diff --git "a/checkpoint-610560/trainer_state.json" "b/checkpoint-610560/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-610560/trainer_state.json"
@@ -0,0 +1,7241 @@
+{
+  "best_metric": 4.105152130126953,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/npi-sent-neg/lstm/0/checkpoints/checkpoint-610560",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 610560,
+  "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": 10.8186,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.553,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0592,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9948,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.948,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.9162,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7528,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.641,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5477,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.473,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.4158,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3505,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2821,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2031,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1533,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.0877,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0439,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 5.9966,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9585,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9177,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8838,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8444,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8116,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7794,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7602,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7233,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6991,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6725,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6432,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6329,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6088,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5993,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5635,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5524,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5409,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.970657373115046e-05,
+      "loss": 5.5228,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5022,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4685,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4697,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4366,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4386,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4212,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4101,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.3925,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3855,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3848,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3673,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.363,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3348,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3217,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.3314,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957241494978587e-05,
+      "loss": 5.3156,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3077,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955567581237229e-05,
+      "loss": 5.2786,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954728986486177e-05,
+      "loss": 5.2929,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.2686,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2897,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2401,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2521,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2318,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2208,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2221,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2075,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.1871,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.1891,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.208,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1806,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1691,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1589,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1426,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1579,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404794272397856e-05,
+      "loss": 5.1498,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9396408324887336e-05,
+      "loss": 5.1414,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388038756180545e-05,
+      "loss": 5.1477,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379652808670025e-05,
+      "loss": 5.1318,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.1268,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362897292452714e-05,
+      "loss": 5.1051,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354511344942194e-05,
+      "loss": 5.1052,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346125397431674e-05,
+      "loss": 5.0958,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337739449921154e-05,
+      "loss": 5.0873,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329353502410634e-05,
+      "loss": 5.0878,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9320967554900114e-05,
+      "loss": 5.0846,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9312581607389594e-05,
+      "loss": 5.0857,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0759,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0512,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0608,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0619,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.063,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0407,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0444,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0357,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.027,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0221,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0192,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0048,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0142,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 4.9967,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918684153114299e-05,
+      "loss": 4.9891,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917845558363247e-05,
+      "loss": 4.9853,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917006963612195e-05,
+      "loss": 4.9887,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 4.9828,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9678,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.97,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.9575,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9601,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9375,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9463,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.94,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9354,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908625929742794e-05,
+      "loss": 4.9402,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907787334991742e-05,
+      "loss": 4.9329,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90694874024069e-05,
+      "loss": 4.912,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9222,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.924,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9163,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9159,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9069,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.902,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.891,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.8984,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.8945,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8985660684909166e-05,
+      "loss": 4.8782,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8804,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8756,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960519221181335e-05,
