      "loss": 4.083,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997397735663142e-05,
+      "loss": 4.0936,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.99655914091209e-05,
+      "loss": 4.085,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995720546161038e-05,
+      "loss": 4.0784,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994881951409986e-05,
+      "loss": 4.0897,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994043356658934e-05,
+      "loss": 4.0698,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9932047619078815e-05,
+      "loss": 4.0816,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9923661671568295e-05,
+      "loss": 4.0876,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9915275724057775e-05,
+      "loss": 4.0815,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9906889776547255e-05,
+      "loss": 4.0825,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9898520207840464e-05,
+      "loss": 4.0846,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9890134260329944e-05,
+      "loss": 4.0733,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988174831281943e-05,
+      "loss": 4.071,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987336236530891e-05,
+      "loss": 4.0681,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986497641779839e-05,
+      "loss": 4.076,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985659047028787e-05,
+      "loss": 4.0768,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984820452277735e-05,
+      "loss": 4.0743,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983983495407056e-05,
+      "loss": 4.0838,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983144900656004e-05,
+      "loss": 4.0919,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982306305904952e-05,
+      "loss": 4.0885,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9814677111539e-05,
+      "loss": 4.0779,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980629116402848e-05,
+      "loss": 4.0801,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979790521651796e-05,
+      "loss": 4.0816,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978951926900744e-05,
+      "loss": 4.0763,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978113332149692e-05,
+      "loss": 4.0769,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.97727473739864e-05,
+      "loss": 4.0687,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976436142647588e-05,
+      "loss": 4.0739,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975597547896536e-05,
+      "loss": 4.0694,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9747605910258575e-05,
+      "loss": 4.0723,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9739219962748055e-05,
+      "loss": 4.0827,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973083401523753e-05,
+      "loss": 4.0814,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.972244806772701e-05,
+      "loss": 4.0782,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.971406212021649e-05,
+      "loss": 4.0744,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705692551509704e-05,
+      "loss": 4.0829,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697322982802913e-05,
+      "loss": 4.076,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9688937035292393e-05,
+      "loss": 4.0651,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680551087781873e-05,
+      "loss": 4.0742,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.967216514027135e-05,
+      "loss": 4.064,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.966377919276083e-05,
+      "loss": 4.0714,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.965539324525031e-05,
+      "loss": 4.0801,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.964700729773979e-05,
+      "loss": 4.0677,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963862135022927e-05,
+      "loss": 4.072,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963025178152248e-05,
+      "loss": 4.0827,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962186583401196e-05,
+      "loss": 4.0787,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.961347988650144e-05,
+      "loss": 4.0791,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960509393899092e-05,
+      "loss": 4.0714,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95967079914804e-05,
+      "loss": 4.0451,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958832204396988e-05,
+      "loss": 4.0863,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957993609645936e-05,
+      "loss": 4.0719,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957155014894884e-05,
+      "loss": 4.0779,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956318058024205e-05,
+      "loss": 4.0728,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955479463273154e-05,
+      "loss": 4.0623,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954640868522102e-05,
+      "loss": 4.0585,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.95380227377105e-05,
+      "loss": 4.0675,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952963679019998e-05,
+      "loss": 4.0612,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952125084268946e-05,
+      "loss": 4.0601,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951288127398267e-05,
+      "loss": 4.0828,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950449532647215e-05,
+      "loss": 4.06,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949610937896163e-05,
+      "loss": 4.055,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9487723431451107e-05,
+      "loss": 4.0796,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9479337483940587e-05,
+      "loss": 4.0494,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9470967915233796e-05,
+      "loss": 4.0562,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462581967723276e-05,
+      "loss": 4.0711,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454196020212756e-05,
+      "loss": 4.0699,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445810072702235e-05,
+      "loss": 4.0543,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437424125191715e-05,
+      "loss": 4.0605,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9429038177681195e-05,
+      "loss": 4.0552,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420652230170675e-05,
+      "loss": 4.0695,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9412266282660155e-05,
+      "loss": 4.066,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9403896713953364e-05,
+      "loss": 4.0651,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9395510766442844e-05,
+      "loss": 4.0699,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9387124818932324e-05,
+      "loss": 4.0697,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9378738871421804e-05,
+      "loss": 4.0788,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9370352923911284e-05,
+      "loss": 4.054,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936199973400823e-05,
+      "loss": 4.0701,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935361378649771e-05,
+      "loss": 4.074,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934522783898719e-05,
+      "loss": 4.0469,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933684189147667e-05,
+      "loss": 4.0705,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932845594396615e-05,
+      "loss": 4.0691,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932006999645563e-05,
+      "loss": 4.0751,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931168404894511e-05,
+      "loss": 4.0677,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.930331448023832e-05,
+      "loss": 4.0556,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.92949285327278e-05,
+      "loss": 4.0488,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.928654258521728e-05,
+      "loss": 4.0615,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.927815663770676e-05,
+      "loss": 4.0688,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926977069019624e-05,
+      "loss": 4.0616,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.926138474268572e-05,
+      "loss": 4.0577,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.92529987951752e-05,
+      "loss": 4.0618,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924461284766468e-05,
+      "loss": 4.0706,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236243278957894e-05,
+      "loss": 4.0582,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9227857331447374e-05,
+      "loss": 4.0562,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219471383936854e-05,
+      "loss": 4.0661,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.921110181523006e-05,
+      "loss": 4.0719,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920271586771954e-05,
+      "loss": 4.0602,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919432992020902e-05,
+      "loss": 4.0525,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.91859439726985e-05,
+      "loss": 4.0555,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917755802518798e-05,
+      "loss": 4.059,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916917207767746e-05,
+      "loss": 4.0699,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916078613016694e-05,
+      "loss": 4.0689,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915241656146015e-05,
+      "loss": 4.0599,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914403061394963e-05,
+      "loss": 4.0563,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913564466643911e-05,
+      "loss": 4.051,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912725871892859e-05,
+      "loss": 4.056,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911888915022181e-05,
+      "loss": 4.0572,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911050320271129e-05,
+      "loss": 4.0462,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910211725520077e-05,
+      "loss": 4.0656,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909373130769025e-05,
+      "loss": 4.0548,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.908534536017973e-05,
+      "loss": 4.0559,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90769594126692e-05,
+      "loss": 4.0507,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906857346515868e-05,
+      "loss": 4.0487,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906018751764816e-05,
+      "loss": 4.0683,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905181794894137e-05,
+      "loss": 4.0494,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.904343200143085e-05,
+      "loss": 4.0773,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9035062432724065e-05,
+      "loss": 4.0649,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026676485213545e-05,
+      "loss": 4.0577,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.901830691650676e-05,
+      "loss": 4.0608,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900993734779997e-05,
+      "loss": 4.055,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.900155140028945e-05,
+      "loss": 4.0552,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899316545277893e-05,
+      "loss": 4.0628,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898477950526841e-05,
+      "loss": 4.0548,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.897639355775789e-05,
+      "loss": 4.0586,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896800761024737e-05,
+      "loss": 4.0516,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895962166273684e-05,
+      "loss": 4.0574,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895123571522632e-05,
+      "loss": 4.057,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894286614651954e-05,
+      "loss": 4.0449,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893448019900902e-05,
+      "loss": 4.0618,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89260942514985e-05,
+      "loss": 4.0473,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891770830398798e-05,
+      "loss": 4.0704,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890932235647746e-05,
+      "loss": 4.0579,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890093640896694e-05,
+      "loss": 4.0594,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889255046145642e-05,
+      "loss": 4.0483,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.88841645139459e-05,
+      "loss": 4.0671,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.887579494523911e-05,
+      "loss": 4.0573,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.886740899772859e-05,
+      "loss": 4.0604,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885902305021807e-05,
+      "loss": 4.0542,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885063710270755e-05,
+      "loss": 4.0625,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.884225115519703e-05,
+      "loss": 4.0579,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.883386520768651e-05,
+      "loss": 4.0748,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882547926017599e-05,
+      "loss": 4.0475,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.881709331266547e-05,
+      "loss": 4.0649,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880870736515495e-05,
+      "loss": 4.0589,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8800321417644434e-05,
+      "loss": 4.0605,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8791935470133914e-05,
+      "loss": 4.0498,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878354952262339e-05,
+      "loss": 4.0621,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.87751799539166e-05,
+      "loss": 4.0564,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876679400640608e-05,
+      "loss": 4.0501,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875842443769929e-05,
+      "loss": 4.058,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.085964679718018,
+      "eval_runtime": 319.8436,
+      "eval_samples_per_second": 1193.055,
+      "eval_steps_per_second": 37.284,
+      "step": 686880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.875003849018877e-05,
+      "loss": 4.0491,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874165254267825e-05,
+      "loss": 4.0426,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873326659516773e-05,
+      "loss": 4.054,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872489702646094e-05,
+      "loss": 4.0675,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871651107895042e-05,
+      "loss": 4.0607,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.87081251314399e-05,
+      "loss": 4.0496,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869973918392939e-05,
+      "loss": 4.065,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869135323641886e-05,
+      "loss": 4.0444,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868298366771208e-05,
+      "loss": 4.0528,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8674597720201557e-05,
+      "loss": 4.0639,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.866621177269103e-05,
+      "loss": 4.052,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657842203984246e-05,
+      "loss": 4.0568,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8649456256473726e-05,
+      "loss": 4.06,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8641070308963206e-05,
+      "loss": 4.0486,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.863268436145268e-05,
+      "loss": 4.0448,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.862429841394216e-05,
+      "loss": 4.0421,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.861591246643164e-05,
+      "loss": 4.0485,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8607526518921125e-05,
+      "loss": 4.0509,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599140571410605e-05,
+      "loss": 4.0481,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8590754623900085e-05,
+      "loss": 4.0602,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8582368676389565e-05,
+      "loss": 4.0649,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8573982728879045e-05,
+      "loss": 4.0627,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8565596781368525e-05,
+      "loss": 4.0546,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8557210833858005e-05,
+      "loss": 4.049,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854885764395494e-05,
+      "loss": 4.0576,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854047169644442e-05,
+      "loss": 4.0511,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.85320857489339e-05,
+      "loss": 4.0482,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852369980142338e-05,
+      "loss": 4.0478,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851531385391286e-05,
+      "loss": 4.0476,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850692790640234e-05,
+      "loss": 4.0455,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849855833769556e-05,
+      "loss": 4.0471,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849017239018504e-05,
+      "loss": 4.0539,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848178644267452e-05,
+      "loss": 4.0612,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8473400495164e-05,
+      "loss": 4.0499,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846501454765348e-05,
+      "loss": 4.0526,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.845664497894669e-05,
+      "loss": 4.0494,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.844825903143617e-05,
+      "loss": 4.0546,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843988946272938e-05,
+      "loss": 4.0403,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843150351521886e-05,
+      "loss": 4.0477,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842311756770834e-05,
+      "loss": 4.0364,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.841473162019782e-05,