+      "loss": 4.883,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952133273670815e-05,
+      "loss": 4.8598,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943747326160294e-05,
+      "loss": 4.8656,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8935361378649774e-05,
+      "loss": 4.853,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8926975431139254e-05,
+      "loss": 4.8599,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918589483628734e-05,
+      "loss": 4.8494,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8632,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.851,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88934480199009e-05,
+      "loss": 4.844,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8432,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766925036836e-05,
+      "loss": 4.8375,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886830655617308e-05,
+      "loss": 4.8448,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885992060866256e-05,
+      "loss": 4.8437,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885153466115204e-05,
+      "loss": 4.8339,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884316509244525e-05,
+      "loss": 4.8289,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883477914493473e-05,
+      "loss": 4.8303,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882639319742421e-05,
+      "loss": 4.8183,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881800724991369e-05,
+      "loss": 4.8061,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880962130240316e-05,
+      "loss": 4.815,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880123535489264e-05,
+      "loss": 4.8067,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879284940738212e-05,
+      "loss": 4.7986,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.798,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.7965,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87677079436543e-05,
+      "loss": 4.7984,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.8001,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7883,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.753722667694092,
+      "eval_runtime": 297.919,
+      "eval_samples_per_second": 1280.855,
+      "eval_steps_per_second": 40.028,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8742566479926466e-05,
+      "loss": 4.7777,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8734180532415946e-05,
+      "loss": 4.7753,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8725794584905426e-05,
+      "loss": 4.7861,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7783,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7805,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700653121177595e-05,
+      "loss": 4.7634,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692267173667075e-05,
+      "loss": 4.7647,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683881226156555e-05,
+      "loss": 4.759,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.867549527864604e-05,
+      "loss": 4.7636,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7644,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7605,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7656,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7361,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7466,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7404,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7309,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7451,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860005450865882e-05,
+      "loss": 4.7345,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85916685611483e-05,
+      "loss": 4.7316,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858329899244151e-05,
+      "loss": 4.7522,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8574913044930995e-05,
+      "loss": 4.7317,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8566527097420475e-05,
+      "loss": 4.7278,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7262,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7386,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7188,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7154,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7148,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851621141235735e-05,
+      "loss": 4.711,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.71,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8499455896140044e-05,
+      "loss": 4.7026,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7243,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.7044,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.7026,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.846592848490169e-05,
+      "loss": 4.7048,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845755891619491e-05,
+      "loss": 4.7023,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844917296868439e-05,
+      "loss": 4.6963,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844078702117387e-05,
+      "loss": 4.6777,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843240107366335e-05,
+      "loss": 4.6842,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842401512615282e-05,
+      "loss": 4.6713,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.84156291786423e-05,
+      "loss": 4.6881,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840724323113178e-05,
+      "loss": 4.6769,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.68,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839047133611074e-05,
+      "loss": 4.6711,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.6752,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6791,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.67,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.6746,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.6606,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.6562,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.6653,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6606,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.667,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6373,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.6617,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289872723591966e-05,
+      "loss": 4.6371,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.6708,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.6332,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.6492,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.6379,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6292,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.6351,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.6342,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.6158,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6226,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6382,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.624,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6218,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6186,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.6063,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6248,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.62,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.6213,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8139007562421265e-05,
+      "loss": 4.6285,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8130621614910745e-05,
+      "loss": 4.6234,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8122235667400224e-05,
+      "loss": 4.6199,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.5992,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.6108,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.605,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.5942,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.6048,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.6041,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8063550413630307e-05,
+      "loss": 4.6135,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.6043,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8046794897413e-05,
+      "loss": 4.584,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803840894990248e-05,
+      "loss": 4.5928,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803002300239196e-05,