+      "loss": 4.05,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84063456726873e-05,
+      "loss": 4.0486,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839795972517678e-05,
+      "loss": 4.0451,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838957377766626e-05,
+      "loss": 4.0486,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8381187830155743e-05,
+      "loss": 4.0555,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8372801882645223e-05,
+      "loss": 4.054,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836443231393843e-05,
+      "loss": 4.054,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835604636642791e-05,
+      "loss": 4.049,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834766041891739e-05,
+      "loss": 4.0184,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8339274471406866e-05,
+      "loss": 4.0586,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8330888523896346e-05,
+      "loss": 4.0483,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8322502576385826e-05,
+      "loss": 4.053,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831413300767904e-05,
+      "loss": 4.0468,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8305747060168515e-05,
+      "loss": 4.0366,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8297361112657995e-05,
+      "loss": 4.034,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828897516514748e-05,
+      "loss": 4.0407,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828058921763696e-05,
+      "loss": 4.0395,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827220327012644e-05,
+      "loss": 4.0335,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826381732261592e-05,
+      "loss": 4.0584,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825544775390913e-05,
+      "loss": 4.0395,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824706180639861e-05,
+      "loss": 4.0262,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823867585888809e-05,
+      "loss": 4.057,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823028991137757e-05,
+      "loss": 4.0279,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822190396386705e-05,
+      "loss": 4.032,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821351801635653e-05,
+      "loss": 4.0479,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.820513206884601e-05,
+      "loss": 4.0412,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.819676250013922e-05,
+      "loss": 4.0348,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81883765526287e-05,
+      "loss": 4.0355,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817999060511818e-05,
+      "loss": 4.031,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8171604657607666e-05,
+      "loss": 4.0439,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8163235088900875e-05,
+      "loss": 4.0423,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8154849141390355e-05,
+      "loss": 4.0436,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8146463193879835e-05,
+      "loss": 4.0426,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8138077246369315e-05,
+      "loss": 4.0467,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8129691298858795e-05,
+      "loss": 4.0563,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8121305351348275e-05,
+      "loss": 4.028,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8112935782641484e-05,
+      "loss": 4.0433,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8104549835130964e-05,
+      "loss": 4.0524,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8096163887620444e-05,
+      "loss": 4.0245,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8087777940109924e-05,
+      "loss": 4.0417,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807940837140313e-05,
+      "loss": 4.0463,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807102242389262e-05,
+      "loss": 4.0482,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80626364763821e-05,
+      "loss": 4.0497,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805425052887158e-05,
+      "loss": 4.026,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804586458136105e-05,
+      "loss": 4.0235,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.803747863385053e-05,
+      "loss": 4.0426,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802909268634001e-05,
+      "loss": 4.0441,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.802070673882949e-05,
+      "loss": 4.0342,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80123371701227e-05,
+      "loss": 4.0375,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800395122261218e-05,
+      "loss": 4.0348,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799556527510166e-05,
+      "loss": 4.0457,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798719570639488e-05,
+      "loss": 4.0369,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797880975888436e-05,
+      "loss": 4.034,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797042381137384e-05,
+      "loss": 4.0375,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.796203786386332e-05,
+      "loss": 4.0475,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953668295156526e-05,
+      "loss": 4.0394,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7945282347646006e-05,
+      "loss": 4.0295,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7936896400135486e-05,
+      "loss": 4.029,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79285268314287e-05,
+      "loss": 4.0358,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7920140883918175e-05,
+      "loss": 4.045,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7911754936407655e-05,
+      "loss": 4.0426,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7903368988897135e-05,
+      "loss": 4.0361,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7894983041386615e-05,
+      "loss": 4.0356,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7886597093876095e-05,
+      "loss": 4.0272,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7878211146365575e-05,
+      "loss": 4.0361,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7869825198855055e-05,
+      "loss": 4.0283,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786145563014827e-05,
+      "loss": 4.0245,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785306968263775e-05,
+      "loss": 4.0421,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.784470011393096e-05,
+      "loss": 4.0292,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.783631416642044e-05,
+      "loss": 4.0338,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.782792821890992e-05,
+      "loss": 4.0283,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.78195422713994e-05,
+      "loss": 4.0241,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.781115632388888e-05,
+      "loss": 4.0414,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.780277037637836e-05,
+      "loss": 4.0309,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.779438442886784e-05,
+      "loss": 4.0505,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.778599848135732e-05,
+      "loss": 4.0412,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.777762891265053e-05,
+      "loss": 4.0358,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.776924296514001e-05,
+      "loss": 4.039,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7760873396433224e-05,
+      "loss": 4.0326,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752487448922704e-05,
+      "loss": 4.0288,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744101501412184e-05,
+      "loss": 4.0416,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735715553901664e-05,
+      "loss": 4.0374,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7727329606391144e-05,
+      "loss": 4.0317,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7718943658880624e-05,
+      "loss": 4.0283,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7710557711370104e-05,
+      "loss": 4.0334,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7702171763859584e-05,
+      "loss": 4.0333,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7693785816349064e-05,
+      "loss": 4.0228,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768541624764227e-05,
+      "loss": 4.039,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767703030013175e-05,
+      "loss": 4.019,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766864435262123e-05,
+      "loss": 4.0524,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766027478391444e-05,
+      "loss": 4.0317,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.765188883640393e-05,
+      "loss": 4.0382,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764350288889341e-05,
+      "loss": 4.0231,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763511694138289e-05,
+      "loss": 4.0436,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.762673099387236e-05,
+      "loss": 4.0345,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.761834504636184e-05,
+      "loss": 4.0377,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760995909885132e-05,
+      "loss": 4.0332,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.76015731513408e-05,
+      "loss": 4.0344,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.759318720383028e-05,
+      "loss": 4.0359,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.758481763512349e-05,
+      "loss": 4.0507,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757643168761297e-05,
+      "loss": 4.0242,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756804574010245e-05,
+      "loss": 4.043,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755965979259193e-05,
+      "loss": 4.039,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755127384508141e-05,
+      "loss": 4.0365,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7542904276374627e-05,
+      "loss": 4.0258,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7534518328864107e-05,
+      "loss": 4.0381,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7526132381353586e-05,
+      "loss": 4.0321,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7517746433843066e-05,
+      "loss": 4.0287,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7509360486332546e-05,
+      "loss": 4.0344,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.070601463317871,
+      "eval_runtime": 289.8375,
+      "eval_samples_per_second": 1316.569,
+      "eval_steps_per_second": 41.144,
+      "step": 763200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7500974538822026e-05,
+      "loss": 4.0258,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7492588591311506e-05,
+      "loss": 4.0178,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7484202643800986e-05,
+      "loss": 4.0326,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7475816696290466e-05,
+      "loss": 4.0431,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7467430748779946e-05,
+      "loss": 4.0395,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7459044801269426e-05,
+      "loss": 4.0289,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.74506588537589e-05,
+      "loss": 4.0369,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744227290624838e-05,
+      "loss": 4.0255,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433886958737866e-05,
+      "loss": 4.0283,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425501011227346e-05,
+      "loss": 4.0402,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417115063716826e-05,
+      "loss": 4.0306,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7408729116206306e-05,
+      "loss": 4.0382,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7400343168695786e-05,
+      "loss": 4.0379,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7391957221185266e-05,
+      "loss": 4.0249,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7383571273674746e-05,
+      "loss": 4.0218,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7375185326164226e-05,
+      "loss": 4.0174,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7366815757457435e-05,
+      "loss": 4.0287,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7358429809946915e-05,
+      "loss": 4.0309,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7350043862436395e-05,
+      "loss": 4.022,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7341657914925875e-05,
+      "loss": 4.0351,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7333271967415355e-05,
+      "loss": 4.0441,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7324902398708564e-05,
+      "loss": 4.0383,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7316516451198044e-05,
+      "loss": 4.0338,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730813050368753e-05,
+      "loss": 4.0276,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729974455617701e-05,
+      "loss": 4.0342,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729135860866649e-05,
+      "loss": 4.0308,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7282972661155963e-05,
+      "loss": 4.0228,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7274586713645443e-05,
+      "loss": 4.0196,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.726620076613492e-05,
+      "loss": 4.0299,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.725783119742814e-05,
+      "loss": 4.0237,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724944524991761e-05,
+      "loss": 4.0257,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.724107568121083e-05,
+      "loss": 4.0321,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723268973370031e-05,
+      "loss": 4.0344,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722430378618979e-05,
+      "loss": 4.0345,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721591783867927e-05,
+      "loss": 4.0231,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7207548269972484e-05,
+      "loss": 4.0324,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7199162322461964e-05,
+      "loss": 4.0351,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719077637495144e-05,
+      "loss": 4.0136,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.718239042744092e-05,
+      "loss": 4.0273,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71740044799304e-05,
+      "loss": 4.0158,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716561853241988e-05,
+      "loss": 4.0276,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715723258490936e-05,
+      "loss": 4.0235,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7148863016202566e-05,
+      "loss": 4.0244,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7140477068692046e-05,
+      "loss": 4.028,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7132091121181526e-05,
+      "loss": 4.0348,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7123705173671006e-05,
+      "loss": 4.0329,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7115319226160486e-05,
+      "loss": 4.0343,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7106933278649966e-05,
+      "loss": 4.0245,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709854733113945e-05,
+      "loss": 4.0001,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709017776243266e-05,
+      "loss": 4.0349,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708179181492214e-05,
+      "loss": 4.0248,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707340586741162e-05,
+      "loss": 4.032,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706503629870483e-05,
+      "loss": 4.0242,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705665035119431e-05,
+      "loss": 4.0186,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704826440368379e-05,
+      "loss": 4.0101,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703987845617327e-05,
+      "loss": 4.0177,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703149250866275e-05,
+      "loss": 4.0218,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.702310656115223e-05,
+      "loss": 4.0067,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.701472061364171e-05,
+      "loss": 4.0369,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.700633466613119e-05,
+      "loss": 4.0197,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.699794871862067e-05,
+      "loss": 4.0063,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6989579149913886e-05,
+      "loss": 4.0323,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6981193202403366e-05,
+      "loss": 4.0081,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6972807254892846e-05,
+      "loss": 4.0111,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6964421307382326e-05,
+      "loss": 4.0253,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69560353598718e-05,
+      "loss": 4.0208,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694764941236128e-05,
+      "loss": 4.0124,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6939279843654495e-05,
+      "loss": 4.0118,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6930893896143975e-05,
+      "loss": 4.011,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692250794863345e-05,
+      "loss": 4.0188,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691412200112293e-05,
+      "loss": 4.0266,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.690573605361241e-05,
+      "loss": 4.0202,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6897366484905624e-05,
+      "loss": 4.0208,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6888980537395104e-05,
+      "loss": 4.0194,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6880594589884584e-05,
+      "loss": 4.0365,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6872208642374064e-05,
+      "loss": 4.0082,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6863822694863544e-05,
+      "loss": 4.023,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6855436747353024e-05,
+      "loss": 4.0251,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684706717864623e-05,
+      "loss": 4.0092,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683868123113571e-05,
+      "loss": 4.0193,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683029528362519e-05,
+      "loss": 4.0252,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682190933611467e-05,
+      "loss": 4.0267,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681352338860415e-05,
+      "loss": 4.0314,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680515381989736e-05,
+      "loss": 4.0011,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679676787238684e-05,
+      "loss": 4.0058,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678838192487632e-05,
+      "loss": 4.0186,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677999597736581e-05,
+      "loss": 4.0229,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677161002985529e-05,
+      "loss": 4.0109,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676322408234477e-05,
+      "loss": 4.0184,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675483813483425e-05,
+      "loss": 4.0155,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674646856612746e-05,
+      "loss": 4.0248,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673808261861694e-05,
+      "loss": 4.0157,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672969667110642e-05,
+      "loss": 4.0108,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.67213107235959e-05,
+      "loss": 4.0203,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6712941154889106e-05,
+      "loss": 4.0256,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6704555207378586e-05,
+      "loss": 4.0223,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6696169259868066e-05,
+      "loss": 4.0067,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6687783312357546e-05,
+      "loss": 4.007,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667941374365076e-05,
+      "loss": 4.0176,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667102779614024e-05,
+      "loss": 4.0237,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666264184862972e-05,
+      "loss": 4.0235,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66542559011192e-05,
+      "loss": 4.0127,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664586995360868e-05,
+      "loss": 4.016,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663748400609816e-05,
+      "loss": 4.0074,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6629098058587635e-05,
+      "loss": 4.0137,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6620712111077115e-05,
+      "loss": 4.0083,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6612326163566595e-05,
+      "loss": 4.0081,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660395659485981e-05,
+      "loss": 4.0194,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6595570647349284e-05,
+      "loss": 4.0099,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65872010786425e-05,
+      "loss": 4.0112,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657881513113198e-05,
+      "loss": 4.0071,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.657042918362146e-05,
+      "loss": 4.0052,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.656204323611094e-05,
+      "loss": 4.0188,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.655365728860042e-05,
+      "loss": 4.0138,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65452713410899e-05,
+      "loss": 4.0227,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.653688539357938e-05,
+      "loss": 4.0211,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.652849944606886e-05,
+      "loss": 4.0168,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6520146256165805e-05,
+      "loss": 4.0181,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6511760308655285e-05,
+      "loss": 4.0147,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.650337436114476e-05,
+      "loss": 4.0104,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.649498841363424e-05,
+      "loss": 4.0191,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.648660246612372e-05,
+      "loss": 4.014,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64782165186132e-05,
+      "loss": 4.0177,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6469830571102684e-05,
+      "loss": 4.0053,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6461444623592164e-05,
+      "loss": 4.0177,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6453058676081644e-05,
+      "loss": 4.0099,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.644470548617858e-05,
+      "loss": 4.0056,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.643631953866806e-05,
+      "loss": 4.0154,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642793359115754e-05,
+      "loss": 3.9971,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641954764364702e-05,
+      "loss": 4.0348,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.64111616961365e-05,
+      "loss": 4.0092,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640277574862598e-05,
+      "loss": 4.0192,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639440617991919e-05,
+      "loss": 4.0045,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638602023240867e-05,
+      "loss": 4.0211,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.637763428489815e-05,
+      "loss": 4.0136,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636924833738764e-05,
+      "loss": 4.0202,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636086238987712e-05,
+      "loss": 4.011,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635249282117033e-05,
+      "loss": 4.0122,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634410687365981e-05,
+      "loss": 4.0164,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633572092614929e-05,
+      "loss": 4.0336,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632733497863877e-05,
+      "loss": 4.0015,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.631894903112825e-05,
+      "loss": 4.0246,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310579462421456e-05,
+      "loss": 4.0225,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6302193514910936e-05,
+      "loss": 4.0136,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6293807567400416e-05,
+      "loss": 4.0071,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6285421619889896e-05,
+      "loss": 4.015,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6277035672379376e-05,
+      "loss": 4.0121,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6268649724868856e-05,
+      "loss": 4.0121,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6260263777358336e-05,
+      "loss": 4.0105,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.057331562042236,
+      "eval_runtime": 302.0736,
+      "eval_samples_per_second": 1263.239,
+      "eval_steps_per_second": 39.477,
+      "step": 839520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251877829847816e-05,
+      "loss": 4.0066,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.624350826114103e-05,
+      "loss": 4.0001,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.623512231363051e-05,
+      "loss": 4.0112,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622675274492372e-05,
+      "loss": 4.0213,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62183667974132e-05,
+      "loss": 4.0201,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620998084990268e-05,
+      "loss": 4.0113,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620159490239216e-05,
+      "loss": 4.0192,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619322533368537e-05,
+      "loss": 4.0078,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618483938617485e-05,
+      "loss": 4.005,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617645343866433e-05,
+      "loss": 4.0163,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616806749115381e-05,
+      "loss": 4.0197,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615968154364329e-05,
+      "loss": 4.0169,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6151311974936505e-05,
+      "loss": 4.0146,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6142926027425985e-05,
+      "loss": 4.0045,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6134540079915465e-05,
+      "loss": 4.0066,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6126154132404945e-05,
+      "loss": 3.9931,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6117784563698154e-05,
+      "loss": 4.0133,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6109398616187634e-05,
+      "loss": 4.01,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6101012668677114e-05,
+      "loss": 4.0012,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.609262672116659e-05,
+      "loss": 4.0143,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.608424077365607e-05,
+      "loss": 4.0244,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.607585482614555e-05,
+      "loss": 4.0182,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.606746887863503e-05,
+      "loss": 4.0144,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605908293112451e-05,
+      "loss": 4.0099,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.605071336241772e-05,
+      "loss": 4.0149,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.60423274149072e-05,
+      "loss": 4.0112,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.603395784620042e-05,
+      "loss": 4.0025,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.602557189868989e-05,
+      "loss": 4.0006,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.601718595117937e-05,
+      "loss": 4.0102,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.600880000366885e-05,
+      "loss": 4.0058,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600043043496206e-05,
+      "loss": 4.0098,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.599204448745154e-05,
+      "loss": 4.0075,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.598365853994102e-05,
+      "loss": 4.0108,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59752725924305e-05,
+      "loss": 4.0183,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596688664491998e-05,
+      "loss": 4.0031,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.59585170762132e-05,
+      "loss": 4.0134,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595013112870268e-05,
+      "loss": 4.0146,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594174518119216e-05,
+      "loss": 3.9929,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5933375612485366e-05,
+      "loss": 4.01,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5924989664974846e-05,
+      "loss": 3.9996,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591662009626806e-05,
+      "loss": 4.003,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5908234148757535e-05,
+      "loss": 4.0067,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5899848201247015e-05,
+      "loss": 4.005,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5891462253736495e-05,
+      "loss": 4.009,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5883076306225975e-05,
+      "loss": 4.0138,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5874690358715455e-05,
+      "loss": 4.0139,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5866304411204935e-05,
+      "loss": 4.0124,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5857918463694414e-05,
+      "loss": 4.0064,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58495325161839e-05,
+      "loss": 3.9804,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584114656867338e-05,
+      "loss": 4.015,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583276062116286e-05,
+      "loss": 4.0063,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582437467365234e-05,
+      "loss": 4.0143,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.581600510494555e-05,
+      "loss": 4.0027,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580761915743503e-05,
+      "loss": 3.9959,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579923320992451e-05,
+      "loss": 4.0,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579084726241399e-05,
+      "loss": 3.9925,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.578246131490347e-05,
+      "loss": 4.0046,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.577407536739295e-05,
+      "loss": 3.9833,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576568941988242e-05,
+      "loss": 4.021,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57573034723719e-05,
+      "loss": 4.0034,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574893390366512e-05,
+      "loss": 3.985,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.57405479561546e-05,
+      "loss": 4.0132,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573216200864408e-05,
+      "loss": 3.991,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5723792439937295e-05,
+      "loss": 3.9928,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5715406492426775e-05,
+      "loss": 4.0025,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570702054491625e-05,
+      "loss": 4.0005,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569863459740573e-05,
+      "loss": 3.9931,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5690265028698944e-05,
+      "loss": 3.9978,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5681879081188424e-05,
+      "loss": 3.9889,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56734931336779e-05,
+      "loss": 4.0002,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.566510718616738e-05,
+      "loss": 4.0062,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.565673761746059e-05,
+      "loss": 4.0006,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.564835166995007e-05,
+      "loss": 4.0063,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563996572243955e-05,
+      "loss": 3.9956,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563159615373277e-05,
+      "loss": 4.0221,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.562321020622225e-05,
+      "loss": 3.9879,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.561482425871172e-05,
+      "loss": 4.0076,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.56064383112012e-05,
+      "loss": 3.9976,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.559805236369068e-05,
+      "loss": 3.9981,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558966641618016e-05,
+      "loss": 3.9982,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.558128046866964e-05,
+      "loss": 4.0062,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557289452115912e-05,
+      "loss": 4.0103,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55645085736486e-05,
+      "loss": 4.0118,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555612262613808e-05,
+      "loss": 3.9853,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554773667862756e-05,
+      "loss": 3.9871,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553935073111704e-05,
+      "loss": 3.9975,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553098116241026e-05,
+      "loss": 4.004,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552259521489974e-05,
+      "loss": 3.9934,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551420926738922e-05,
+      "loss": 3.9968,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.55058233198787e-05,
+      "loss": 3.9981,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5497453751171906e-05,
+      "loss": 4.0015,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5489084182465115e-05,
+      "loss": 3.999,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5480698234954595e-05,
+      "loss": 3.9925,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5472312287444075e-05,
+      "loss": 4.0021,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5463926339933555e-05,
+      "loss": 4.0055,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5455540392423035e-05,
+      "loss": 4.0063,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5447170823716244e-05,
+      "loss": 3.9876,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5438784876205724e-05,
+      "loss": 3.9878,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543039892869521e-05,
+      "loss": 4.0006,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542201298118469e-05,
+      "loss": 4.0027,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541362703367417e-05,
+      "loss": 4.0065,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.540525746496738e-05,
+      "loss": 3.996,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539687151745686e-05,
+      "loss": 3.997,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538848556994634e-05,
+      "loss": 3.9899,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538009962243582e-05,
+      "loss": 3.9946,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537173005372903e-05,
+      "loss": 3.9924,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536334410621851e-05,
+      "loss": 3.9897,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535495815870799e-05,
+      "loss": 3.9977,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534657221119747e-05,
+      "loss": 3.9942,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.533818626368695e-05,
+      "loss": 3.9928,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532980031617643e-05,
+      "loss": 3.9891,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5321430747469644e-05,
+      "loss": 3.9867,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5313044799959124e-05,
+      "loss": 4.004,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304658852448604e-05,
+      "loss": 3.996,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296272904938084e-05,
+      "loss": 4.0042,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.528790333623129e-05,
+      "loss": 3.9984,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527951738872077e-05,