+      "loss": 4.603,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.6049,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.5946,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5987,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799649559115361e-05,
+      "loss": 4.5934,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798810964364309e-05,
+      "loss": 4.5815,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797972369613257e-05,
+      "loss": 4.5883,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797133774862205e-05,
+      "loss": 4.5879,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.5774,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795458223240474e-05,
+      "loss": 4.5928,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794619628489423e-05,
+      "loss": 4.575,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793781033738371e-05,
+      "loss": 4.5723,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.5741,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.5811,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.5785,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.5667,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.5663,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887494652320587e-05,
+      "loss": 4.5614,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879108704810067e-05,
+      "loss": 4.5712,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5476,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786233680978902e-05,
+      "loss": 4.5558,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.565,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5587,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837211724864925e-05,
+      "loss": 4.5587,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.5572,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782043982984389e-05,
+      "loss": 4.5471,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.781205388233337e-05,
+      "loss": 4.556,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5537,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.553,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5567,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5559,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.5462,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776175457607397e-05,
+      "loss": 4.5402,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775336862856345e-05,
+      "loss": 4.5509,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5447,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.5318,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.5421,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.5356,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5459,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.5295,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.5369,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5144,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767792785857624e-05,
+      "loss": 4.538,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766954191106572e-05,
+      "loss": 4.5253,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766117234235893e-05,
+      "loss": 4.5435,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765278639484841e-05,
+      "loss": 4.5336,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764440044733789e-05,
+      "loss": 4.531,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763601449982737e-05,
+      "loss": 4.5212,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762762855231685e-05,
+      "loss": 4.5243,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761924260480633e-05,
+      "loss": 4.5351,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5393,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760248708858902e-05,
+      "loss": 4.5278,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.5237,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5221,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5189,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.5113,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756055735103642e-05,
+      "loss": 4.5196,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.755218778232963e-05,
+      "loss": 4.5119,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5069,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753541588730859e-05,
+      "loss": 4.5095,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.752702993979807e-05,
+      "loss": 4.5075,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751864399228755e-05,
+      "loss": 4.5108,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751027442358077e-05,
+      "loss": 4.5118,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.5088,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.475302696228027,
+      "eval_runtime": 295.3079,
+      "eval_samples_per_second": 1292.18,
+      "eval_steps_per_second": 40.382,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.4942,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.4973,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747673063353868e-05,
+      "loss": 4.5066,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.5038,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.5091,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745157279100712e-05,
+      "loss": 4.4866,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74431868434966e-05,
+      "loss": 4.4975,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.4891,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.4978,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.4982,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.4991,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401257105944e-05,
+      "loss": 4.4996,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.4754,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384501589726696e-05,
+      "loss": 4.4845,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376115642216176e-05,
+      "loss": 4.4863,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.4776,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.4885,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.4801,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.48,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.5073,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.4793,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4829,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4795,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4944,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292272545914704e-05,
+      "loss": 4.4736,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728390297720791e-05,
+      "loss": 4.472,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727551702969739e-05,
+      "loss": 4.4748,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.726713108218687e-05,
+      "loss": 4.4763,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.725874513467635e-05,
+      "loss": 4.4668,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725035918716583e-05,
+      "loss": 4.4694,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724197323965531e-05,
+      "loss": 4.4861,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723360367094852e-05,
+      "loss": 4.4728,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225217723438e-05,
+      "loss": 4.4693,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721683177592748e-05,
+      "loss": 4.475,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720846220722069e-05,
+      "loss": 4.4704,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720007625971017e-05,
+      "loss": 4.4669,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4494,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.4634,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717491841717862e-05,
+      "loss": 4.447,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71665324696681e-05,
+      "loss": 4.4617,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715814652215758e-05,
+      "loss": 4.4536,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714976057464706e-05,
+      "loss": 4.4612,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.453,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7133021437233476e-05,