+      "loss": 4.0023,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527113144121025e-05,
+      "loss": 4.0013,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526274549369973e-05,
+      "loss": 3.9914,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5254359546189206e-05,
+      "loss": 4.0004,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524598997748242e-05,
+      "loss": 3.9985,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52376040299719e-05,
+      "loss": 3.9949,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522921808246138e-05,
+      "loss": 4.0011,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522083213495086e-05,
+      "loss": 3.9919,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521246256624408e-05,
+      "loss": 3.9962,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520407661873356e-05,
+      "loss": 3.9904,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519569067122303e-05,
+      "loss": 3.9898,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518730472371251e-05,
+      "loss": 3.9987,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517893515500573e-05,
+      "loss": 3.9853,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517054920749521e-05,
+      "loss": 4.0126,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.516216325998468e-05,
+      "loss": 3.9949,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.515377731247416e-05,
+      "loss": 3.9998,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5145407743767376e-05,
+      "loss": 3.987,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5137021796256856e-05,
+      "loss": 4.0028,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5128635848746336e-05,
+      "loss": 3.9992,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5120249901235816e-05,
+      "loss": 4.0023,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5111863953725296e-05,
+      "loss": 3.9941,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5103478006214776e-05,
+      "loss": 3.994,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5095092058704256e-05,
+      "loss": 3.9942,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5086706111193736e-05,
+      "loss": 4.0159,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5078320163683216e-05,
+      "loss": 3.9887,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5069934216172696e-05,
+      "loss": 4.0065,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5061548268662175e-05,
+      "loss": 4.0051,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5053178699955385e-05,
+      "loss": 3.9932,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5044809131248594e-05,
+      "loss": 3.9969,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5036423183738074e-05,
+      "loss": 3.9946,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5028037236227554e-05,
+      "loss": 3.9946,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501965128871704e-05,
+      "loss": 3.9939,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501126534120652e-05,
+      "loss": 3.9935,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.046631813049316,
+      "eval_runtime": 309.2483,
+      "eval_samples_per_second": 1233.931,
+      "eval_steps_per_second": 38.561,
+      "step": 915840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.5002879393696e-05,
+      "loss": 4.0069,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499449344618548e-05,
+      "loss": 3.9787,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498610749867496e-05,
+      "loss": 3.9898,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497772155116444e-05,
+      "loss": 4.0085,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496933560365392e-05,
+      "loss": 3.9993,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496094965614339e-05,
+      "loss": 3.9942,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495256370863287e-05,
+      "loss": 4.0075,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494417776112235e-05,
+      "loss": 3.9882,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493579181361183e-05,
+      "loss": 3.9865,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492740586610131e-05,
+      "loss": 4.0021,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491901991859079e-05,
+      "loss": 4.0019,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491063397108027e-05,
+      "loss": 3.9965,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490226440237349e-05,
+      "loss": 4.006,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489387845486297e-05,
+      "loss": 3.985,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488549250735245e-05,
+      "loss": 3.9858,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487710655984193e-05,
+      "loss": 3.9775,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486872061233141e-05,
+      "loss": 3.9958,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486033466482089e-05,
+      "loss": 3.9932,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.485194871731037e-05,
+      "loss": 3.9878,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484356276979985e-05,
+      "loss": 3.9903,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.483517682228933e-05,
+      "loss": 4.0146,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.482679087477881e-05,
+      "loss": 3.9948,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481840492726828e-05,
+      "loss": 3.9979,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.481001897975776e-05,
+      "loss": 3.9929,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4801665789854707e-05,
+      "loss": 4.0001,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4793279842344187e-05,
+      "loss": 3.9916,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.47849102736374e-05,
+      "loss": 3.9912,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477652432612688e-05,
+      "loss": 3.9814,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476813837861636e-05,
+      "loss": 3.9954,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475975243110584e-05,
+      "loss": 3.9844,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475136648359532e-05,
+      "loss": 3.9935,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47429805360848e-05,
+      "loss": 3.9909,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473459458857428e-05,
+      "loss": 3.9972,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726208641063755e-05,
+      "loss": 4.0,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4717822693553235e-05,
+      "loss": 3.9892,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470946950365018e-05,
+      "loss": 3.9962,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470108355613966e-05,
+      "loss": 3.9995,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469269760862914e-05,
+      "loss": 3.9738,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468431166111863e-05,
+      "loss": 3.994,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467592571360811e-05,
+      "loss": 3.9842,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466753976609758e-05,
+      "loss": 3.9845,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465915381858706e-05,
+      "loss": 3.9952,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465076787107654e-05,
+      "loss": 3.9888,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.464238192356602e-05,
+      "loss": 3.9897,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46339959760555e-05,
+      "loss": 3.9948,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.462561002854498e-05,
+      "loss": 3.9982,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461724045983819e-05,
+      "loss": 3.9976,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460885451232767e-05,
+      "loss": 3.9927,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460046856481715e-05,
+      "loss": 3.9656,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459208261730663e-05,
+      "loss": 3.9953,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458369666979611e-05,
+      "loss": 3.9927,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575310722285595e-05,
+      "loss": 3.9987,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566941153578805e-05,
+      "loss": 3.9905,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558555206068285e-05,
+      "loss": 3.9822,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550169258557764e-05,
+      "loss": 3.9766,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4541783311047244e-05,
+      "loss": 3.9793,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4533397363536724e-05,
+      "loss": 3.9858,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4525011416026204e-05,
+      "loss": 3.9676,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4516625468515684e-05,
+      "loss": 4.0055,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508239521005164e-05,
+      "loss": 3.9879,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449986995229837e-05,
+      "loss": 3.9702,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449148400478785e-05,
+      "loss": 3.9943,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448309805727733e-05,
+      "loss": 3.9751,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447471210976681e-05,
+      "loss": 3.9753,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.446634254106003e-05,
+      "loss": 3.9861,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445795659354951e-05,
+      "loss": 3.9883,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444958702484272e-05,
+      "loss": 3.972,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44412010773322e-05,
+      "loss": 3.9881,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443281512982168e-05,
+      "loss": 3.9714,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.442442918231116e-05,
+      "loss": 3.9825,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441604323480064e-05,
+      "loss": 3.9922,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440765728729012e-05,
+      "loss": 3.9874,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439928771858333e-05,
+      "loss": 3.9847,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439090177107281e-05,
+      "loss": 3.9824,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438251582356229e-05,
+      "loss": 4.0068,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4374146254855496e-05,
+      "loss": 3.9691,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436576030734498e-05,
+      "loss": 3.9936,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435737435983446e-05,
+      "loss": 3.9808,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434898841232394e-05,
+      "loss": 3.9808,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4340602464813416e-05,
+      "loss": 3.9852,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4332216517302896e-05,
+      "loss": 3.9848,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4323830569792376e-05,
+      "loss": 3.9978,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315444622281856e-05,
+      "loss": 3.9948,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4307058674771336e-05,
+      "loss": 3.9701,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4298672727260816e-05,
+      "loss": 3.9707,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4290286779750296e-05,
+      "loss": 3.9773,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4281900832239776e-05,
+      "loss": 3.9904,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4273531263532985e-05,
+      "loss": 3.9798,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4265145316022465e-05,
+      "loss": 3.9785,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425677574731568e-05,
+      "loss": 3.9902,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424838979980516e-05,
+      "loss": 3.9831,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424000385229464e-05,
+      "loss": 3.9871,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423163428358785e-05,
+      "loss": 3.9751,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422324833607733e-05,
+      "loss": 3.9853,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421486238856681e-05,
+      "loss": 3.9866,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420647644105629e-05,
+      "loss": 3.9922,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419809049354577e-05,
+      "loss": 3.9717,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418972092483898e-05,
+      "loss": 3.9726,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418133497732846e-05,
+      "loss": 3.985,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417294902981794e-05,
+      "loss": 3.9845,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416456308230742e-05,
+      "loss": 3.9953,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156177134796905e-05,
+      "loss": 3.9772,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147791187286385e-05,
+      "loss": 3.9795,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139405239775865e-05,
+      "loss": 3.9769,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4131019292265345e-05,
+      "loss": 3.9746,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122633344754825e-05,
+      "loss": 3.9777,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114263776048034e-05,
+      "loss": 3.9781,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4105877828537514e-05,
+      "loss": 3.9814,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4097491881026994e-05,
+      "loss": 3.9781,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4089105933516474e-05,
+      "loss": 3.9759,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.408071998600595e-05,
+      "loss": 3.9767,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407233403849543e-05,
+      "loss": 3.9683,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406396446978864e-05,
+      "loss": 3.9877,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405557852227812e-05,
+      "loss": 3.986,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40471925747676e-05,
+      "loss": 3.9846,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403880662725708e-05,
+      "loss": 3.9888,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403042067974656e-05,
+      "loss": 3.9869,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402205111103978e-05,
+      "loss": 3.9827,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.401366516352925e-05,
+      "loss": 3.9755,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400527921601873e-05,
+      "loss": 3.9858,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399689326850821e-05,
+      "loss": 3.9787,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398852369980142e-05,
+      "loss": 3.9801,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39801377522909e-05,
+      "loss": 3.985,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397175180478038e-05,
+      "loss": 3.9739,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396336585726986e-05,
+      "loss": 3.9822,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395497990975934e-05,
+      "loss": 3.9708,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.394659396224882e-05,
+      "loss": 3.9753,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393820801473831e-05,
+      "loss": 3.9823,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3929838446031516e-05,
+      "loss": 3.9733,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921452498520996e-05,
+      "loss": 3.9923,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3913066551010476e-05,
+      "loss": 3.9816,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3904680603499956e-05,
+      "loss": 3.9829,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3896294655989436e-05,
+      "loss": 3.9753,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3887908708478916e-05,
+      "loss": 3.9826,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3879522760968396e-05,
+      "loss": 3.9878,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3871153192261605e-05,
+      "loss": 3.9867,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3862767244751085e-05,
+      "loss": 3.9804,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3854397676044294e-05,
+      "loss": 3.9757,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3846011728533774e-05,
+      "loss": 3.9817,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383762578102326e-05,
+      "loss": 3.9998,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382923983351274e-05,
+      "loss": 3.9781,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382085388600222e-05,
+      "loss": 3.988,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38124679384917e-05,
+      "loss": 3.9903,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.380408199098118e-05,
+      "loss": 3.9813,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.379569604347066e-05,
+      "loss": 3.9816,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.378732647476387e-05,
+      "loss": 3.978,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377894052725335e-05,
+      "loss": 3.9792,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.377057095854656e-05,
+      "loss": 3.9825,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376218501103604e-05,
+      "loss": 3.9739,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.037889003753662,
+      "eval_runtime": 314.8622,
+      "eval_samples_per_second": 1211.93,
+      "eval_steps_per_second": 37.874,
+      "step": 992160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.375379906352552e-05,
+      "loss": 3.9785,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3745413116015e-05,
+      "loss": 3.9631,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.373702716850448e-05,
+      "loss": 3.9756,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372864122099396e-05,