+      "loss": 4.4553,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124635489722956e-05,
+      "loss": 4.458,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7116249542212436e-05,
+      "loss": 4.4559,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7107863594701916e-05,
+      "loss": 4.4604,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099477647191396e-05,
+      "loss": 4.4454,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091091699680876e-05,
+      "loss": 4.4397,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082705752170356e-05,
+      "loss": 4.457,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707433618346357e-05,
+      "loss": 4.4485,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706595023595305e-05,
+      "loss": 4.4521,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705756428844253e-05,
+      "loss": 4.431,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704917834093201e-05,
+      "loss": 4.4565,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704079239342149e-05,
+      "loss": 4.43,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703240644591097e-05,
+      "loss": 4.4617,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702402049840045e-05,
+      "loss": 4.4222,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015634550889924e-05,
+      "loss": 4.4472,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007248603379404e-05,
+      "loss": 4.4354,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4333,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990476708358364e-05,
+      "loss": 4.4319,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4354,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.697372119214106e-05,
+      "loss": 4.4184,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4239,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956965675923756e-05,
+      "loss": 4.4344,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694857972841323e-05,
+      "loss": 4.4315,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694019378090271e-05,
+      "loss": 4.4241,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931824212195925e-05,
+      "loss": 4.423,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69234382646854e-05,
+      "loss": 4.4186,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691505231717488e-05,
+      "loss": 4.4273,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690666636966436e-05,
+      "loss": 4.4269,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689828042215384e-05,
+      "loss": 4.4286,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4373,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4361,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4337,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864753010915494e-05,
+      "loss": 4.4117,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856367063404974e-05,
+      "loss": 4.4232,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.4254,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839595168383934e-05,
+      "loss": 4.4072,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831209220873414e-05,
+      "loss": 4.4212,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822823273362894e-05,
+      "loss": 4.4194,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814437325852373e-05,
+      "loss": 4.429,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6806051378341853e-05,
+      "loss": 4.4202,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4017,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678931224092827e-05,
+      "loss": 4.4145,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678092629341775e-05,
+      "loss": 4.4203,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677254034590723e-05,
+      "loss": 4.4248,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676415439839671e-05,
+      "loss": 4.4169,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675578482968993e-05,
+      "loss": 4.4227,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674739888217941e-05,
+      "loss": 4.4185,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673901293466889e-05,
+      "loss": 4.4025,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673062698715837e-05,
+      "loss": 4.4166,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672224103964785e-05,
+      "loss": 4.4096,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.402,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6705485523430536e-05,
+      "loss": 4.423,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6697099575920016e-05,
+      "loss": 4.4047,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688713628409496e-05,
+      "loss": 4.3965,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680327680898976e-05,
+      "loss": 4.405,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6671941733388456e-05,
+      "loss": 4.4113,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.4096,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655186217171145e-05,
+      "loss": 4.4004,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664681664846436e-05,
+      "loss": 4.3942,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663843070095384e-05,
+      "loss": 4.3941,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663004475344332e-05,
+      "loss": 4.4058,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66216588059328e-05,
+      "loss": 4.3811,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661327285842228e-05,
+      "loss": 4.3936,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660488691091176e-05,
+      "loss": 4.3997,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596500963401234e-05,
+      "loss": 4.3944,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588115015890714e-05,
+      "loss": 4.3943,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579729068380194e-05,
+      "loss": 4.3962,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571343120869674e-05,
+      "loss": 4.3874,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.393,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655458760465237e-05,
+      "loss": 4.3933,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654620165714185e-05,
+      "loss": 4.3904,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.3973,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.3978,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652106019341402e-05,
+      "loss": 4.3935,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65126742459035e-05,
+      "loss": 4.3799,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650428829839298e-05,
+      "loss": 4.3887,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649590235088246e-05,
+      "loss": 4.3871,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648751640337194e-05,
+      "loss": 4.3748,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647913045586142e-05,
+      "loss": 4.3902,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647076088715463e-05,
+      "loss": 4.3759,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.3903,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.3751,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644560304462307e-05,
+      "loss": 4.3838,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.365,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642884752840576e-05,
+      "loss": 4.3846,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.3705,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.3915,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.3868,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.3797,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638691779085316e-05,
+      "loss": 4.3694,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637854822214637e-05,
+      "loss": 4.3743,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637016227463585e-05,
+      "loss": 4.3868,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636177632712533e-05,
+      "loss": 4.3923,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635339037961481e-05,
+      "loss": 4.3774,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634500443210429e-05,
+      "loss": 4.3786,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633661848459377e-05,
+      "loss": 4.3771,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632823253708325e-05,