+      "loss": 3.9961,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.372025527348344e-05,
+      "loss": 3.983,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.371186932597292e-05,
+      "loss": 3.9793,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.37034833784624e-05,
+      "loss": 3.9862,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.369509743095188e-05,
+      "loss": 3.9767,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.368672786224509e-05,
+      "loss": 3.971,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367834191473457e-05,
+      "loss": 3.9896,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366995596722405e-05,
+      "loss": 3.9801,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366157001971353e-05,
+      "loss": 3.9836,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365321682981047e-05,
+      "loss": 3.9894,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.364483088229995e-05,
+      "loss": 3.9682,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363644493478943e-05,
+      "loss": 3.9728,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362805898727891e-05,
+      "loss": 3.9628,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361967303976839e-05,
+      "loss": 3.9815,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361130347106161e-05,
+      "loss": 3.9777,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.360291752355109e-05,
+      "loss": 3.9727,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359453157604056e-05,
+      "loss": 3.9732,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358614562853004e-05,
+      "loss": 3.999,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357775968101952e-05,
+      "loss": 3.979,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3569373733509e-05,
+      "loss": 3.9837,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356098778599848e-05,
+      "loss": 3.9818,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.355260183848796e-05,
+      "loss": 3.9848,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354421589097744e-05,
+      "loss": 3.9757,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353582994346692e-05,
+      "loss": 3.9737,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35274439959564e-05,
+      "loss": 3.9724,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351905804844588e-05,
+      "loss": 3.9727,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3510704858542826e-05,
+      "loss": 3.9719,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3502318911032306e-05,
+      "loss": 3.9787,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3493932963521786e-05,
+      "loss": 3.9717,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3485547016011266e-05,
+      "loss": 3.988,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3477161068500746e-05,
+      "loss": 3.9807,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3468775120990226e-05,
+      "loss": 3.9774,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3460405552283435e-05,
+      "loss": 3.9844,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3452035983576644e-05,
+      "loss": 3.9809,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3443650036066124e-05,
+      "loss": 3.9567,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3435264088555604e-05,
+      "loss": 3.9794,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342687814104509e-05,
+      "loss": 3.9694,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341849219353457e-05,
+      "loss": 3.9695,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341010624602405e-05,
+      "loss": 3.9793,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340172029851353e-05,
+      "loss": 3.9776,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339333435100301e-05,
+      "loss": 3.9742,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.338494840349249e-05,
+      "loss": 3.9838,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337656245598197e-05,
+      "loss": 3.9808,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336817650847144e-05,
+      "loss": 3.9805,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335979056096092e-05,
+      "loss": 3.9795,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.335142099225414e-05,
+      "loss": 3.954,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.334303504474362e-05,
+      "loss": 3.9762,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.333464909723309e-05,
+      "loss": 3.9826,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.332626314972257e-05,
+      "loss": 3.9796,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.331787720221205e-05,
+      "loss": 3.9741,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330950763350527e-05,
+      "loss": 3.9678,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.330112168599475e-05,
+      "loss": 3.9631,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.329273573848423e-05,
+      "loss": 3.9616,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.328434979097371e-05,
+      "loss": 3.9769,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.327598022226692e-05,
+      "loss": 3.954,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32675942747564e-05,
+      "loss": 3.9836,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325922470604961e-05,
+      "loss": 3.9758,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.325083875853909e-05,
+      "loss": 3.9594,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3242452811028566e-05,
+      "loss": 3.9771,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3234066863518046e-05,
+      "loss": 3.9594,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322569729481126e-05,
+      "loss": 3.9638,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321731134730074e-05,
+      "loss": 3.9675,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320892539979022e-05,
+      "loss": 3.9779,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.32005394522797e-05,
+      "loss": 3.9603,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319215350476918e-05,
+      "loss": 3.9719,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318376755725866e-05,
+      "loss": 3.9554,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317538160974814e-05,
+      "loss": 3.9687,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316699566223762e-05,
+      "loss": 3.9753,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315862609353083e-05,
+      "loss": 3.9745,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315024014602031e-05,
+      "loss": 3.9681,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314185419850979e-05,
+      "loss": 3.9706,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.313346825099927e-05,
+      "loss": 3.9898,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.312509868229248e-05,
+      "loss": 3.9566,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.311671273478196e-05,
+      "loss": 3.981,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3108326787271446e-05,
+      "loss": 3.9678,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3099940839760926e-05,
+      "loss": 3.9639,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3091554892250406e-05,
+      "loss": 3.9676,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3083168944739886e-05,
+      "loss": 3.972,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3074799376033095e-05,
+      "loss": 3.9815,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3066413428522575e-05,
+      "loss": 3.9843,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3058027481012055e-05,
+      "loss": 3.9585,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3049657912305264e-05,
+      "loss": 3.9565,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3041271964794744e-05,
+      "loss": 3.9639,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3032886017284224e-05,
+      "loss": 3.9746,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3024500069773704e-05,
+      "loss": 3.9661,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.301613050106691e-05,
+      "loss": 3.9651,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30077445535564e-05,
+      "loss": 3.9745,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299935860604588e-05,
+      "loss": 3.9664,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299097265853536e-05,
+      "loss": 3.9763,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298258671102484e-05,
+      "loss": 3.9597,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.297420076351432e-05,
+      "loss": 3.973,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29658148160038e-05,
+      "loss": 3.9684,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295744524729701e-05,
+      "loss": 3.9824,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294905929978649e-05,
+      "loss": 3.9555,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294067335227597e-05,
+      "loss": 3.9588,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293228740476545e-05,
+      "loss": 3.9754,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.292390145725493e-05,
+      "loss": 3.9682,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.29155155097444e-05,
+      "loss": 3.9786,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.290712956223388e-05,
+      "loss": 3.9657,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289874361472337e-05,
+      "loss": 3.9644,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.289035766721285e-05,
+      "loss": 3.9625,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.288197171970233e-05,
+      "loss": 3.9657,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.287360215099554e-05,
+      "loss": 3.9642,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.286523258228875e-05,
+      "loss": 3.9623,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2856846634778227e-05,
+      "loss": 3.9684,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2848460687267707e-05,
+      "loss": 3.9618,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2840074739757186e-05,
+      "loss": 3.9647,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2831688792246666e-05,
+      "loss": 3.9626,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2823302844736146e-05,
+      "loss": 3.9524,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2814933276029355e-05,
+      "loss": 3.9747,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2806547328518835e-05,
+      "loss": 3.9676,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279816138100832e-05,
+      "loss": 3.9711,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27897754334978e-05,
+      "loss": 3.9769,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278138948598728e-05,
+      "loss": 3.9727,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277300353847676e-05,
+      "loss": 3.967,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276461759096624e-05,
+      "loss": 3.9615,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275623164345572e-05,
+      "loss": 3.9699,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.27478456959452e-05,
+      "loss": 3.9642,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273945974843468e-05,
+      "loss": 3.9628,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273107380092416e-05,
+      "loss": 3.972,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272268785341364e-05,
+      "loss": 3.965,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271431828470685e-05,
+      "loss": 3.9661,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270593233719633e-05,
+      "loss": 3.9603,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269754638968581e-05,
+      "loss": 3.9585,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2689193199782756e-05,
+      "loss": 3.9708,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2680807252272236e-05,
+      "loss": 3.958,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2672421304761716e-05,
+      "loss": 3.9704,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2664035357251196e-05,
+      "loss": 3.9751,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2655649409740676e-05,
+      "loss": 3.967,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2647263462230156e-05,
+      "loss": 3.9595,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2638877514719636e-05,
+      "loss": 3.9729,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2630491567209115e-05,
+      "loss": 3.9717,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2622121998502325e-05,
+      "loss": 3.9707,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2613736050991805e-05,
+      "loss": 3.9628,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2605350103481284e-05,
+      "loss": 3.9627,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2596964155970764e-05,
+      "loss": 3.9739,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258857820846024e-05,
+      "loss": 3.9852,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2580192260949724e-05,
+      "loss": 3.9607,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2571806313439204e-05,
+      "loss": 3.9727,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2563420365928684e-05,
+      "loss": 3.9762,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255505079722189e-05,
+      "loss": 3.9659,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254668122851511e-05,
+      "loss": 3.9693,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.253829528100459e-05,
+      "loss": 3.9643,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252990933349406e-05,
+      "loss": 3.9638,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252152338598354e-05,
+      "loss": 3.9763,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.251313743847302e-05,
+      "loss": 3.9532,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0302228927612305,
+      "eval_runtime": 317.25,
+      "eval_samples_per_second": 1202.808,
+      "eval_steps_per_second": 37.589,
+      "step": 1068480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.250476786976624e-05,
+      "loss": 3.9771,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.249638192225571e-05,
+      "loss": 3.9528,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.248799597474519e-05,
+      "loss": 3.9601,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247961002723468e-05,
+      "loss": 3.9806,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.247122407972416e-05,
+      "loss": 3.9734,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.246283813221364e-05,
+      "loss": 3.9638,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.245445218470312e-05,
+      "loss": 3.9706,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.24460662371926e-05,
+      "loss": 3.96,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.243768028968208e-05,
+      "loss": 3.9583,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242929434217156e-05,
+      "loss": 3.973,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.242090839466104e-05,
+      "loss": 3.9696,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.241252244715052e-05,
+      "loss": 3.9713,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.240415287844373e-05,
+      "loss": 3.9764,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.239576693093321e-05,
+      "loss": 3.956,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.238738098342269e-05,
+      "loss": 3.9556,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.237899503591217e-05,
+      "loss": 3.9504,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2370609088401647e-05,
+      "loss": 3.9685,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236223951969486e-05,
+      "loss": 3.9618,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235385357218434e-05,
+      "loss": 3.9618,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234546762467382e-05,
+      "loss": 3.9642,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.23370816771633e-05,
+      "loss": 3.9819,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2328695729652775e-05,
+      "loss": 3.9712,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2320309782142255e-05,
+      "loss": 3.9638,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2311923834631735e-05,
+      "loss": 3.9686,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2303537887121215e-05,
+      "loss": 3.9722,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2295151939610695e-05,
+      "loss": 3.9648,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2286782370903904e-05,
+      "loss": 3.9577,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227841280219712e-05,
+      "loss": 3.9586,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22700268546866e-05,
+      "loss": 3.9568,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226164090717608e-05,
+      "loss": 3.9608,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225325495966556e-05,
+      "loss": 3.9654,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224486901215504e-05,
+      "loss": 3.9629,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223648306464452e-05,
+      "loss": 3.9678,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2228097117134e-05,
+      "loss": 3.9675,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221971116962348e-05,
+      "loss": 3.9668,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.221134160091669e-05,
+      "loss": 3.97,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.220295565340617e-05,
+      "loss": 3.9714,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.219456970589565e-05,
+      "loss": 3.9443,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218618375838513e-05,
+      "loss": 3.9705,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217781418967834e-05,
+      "loss": 3.949,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216942824216782e-05,
+      "loss": 3.9598,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21610422946573e-05,
+      "loss": 3.9624,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.215265634714678e-05,
+      "loss": 3.9697,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2144270399636265e-05,
+      "loss": 3.9593,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2135900830929474e-05,
+      "loss": 3.9664,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2127514883418954e-05,
+      "loss": 3.9732,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2119128935908434e-05,
+      "loss": 3.9644,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2110742988397914e-05,
+      "loss": 3.9691,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.210237341969112e-05,
+      "loss": 3.9389,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20939874721806e-05,
+      "loss": 3.9605,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.208560152467008e-05,
+      "loss": 3.9705,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207721557715956e-05,
+      "loss": 3.9673,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206884600845277e-05,
+      "loss": 3.962,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206046006094225e-05,
+      "loss": 3.9512,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205207411343173e-05,
+      "loss": 3.9559,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.204368816592122e-05,
+      "loss": 3.9424,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20353022184107e-05,
+      "loss": 3.9686,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202691627090018e-05,
+      "loss": 3.9404,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201853032338966e-05,
+      "loss": 3.9679,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2010177133486596e-05,
+      "loss": 3.9631,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2001791185976076e-05,
+      "loss": 3.9486,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1993405238465556e-05,
+      "loss": 3.9645,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1985019290955036e-05,
+      "loss": 3.9463,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1976633343444516e-05,
+      "loss": 3.9498,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1968247395933996e-05,
+      "loss": 3.9521,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1959861448423476e-05,
+      "loss": 3.9629,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1951475500912956e-05,
+      "loss": 3.946,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1943089553402436e-05,
+      "loss": 3.9562,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1934703605891916e-05,
+      "loss": 3.943,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192633403718513e-05,
+      "loss": 3.9516,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191794808967461e-05,
+      "loss": 3.9651,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1909562142164085e-05,
+      "loss": 3.9628,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1901176194653565e-05,
+      "loss": 3.9544,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189280662594678e-05,
+      "loss": 3.9605,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.188443705723999e-05,
+      "loss": 3.9724,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187605110972947e-05,
+      "loss": 3.9479,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.186766516221895e-05,
+      "loss": 3.9648,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185927921470843e-05,
+      "loss": 3.9542,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185089326719791e-05,
+      "loss": 3.9548,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184250731968739e-05,
+      "loss": 3.9531,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183412137217687e-05,
+      "loss": 3.9577,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182573542466635e-05,
+      "loss": 3.9707,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181734947715583e-05,
+      "loss": 3.967,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180896352964531e-05,
+      "loss": 3.9551,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180057758213479e-05,
+      "loss": 3.9383,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.179219163462427e-05,
+      "loss": 3.9493,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.178382206591748e-05,
+      "loss": 3.9615,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.177545249721069e-05,
+      "loss": 3.9535,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.176706654970017e-05,
+      "loss": 3.9525,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.175868060218965e-05,
+      "loss": 3.9629,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1750311033482863e-05,
+      "loss": 3.9546,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174192508597234e-05,
+      "loss": 3.959,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.173353913846182e-05,
+      "loss": 3.9487,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.17251531909513e-05,
+      "loss": 3.9611,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.171676724344078e-05,
+      "loss": 3.9518,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170838129593026e-05,
+      "loss": 3.9695,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170001172722347e-05,
+      "loss": 3.9464,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.169162577971295e-05,
+      "loss": 3.9457,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.168323983220243e-05,
+      "loss": 3.9633,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.167485388469191e-05,
+      "loss": 3.9554,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166648431598512e-05,
+      "loss": 3.9641,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16580983684746e-05,
+      "loss": 3.9535,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164971242096408e-05,
+      "loss": 3.953,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164132647345356e-05,
+      "loss": 3.9492,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163294052594304e-05,
+      "loss": 3.9532,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162455457843253e-05,
+      "loss": 3.946,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161616863092201e-05,
+      "loss": 3.952,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.160778268341149e-05,
+      "loss": 3.9552,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159939673590097e-05,
+      "loss": 3.9486,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159102716719418e-05,
+      "loss": 3.9481,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158264121968366e-05,
+      "loss": 3.9558,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157425527217314e-05,
+      "loss": 3.9389,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.156586932466262e-05,
+      "loss": 3.9597,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.155748337715209e-05,
+      "loss": 3.9535,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154909742964157e-05,
+      "loss": 3.9592,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154071148213105e-05,
+      "loss": 3.9617,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153232553462053e-05,
+      "loss": 3.9604,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152397234471748e-05,
+      "loss": 3.9541,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151558639720696e-05,
+      "loss": 3.9517,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150720044969644e-05,
+      "loss": 3.9549,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149881450218592e-05,
+      "loss": 3.953,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1490428554675395e-05,
+      "loss": 3.9458,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1482042607164874e-05,
+      "loss": 3.9628,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1473656659654354e-05,
+      "loss": 3.9483,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1465270712143834e-05,
+      "loss": 3.9569,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1456884764633314e-05,
+      "loss": 3.9458,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1448498817122794e-05,
+      "loss": 3.9444,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1440129248416003e-05,
+      "loss": 3.9611,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143174330090548e-05,
+      "loss": 3.9493,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142335735339496e-05,
+      "loss": 3.9531,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141497140588445e-05,
+      "loss": 3.9644,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140658545837393e-05,
+      "loss": 3.9511,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.139819951086341e-05,
+      "loss": 3.9486,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.138981356335289e-05,
+      "loss": 3.9601,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13814439946461e-05,
+      "loss": 3.9593,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.137305804713558e-05,
+      "loss": 3.9586,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.136467209962506e-05,
+      "loss": 3.949,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.135628615211454e-05,
+      "loss": 3.953,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.134790020460402e-05,
+      "loss": 3.9649,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.13395142570935e-05,
+      "loss": 3.9675,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.133112830958298e-05,
+      "loss": 3.9535,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.132274236207245e-05,
+      "loss": 3.9545,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.131437279336567e-05,
+      "loss": 3.9644,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.130598684585515e-05,
+      "loss": 3.9562,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1297600898344634e-05,
+      "loss": 3.9548,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.128921495083411e-05,
+      "loss": 3.9515,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1280845382127324e-05,
+      "loss": 3.9502,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.127247581342053e-05,
+      "loss": 3.9636,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.126408986591001e-05,
+      "loss": 3.9432,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.023078918457031,
+      "eval_runtime": 294.5903,
+      "eval_samples_per_second": 1295.328,
+      "eval_steps_per_second": 40.48,
+      "step": 1144800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.125570391839949e-05,
+      "loss": 3.951,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.124731797088897e-05,
+      "loss": 3.9443,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123893202337845e-05,
+      "loss": 3.9452,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1230546075867926e-05,
+      "loss": 3.9676,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1222160128357406e-05,
+      "loss": 3.9638,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1213774180846886e-05,
+      "loss": 3.95,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.120538823333637e-05,
+      "loss": 3.9584,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119701866462958e-05,
+      "loss": 3.9498,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11886490959228e-05,
+      "loss": 3.9504,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118026314841228e-05,
+      "loss": 3.9588,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117187720090176e-05,
+      "loss": 3.9534,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116349125339123e-05,
+      "loss": 3.9581,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1155138063488175e-05,
+      "loss": 3.9668,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1146752115977655e-05,
+      "loss": 3.9458,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1138366168467135e-05,
+      "loss": 3.9438,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1129980220956615e-05,
+      "loss": 3.9371,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1121594273446095e-05,
+      "loss": 3.9543,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1113208325935575e-05,
+      "loss": 3.9537,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1104822378425055e-05,
+      "loss": 3.9483,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1096436430914535e-05,
+      "loss": 3.9507,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1088050483404015e-05,
+      "loss": 3.9672,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1079664535893495e-05,
+      "loss": 3.9546,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1071278588382975e-05,
+      "loss": 3.9572,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1062909019676184e-05,
+      "loss": 3.9524,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1054523072165664e-05,
+      "loss": 3.9632,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1046137124655144e-05,
+      "loss": 3.9486,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1037751177144624e-05,
+      "loss": 3.9497,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1029365229634104e-05,
+      "loss": 3.9461,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102099566092731e-05,
+      "loss": 3.9451,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101260971341679e-05,
+      "loss": 3.9489,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100424014471001e-05,
+      "loss": 3.9487,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099585419719949e-05,
+      "loss": 3.9503,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098746824968897e-05,
+      "loss": 3.9542,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097908230217845e-05,
+      "loss": 3.9542,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097069635466793e-05,
+      "loss": 3.9592,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096232678596114e-05,
+      "loss": 3.9549,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095394083845062e-05,
+      "loss": 3.9579,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09455548909401e-05,
+      "loss": 3.9338,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093716894342958e-05,
+      "loss": 3.9593,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092878299591906e-05,
+      "loss": 3.9335,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092039704840854e-05,
+      "loss": 3.9499,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.091201110089802e-05,
+      "loss": 3.9508,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09036251533875e-05,
+      "loss": 3.9585,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.089523920587698e-05,
+      "loss": 3.9443,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.088686963717019e-05,
+      "loss": 3.9576,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087848368965967e-05,
+      "loss": 3.9578,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.087009774214915e-05,
+      "loss": 3.9558,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.086171179463863e-05,
+      "loss": 3.9547,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.085334222593184e-05,
+      "loss": 3.9285,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.084497265722505e-05,
+      "loss": 3.9492,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.083658670971453e-05,
+      "loss": 3.9602,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.082820076220401e-05,
+      "loss": 3.9536,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081981481469349e-05,
+      "loss": 3.9494,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.081142886718297e-05,
+      "loss": 3.9396,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.080304291967245e-05,
+      "loss": 3.9434,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.079467335096567e-05,
+      "loss": 3.9325,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.078628740345515e-05,
+      "loss": 3.9597,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077790145594463e-05,
+      "loss": 3.9305,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076951550843411e-05,
+      "loss": 3.9542,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076112956092359e-05,
+      "loss": 3.9527,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075274361341306e-05,
+      "loss": 3.9408,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074435766590254e-05,
+      "loss": 3.9469,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073597171839202e-05,
+      "loss": 3.936,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.07275857708815e-05,
+      "loss": 3.94,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071919982337098e-05,
+      "loss": 3.936,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071081387586046e-05,
+      "loss": 3.9568,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.070242792834994e-05,
+      "loss": 3.9391,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.069405835964315e-05,
+      "loss": 3.9382,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0685672412132635e-05,
+      "loss": 3.9335,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0677286464622115e-05,
+      "loss": 3.943,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0668900517111595e-05,
+      "loss": 3.9558,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0660514569601075e-05,
+      "loss": 3.9493,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0652161379698014e-05,
+      "loss": 3.9405,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0643775432187494e-05,
+      "loss": 3.951,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0635389484676973e-05,
+      "loss": 3.9604,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0627003537166453e-05,
+      "loss": 3.9358,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0618617589655933e-05,
+      "loss": 3.9553,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.061023164214541e-05,
+      "loss": 3.9437,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060186207343862e-05,