+      "loss": 4.3755,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.3671,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.3709,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630309107335542e-05,
+      "loss": 4.3696,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62947051258449e-05,
+      "loss": 4.3644,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628631917833438e-05,
+      "loss": 4.3651,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.369,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626954728331334e-05,
+      "loss": 4.3627,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626117771460655e-05,
+      "loss": 4.3693,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.3647,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.341277122497559,
+      "eval_runtime": 300.372,
+      "eval_samples_per_second": 1270.395,
+      "eval_steps_per_second": 39.701,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.3616,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3538,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3705,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.361,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210862029543436e-05,
+      "loss": 4.3702,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202476082032916e-05,
+      "loss": 4.348,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194106513326125e-05,
+      "loss": 4.3599,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185720565815605e-05,
+      "loss": 4.3488,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.3648,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168965049598294e-05,
+      "loss": 4.3605,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6160579102087774e-05,
+      "loss": 4.3626,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152193154577254e-05,
+      "loss": 4.3637,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143807207066734e-05,
+      "loss": 4.3417,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135421259556214e-05,
+      "loss": 4.3507,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612705169084942e-05,
+      "loss": 4.351,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.3482,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3516,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.352,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3449,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3786,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3471,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3499,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059980489569e-05,
+      "loss": 4.3513,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605161092086221e-05,
+      "loss": 4.3616,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.3433,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.3483,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3438,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6018083509623864e-05,
+      "loss": 4.3462,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6009697562113344e-05,
+      "loss": 4.3361,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6001311614602824e-05,
+      "loss": 4.3455,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5992925667092303e-05,
+      "loss": 4.355,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598455609838551e-05,
+      "loss": 4.3473,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597617015087499e-05,
+      "loss": 4.3465,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596778420336447e-05,
+      "loss": 4.346,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595941463465768e-05,
+      "loss": 4.3446,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595102868714716e-05,
+      "loss": 4.3481,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594264273963664e-05,
+      "loss": 4.3249,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593425679212612e-05,
+      "loss": 4.3405,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59258708446156e-05,
+      "loss": 4.3229,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591748489710508e-05,
+      "loss": 4.3389,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59091153283983e-05,
+      "loss": 4.3332,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590072938088778e-05,
+      "loss": 4.3407,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589234343337726e-05,
+      "loss": 4.3328,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588395748586673e-05,
+      "loss": 4.3318,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587557153835621e-05,
+      "loss": 4.3399,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586718559084569e-05,
+      "loss": 4.3358,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3362,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3297,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.3156,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583365817960734e-05,
+      "loss": 4.3369,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.3288,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58168862845863e-05,
+      "loss": 4.3319,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808516715879515e-05,
+      "loss": 4.3182,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800130768368995e-05,
+      "loss": 4.3408,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791744820858475e-05,
+      "loss": 4.3075,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3444,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3079,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3267,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3205,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749815083305875e-05,
+      "loss": 4.3193,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741429135795355e-05,
+      "loss": 4.3146,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3265,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572468999838177e-05,
+      "loss": 4.3017,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571630405087125e-05,
+      "loss": 4.3088,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570791810336074e-05,
+      "loss": 4.3236,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569953215585022e-05,
+      "loss": 4.3179,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56911462083397e-05,
+      "loss": 4.313,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568276026082918e-05,
+      "loss": 4.3128,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.567437431331866e-05,
+      "loss": 4.3015,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.566598836580814e-05,
+      "loss": 4.314,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565763517590508e-05,
+      "loss": 4.3163,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564924922839456e-05,
+      "loss": 4.3185,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564086328088404e-05,
+      "loss": 4.3272,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563247733337352e-05,
+      "loss": 4.3237,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624107764666726e-05,
+      "loss": 4.3237,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5615721817156206e-05,
+      "loss": 4.3029,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560735224844942e-05,
+      "loss": 4.3106,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55989663009389e-05,
+      "loss": 4.3175,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559058035342838e-05,
+      "loss": 4.2997,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558219440591786e-05,
+      "loss": 4.3137,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557380845840734e-05,
+      "loss": 4.3086,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556543888970055e-05,
+      "loss": 4.3189,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555706932099376e-05,
+      "loss": 4.3138,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554869975228697e-05,
+      "loss": 4.2971,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554031380477645e-05,
+      "loss": 4.3101,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553192785726593e-05,
+      "loss": 4.3096,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5523541909755416e-05,
+      "loss": 4.3196,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5515155962244896e-05,
+      "loss": 4.3116,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506770014734376e-05,