+      "loss": 3.9432,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05934761259281e-05,
+      "loss": 3.939,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058509017841759e-05,
+      "loss": 3.9467,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057670423090707e-05,
+      "loss": 3.9577,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056833466220028e-05,
+      "loss": 3.9545,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055994871468976e-05,
+      "loss": 3.9441,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055156276717924e-05,
+      "loss": 3.9231,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.054317681966872e-05,
+      "loss": 3.9427,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05347908721582e-05,
+      "loss": 3.9463,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.052640492464768e-05,
+      "loss": 3.9479,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051801897713716e-05,
+      "loss": 3.9364,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509633029626638e-05,
+      "loss": 3.9505,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501247082116118e-05,
+      "loss": 3.947,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.049287751340933e-05,
+      "loss": 3.9485,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.048449156589881e-05,
+      "loss": 3.9355,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.047610561838829e-05,
+      "loss": 3.95,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.046771967087777e-05,
+      "loss": 3.9378,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045935010217098e-05,
+      "loss": 3.9566,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045096415466046e-05,
+      "loss": 3.9407,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0442578207149943e-05,
+      "loss": 3.9317,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0434192259639423e-05,
+      "loss": 3.9472,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0425806312128896e-05,
+      "loss": 3.947,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.041743674342211e-05,
+      "loss": 3.9524,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040905079591159e-05,
+      "loss": 3.9448,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040066484840107e-05,
+      "loss": 3.9406,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0392278900890548e-05,
+      "loss": 3.9365,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0383892953380028e-05,
+      "loss": 3.9467,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0375507005869508e-05,
+      "loss": 3.9312,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0367137437162717e-05,
+      "loss": 3.9449,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0358751489652197e-05,
+      "loss": 3.9427,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.035036554214168e-05,
+      "loss": 3.9413,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.034197959463116e-05,
+      "loss": 3.9323,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.033361002592437e-05,
+      "loss": 3.9436,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.032522407841385e-05,
+      "loss": 3.9324,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.031683813090333e-05,
+      "loss": 3.9471,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.030845218339281e-05,
+      "loss": 3.9414,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.030006623588229e-05,
+      "loss": 3.943,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0291680288371773e-05,
+      "loss": 3.95,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.028331071966498e-05,
+      "loss": 3.9501,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.027492477215446e-05,
+      "loss": 3.946,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.026653882464394e-05,
+      "loss": 3.9374,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.025815287713342e-05,
+      "loss": 3.9448,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.02497669296229e-05,
+      "loss": 3.9434,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.024138098211238e-05,
+      "loss": 3.9362,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.023299503460186e-05,
+      "loss": 3.9481,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0224609087091345e-05,
+      "loss": 3.939,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0216223139580825e-05,
+      "loss": 3.9467,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0207853570874034e-05,
+      "loss": 3.9351,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0199484002167243e-05,
+      "loss": 3.9362,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0191098054656726e-05,
+      "loss": 3.944,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0182712107146206e-05,
+      "loss": 3.9382,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0174326159635686e-05,
+      "loss": 3.9371,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0165940212125166e-05,
+      "loss": 3.9571,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0157570643418375e-05,
+      "loss": 3.944,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0149184695907855e-05,
+      "loss": 3.9377,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0140798748397335e-05,
+      "loss": 3.948,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132412800886815e-05,
+      "loss": 3.946,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01240268533763e-05,
+      "loss": 3.9493,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115657284669508e-05,
+      "loss": 3.9416,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0107271337158987e-05,
+      "loss": 3.9388,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098885389648467e-05,
+      "loss": 3.9521,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0090499442137947e-05,
+      "loss": 3.9553,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0082113494627427e-05,
+      "loss": 3.9436,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.007374392592064e-05,
+      "loss": 3.9422,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.006535797841012e-05,
+      "loss": 3.956,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.00569720308996e-05,
+      "loss": 3.9408,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.004860246219281e-05,
+      "loss": 3.9446,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.004021651468229e-05,
+      "loss": 3.9412,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.003183056717177e-05,
+      "loss": 3.9397,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0023444619661252e-05,
+      "loss": 3.9521,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0015058672150732e-05,
+      "loss": 3.9342,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.0006672724640205e-05,
+      "loss": 3.939,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.017744541168213,
+      "eval_runtime": 295.6323,
+      "eval_samples_per_second": 1290.762,
+      "eval_steps_per_second": 40.337,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9998286777129685e-05,
+      "loss": 3.9309,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989900829619165e-05,
+      "loss": 3.9332,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9981514882108645e-05,
+      "loss": 3.9562,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.997312893459813e-05,
+      "loss": 3.9532,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.996474298708761e-05,
+      "loss": 3.9406,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.995635703957709e-05,
+      "loss": 3.9477,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9947971092066568e-05,
+      "loss": 3.9378,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9939585144556048e-05,
+      "loss": 3.9345,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931199197045528e-05,
+      "loss": 3.9511,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922813249535008e-05,
+      "loss": 3.9419,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914427302024488e-05,
+      "loss": 3.9503,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9906041354513968e-05,
+      "loss": 3.9498,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.989767178580718e-05,
+      "loss": 3.9414,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988928583829666e-05,
+      "loss": 3.9346,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988089989078614e-05,
+      "loss": 3.9251,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.987251394327562e-05,
+      "loss": 3.9393,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9864127995765097e-05,
+      "loss": 3.9434,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9855742048254577e-05,
+      "loss": 3.9392,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9847372479547793e-05,
+      "loss": 3.9391,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9838986532037266e-05,
+      "loss": 3.9604,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9830600584526746e-05,
+      "loss": 3.9435,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822214637016226e-05,
+      "loss": 3.9444,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813828689505706e-05,
+      "loss": 3.9417,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.980544274199519e-05,
+      "loss": 3.9576,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.979705679448467e-05,
+      "loss": 3.9327,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978867084697415e-05,
+      "loss": 3.9389,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978028489946363e-05,
+      "loss": 3.934,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9771915330756838e-05,
+      "loss": 3.9384,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9763529383246318e-05,
+      "loss": 3.9386,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9755143435735798e-05,
+      "loss": 3.9373,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.974675748822528e-05,
+      "loss": 3.9406,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.973837154071476e-05,
+      "loss": 3.9451,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972998559320424e-05,
+      "loss": 3.9407,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.972159964569372e-05,
+      "loss": 3.9512,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.971323007698693e-05,
+      "loss": 3.9407,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.970484412947641e-05,
+      "loss": 3.9491,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696474560769623e-05,
+      "loss": 3.9245,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9688088613259103e-05,
+      "loss": 3.9495,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9679702665748583e-05,
+      "loss": 3.9248,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9671316718238063e-05,
+      "loss": 3.9417,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662930770727543e-05,
+      "loss": 3.9352,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654544823217023e-05,
+      "loss": 3.9467,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9646158875706503e-05,
+      "loss": 3.9339,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9637772928195983e-05,
+      "loss": 3.9462,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962938698068546e-05,
+      "loss": 3.9445,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9621017411978675e-05,
+      "loss": 3.9442,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9612631464468155e-05,
+      "loss": 3.9451,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9604261895761364e-05,
+      "loss": 3.9201,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9595875948250844e-05,
+      "loss": 3.9388,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9587490000740324e-05,
+      "loss": 3.9477,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9579104053229807e-05,
+      "loss": 3.9415,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.957071810571928e-05,
+      "loss": 3.9422,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.956233215820876e-05,
+      "loss": 3.9281,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.955394621069824e-05,
+      "loss": 3.9344,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.954556026318772e-05,
+      "loss": 3.9203,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.95371743156772e-05,
+      "loss": 3.9507,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528788368166684e-05,
+      "loss": 3.9174,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520402420656164e-05,
+      "loss": 3.9447,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9512016473145644e-05,
+      "loss": 3.9398,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9503663283242582e-05,
+      "loss": 3.9293,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9495277335732062e-05,
+      "loss": 3.9338,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486891388221545e-05,
+      "loss": 3.9308,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478505440711025e-05,
+      "loss": 3.93,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9470119493200505e-05,
+      "loss": 3.9239,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9461733545689985e-05,
+      "loss": 3.9459,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9453347598179465e-05,
+      "loss": 3.9312,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9444961650668945e-05,
+      "loss": 3.9273,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9436592081962154e-05,
+      "loss": 3.9244,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9428222513255366e-05,
+      "loss": 3.9302,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9419836565744846e-05,
+      "loss": 3.9434,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411450618234326e-05,
+      "loss": 3.9428,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9403064670723806e-05,
+      "loss": 3.9293,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394695102017015e-05,
+      "loss": 3.9424,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.93863091545065e-05,
+      "loss": 3.9464,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377939585799708e-05,
+      "loss": 3.9263,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369553638289188e-05,
+      "loss": 3.9412,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9361167690778668e-05,
+      "loss": 3.9359,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9352781743268148e-05,
+      "loss": 3.9356,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344395795757628e-05,
+      "loss": 3.9301,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9336009848247108e-05,
+      "loss": 3.9352,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.932762390073659e-05,
+      "loss": 3.9444,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931923795322607e-05,
+      "loss": 3.9512,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.931085200571555e-05,
+      "loss": 3.9289,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.930246605820503e-05,
+      "loss": 3.9119,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.929409648949824e-05,
+      "loss": 3.9318,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.928571054198772e-05,
+      "loss": 3.936,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.92773245944772e-05,
+      "loss": 3.9368,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9268938646966683e-05,
+      "loss": 3.9315,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260552699456163e-05,
+      "loss": 3.9373,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9252166751945643e-05,
+      "loss": 3.9376,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243797183238852e-05,
+      "loss": 3.9364,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235411235728332e-05,
+      "loss": 3.9248,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9227025288217812e-05,
+      "loss": 3.9386,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921863934070729e-05,
+      "loss": 3.9287,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.921025339319677e-05,
+      "loss": 3.9445,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9201883824489984e-05,
+      "loss": 3.9332,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193497876979464e-05,
+      "loss": 3.9184,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9185111929468938e-05,
+      "loss": 3.9371,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.917672598195842e-05,
+      "loss": 3.9377,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.91683400344479e-05,
+      "loss": 3.9441,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915995408693738e-05,
+      "loss": 3.9316,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.915156813942686e-05,
+      "loss": 3.9331,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.914318219191634e-05,
+      "loss": 3.924,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.913481262320955e-05,
+      "loss": 3.9383,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126443054502762e-05,
+      "loss": 3.9198,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9118057106992242e-05,
+      "loss": 3.9308,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109671159481722e-05,
+      "loss": 3.9337,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9101285211971202e-05,
+      "loss": 3.932,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092899264460682e-05,
+      "loss": 3.924,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.908452969575389e-05,
+      "loss": 3.93,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9076143748243375e-05,
+      "loss": 3.9268,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067757800732855e-05,
+      "loss": 3.9368,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059371853222335e-05,
+      "loss": 3.9293,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9050985905711815e-05,
+      "loss": 3.9315,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9042599958201294e-05,
+      "loss": 3.9455,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9034214010690774e-05,
+      "loss": 3.9411,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9025828063180254e-05,
+      "loss": 3.9337,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9017442115669734e-05,
+      "loss": 3.9277,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9009072546962947e-05,
+      "loss": 3.9346,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9000686599452427e-05,
+      "loss": 3.9283,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8992317030745636e-05,
+      "loss": 3.9279,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8983931083235116e-05,
+      "loss": 3.9424,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8975545135724596e-05,
+      "loss": 3.9262,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8967159188214076e-05,
+      "loss": 3.9373,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958773240703556e-05,
+      "loss": 3.9238,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.895038729319304e-05,
+      "loss": 3.9265,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.894200134568252e-05,
+      "loss": 3.9347,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8933615398172e-05,
+      "loss": 3.9288,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925262208268937e-05,
+      "loss": 3.9223,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916876260758417e-05,
+      "loss": 3.9488,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.89084903132479e-05,
+      "loss": 3.9344,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.890010436573738e-05,
+      "loss": 3.9312,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.889171841822686e-05,
+      "loss": 3.9319,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.888333247071634e-05,
+      "loss": 3.9335,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887496290200955e-05,
+      "loss": 3.9392,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886657695449903e-05,
+      "loss": 3.9335,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8858207385792242e-05,
+      "loss": 3.9286,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849821438281722e-05,
+      "loss": 3.944,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841435490771202e-05,
+      "loss": 3.9388,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8833049543260682e-05,
+      "loss": 3.9373,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8824663595750162e-05,
+      "loss": 3.931,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881627764823964e-05,
+      "loss": 3.9389,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880789170072912e-05,
+      "loss": 3.9352,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8799505753218598e-05,
+      "loss": 3.9341,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8791119805708078e-05,
+      "loss": 3.9335,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8782733858197558e-05,
+      "loss": 3.9315,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774364289490774e-05,
+      "loss": 3.9394,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765978341980247e-05,
+      "loss": 3.9233,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.875759239446973e-05,
+      "loss": 3.9282,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.013113498687744,
+      "eval_runtime": 297.6157,
+      "eval_samples_per_second": 1282.16,
+      "eval_steps_per_second": 40.068,
+      "step": 1297440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.874920644695921e-05,
+      "loss": 3.9242,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8740836878252426e-05,
+      "loss": 3.9247,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.87324509307419e-05,
+      "loss": 3.9482,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.872406498323138e-05,
+      "loss": 3.9385,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715695414524595e-05,
+      "loss": 3.9344,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707309467014072e-05,
+      "loss": 3.9329,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698923519503552e-05,
+      "loss": 3.9303,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690553950796768e-05,
+      "loss": 3.9248,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8682168003286248e-05,
+      "loss": 3.9376,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.867378205577572e-05,
+      "loss": 3.9352,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.86653961082652e-05,
+      "loss": 3.9362,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8657010160754684e-05,
+      "loss": 3.9422,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8648640592047893e-05,
+      "loss": 3.9292,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640254644537373e-05,
+      "loss": 3.9262,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631868697026853e-05,
+      "loss": 3.9196,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8623482749516333e-05,
+      "loss": 3.9223,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8615096802005813e-05,
+      "loss": 3.9335,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8606710854495293e-05,
+      "loss": 3.9253,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8598324906984776e-05,
+      "loss": 3.9316,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8589938959474256e-05,
+      "loss": 3.9528,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8581553011963736e-05,
+      "loss": 3.9342,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8573167064453216e-05,
+      "loss": 3.9376,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564781116942696e-05,
+      "loss": 3.9294,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8556395169432176e-05,
+      "loss": 3.9442,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8548009221921656e-05,
+      "loss": 3.9256,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539623274411136e-05,
+      "loss": 3.9296,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531237326900613e-05,
+      "loss": 3.9225,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.852286775819383e-05,
+      "loss": 3.9321,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.851448181068331e-05,
+      "loss": 3.9266,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.850609586317278e-05,
+      "loss": 3.926,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497726294465998e-05,
+      "loss": 3.9313,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489340346955477e-05,
+      "loss": 3.9399,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8480954399444957e-05,
+      "loss": 3.9289,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472568451934434e-05,
+      "loss": 3.9365,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.846419888322765e-05,
+      "loss": 3.9276,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.845581293571713e-05,
+      "loss": 3.9397,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.844742698820661e-05,
+      "loss": 3.9171,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8439041040696086e-05,
+      "loss": 3.9337,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8430655093185566e-05,
+      "loss": 3.9141,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8422269145675046e-05,
+      "loss": 3.9345,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8413899576968255e-05,
+      "loss": 3.9226,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8405513629457735e-05,
+      "loss": 3.9388,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8397127681947215e-05,
+      "loss": 3.9215,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8388741734436695e-05,
+      "loss": 3.9353,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838035578692618e-05,
+      "loss": 3.9375,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.837196983941566e-05,
+      "loss": 3.9345,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8363600270708868e-05,
+      "loss": 3.9401,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8355214323198348e-05,
+      "loss": 3.9107,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8346828375687828e-05,
+      "loss": 3.9223,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338442428177307e-05,
+      "loss": 3.9356,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8330056480666787e-05,
+      "loss": 3.9331,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.832167053315627e-05,
+      "loss": 3.9279,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.831328458564575e-05,
+      "loss": 3.9237,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.830489863813523e-05,
+      "loss": 3.923,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.829654544823217e-05,
+      "loss": 3.9155,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.828815950072165e-05,
+      "loss": 3.931,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279773553211132e-05,
+      "loss": 3.914,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271387605700612e-05,
+      "loss": 3.9273,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8263001658190092e-05,
+      "loss": 3.9355,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.82546320894833e-05,
+      "loss": 3.9205,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.824624614197278e-05,
+      "loss": 3.9191,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.823786019446226e-05,
+      "loss": 3.9207,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229490625755474e-05,
+      "loss": 3.9142,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8221104678244954e-05,
+      "loss": 3.9156,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8212718730734434e-05,
+      "loss": 3.9359,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8204332783223913e-05,
+      "loss": 3.9228,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8195946835713393e-05,
+      "loss": 3.9166,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8187560888202873e-05,
+      "loss": 3.9166,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8179174940692353e-05,
+      "loss": 3.9182,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8170788993181833e-05,
+      "loss": 3.9327,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8162403045671317e-05,
+      "loss": 3.9306,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8154017098160797e-05,
+      "loss": 3.9207,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.814563115065027e-05,
+      "loss": 3.9319,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.813724520313975e-05,
+      "loss": 3.9362,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128875634432966e-05,
+      "loss": 3.9201,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8120489686922442e-05,
+      "loss": 3.9274,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8112103739411922e-05,
+      "loss": 3.927,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8103734170705138e-05,
+      "loss": 3.9276,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8095348223194618e-05,
+      "loss": 3.9214,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.808696227568409e-05,
+      "loss": 3.923,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.807857632817357e-05,
+      "loss": 3.9367,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070190380663054e-05,
+      "loss": 3.9341,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061804433152534e-05,
+      "loss": 3.9253,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053418485642014e-05,
+      "loss": 3.9039,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.804506529573896e-05,
+      "loss": 3.9209,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.803667934822844e-05,
+      "loss": 3.9261,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8028293400717916e-05,
+      "loss": 3.9263,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8019907453207396e-05,
+      "loss": 3.9234,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8011537884500612e-05,
+      "loss": 3.9237,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800315193699009e-05,
+      "loss": 3.9295,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7994765989479565e-05,
+      "loss": 3.9214,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7986380041969045e-05,
+      "loss": 3.9158,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7977994094458525e-05,
+      "loss": 3.9311,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7969608146948008e-05,
+      "loss": 3.9153,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7961222199437488e-05,
+      "loss": 3.937,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7952836251926968e-05,
+      "loss": 3.9267,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7944450304416448e-05,
+      "loss": 3.9077,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7936064356905928e-05,
+      "loss": 3.9261,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7927678409395408e-05,
+      "loss": 3.924,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919292461884888e-05,
+      "loss": 3.9353,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910906514374368e-05,
+      "loss": 3.9252,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902520566863848e-05,
+      "loss": 3.9223,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7894134619353328e-05,
+      "loss": 3.9147,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.788576505064654e-05,
+      "loss": 3.9302,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.787737910313602e-05,
+      "loss": 3.91,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.78689931556255e-05,
+      "loss": 3.9193,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.786060720811498e-05,
+      "loss": 3.925,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7852221260604457e-05,
+      "loss": 3.9202,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7843835313093937e-05,
+      "loss": 3.913,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7835465744387152e-05,
+      "loss": 3.9247,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7827079796876632e-05,
+      "loss": 3.9176,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7818693849366106e-05,
+      "loss": 3.9233,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7810307901855586e-05,
+      "loss": 3.9227,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7801921954345066e-05,
+      "loss": 3.9151,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.779353600683455e-05,
+      "loss": 3.9428,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.778515005932403e-05,
+      "loss": 3.9312,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.777676411181351e-05,
+      "loss": 3.9226,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768394543106718e-05,
+      "loss": 3.9172,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7760008595596198e-05,
+      "loss": 3.9233,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.775163902688941e-05,
+      "loss": 3.9219,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.774325307937889e-05,
+      "loss": 3.9156,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.773486713186837e-05,
+      "loss": 3.9324,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.772648118435785e-05,
+      "loss": 3.9214,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.771811161565106e-05,
+      "loss": 3.9228,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.770972566814054e-05,
+      "loss": 3.9197,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.770133972063002e-05,
+      "loss": 3.9187,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7692953773119503e-05,
+      "loss": 3.9226,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684567825608982e-05,
+      "loss": 3.9226,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.767619825690219e-05,
+      "loss": 3.9089,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.766781230939167e-05,
+      "loss": 3.9389,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765942636188115e-05,
+      "loss": 3.9311,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.765104041437063e-05,
+      "loss": 3.9212,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.764265446686011e-05,
+      "loss": 3.9225,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634268519349595e-05,
+      "loss": 3.9199,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7625882571839075e-05,
+      "loss": 3.9334,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7617496624328555e-05,
+      "loss": 3.9265,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7609127055621764e-05,
+      "loss": 3.9201,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7600741108111244e-05,
+      "loss": 3.9339,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7592371539404456e-05,
+      "loss": 3.9253,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7583985591893936e-05,
+      "loss": 3.9293,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7575599644383416e-05,
+      "loss": 3.9216,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7567213696872896e-05,
+      "loss": 3.932,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7558827749362376e-05,
+      "loss": 3.9257,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7550441801851856e-05,
+      "loss": 3.9242,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7542055854341336e-05,
+      "loss": 3.9242,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.753368628563455e-05,
+      "loss": 3.9244,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.752530033812403e-05,
+      "loss": 3.9274,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.751691439061351e-05,
+      "loss": 3.9147,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508528443102988e-05,
+      "loss": 3.9195,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.008997440338135,
+      "eval_runtime": 292.3581,
+      "eval_samples_per_second": 1305.218,
+      "eval_steps_per_second": 40.789,
+      "step": 1373760
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 5.676403438262292e+17,
+  "trial_name": null,
+  "trial_params": null
+}