+      "loss": 4.3141,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498384067223856e-05,
+      "loss": 4.3166,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489998119713336e-05,
+      "loss": 4.2962,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481612172202816e-05,
+      "loss": 4.312,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473226224692296e-05,
+      "loss": 4.3011,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464840277181776e-05,
+      "loss": 4.3024,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456454329671256e-05,
+      "loss": 4.3156,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5448068382160736e-05,
+      "loss": 4.3029,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5439682434650216e-05,
+      "loss": 4.2916,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543129648713969e-05,
+      "loss": 4.3007,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542291053962917e-05,
+      "loss": 4.3109,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541452459211865e-05,
+      "loss": 4.3083,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540613864460813e-05,
+      "loss": 4.2976,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539775269709761e-05,
+      "loss": 4.292,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389366749587095e-05,
+      "loss": 4.2926,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380997180880304e-05,
+      "loss": 4.306,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372611233369784e-05,
+      "loss": 4.279,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364225285859264e-05,
+      "loss": 4.2918,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5355839338348744e-05,
+      "loss": 4.2994,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5347469769641953e-05,
+      "loss": 4.2973,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533910020093516e-05,
+      "loss": 4.2897,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533071425342464e-05,
+      "loss": 4.2997,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.532232830591412e-05,
+      "loss": 4.29,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53139423584036e-05,
+      "loss": 4.2915,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530555641089308e-05,
+      "loss": 4.295,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52971868421863e-05,
+      "loss": 4.2913,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528880089467578e-05,
+      "loss": 4.3032,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528041494716526e-05,
+      "loss": 4.2957,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527202899965474e-05,
+      "loss": 4.2963,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.526364305214422e-05,
+      "loss": 4.282,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.525527348343743e-05,
+      "loss": 4.2923,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.524688753592691e-05,
+      "loss": 4.2912,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523850158841639e-05,
+      "loss": 4.2789,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523011564090587e-05,
+      "loss": 4.2926,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522172969339535e-05,
+      "loss": 4.2827,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521334374588483e-05,
+      "loss": 4.2922,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5204974177178036e-05,
+      "loss": 4.2804,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5196588229667516e-05,
+      "loss": 4.2897,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5188202282157e-05,
+      "loss": 4.2642,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517981633464648e-05,
+      "loss": 4.2922,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517143038713596e-05,
+      "loss": 4.2778,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516304443962544e-05,
+      "loss": 4.2931,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515467487091865e-05,
+      "loss": 4.291,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514628892340813e-05,
+      "loss": 4.2905,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513790297589761e-05,
+      "loss": 4.2723,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512953340719082e-05,
+      "loss": 4.2849,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51211474596803e-05,
+      "loss": 4.2909,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511276151216978e-05,
+      "loss": 4.2961,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510437556465926e-05,
+      "loss": 4.2861,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509598961714874e-05,
+      "loss": 4.2846,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.508760366963822e-05,
+      "loss": 4.2851,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50792177221277e-05,
+      "loss": 4.2864,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.507083177461718e-05,
+      "loss": 4.2718,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.506244582710666e-05,
+      "loss": 4.2788,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505405987959614e-05,
+      "loss": 4.2766,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504569031088935e-05,
+      "loss": 4.2731,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503730436337883e-05,
+      "loss": 4.2738,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502891841586831e-05,
+      "loss": 4.2823,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502053246835779e-05,
+      "loss": 4.2713,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012162899651e-05,
+      "loss": 4.2757,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500377695214048e-05,
+      "loss": 4.2824,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.258530616760254,
+      "eval_runtime": 297.5439,
+      "eval_samples_per_second": 1282.47,
+      "eval_steps_per_second": 40.078,
+      "step": 305280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.499539100462996e-05,
+      "loss": 4.2631,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.498700505711944e-05,
+      "loss": 4.267,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978619109608925e-05,
+      "loss": 4.2794,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970233162098405e-05,
+      "loss": 4.2717,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961847214587885e-05,
+      "loss": 4.2845,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953461267077365e-05,
+      "loss": 4.2606,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945075319566845e-05,
+      "loss": 4.2694,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936689372056325e-05,
+      "loss": 4.2617,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4928303424545805e-05,
+      "loss": 4.2765,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919917477035285e-05,
+      "loss": 4.2743,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911531529524765e-05,
+      "loss": 4.2669,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490314558201424e-05,
+      "loss": 4.282,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489475963450372e-05,
+      "loss": 4.2524,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48863736869932e-05,
+      "loss": 4.2639,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.487800411828641e-05,
+      "loss": 4.2673,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486961817077589e-05,
+      "loss": 4.2584,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861232223265373e-05,
+      "loss": 4.2644,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852846275754853e-05,
+      "loss": 4.2646,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.484446032824433e-05,
+      "loss": 4.2614,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483609075953754e-05,
+      "loss": 4.2894,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482770481202702e-05,
+      "loss": 4.2625,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48193188645165e-05,
+      "loss": 4.2616,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481093291700598e-05,
+      "loss": 4.2697,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480254696949546e-05,
+      "loss": 4.2754,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479416102198494e-05,
+      "loss": 4.2565,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.478577507447442e-05,
+      "loss": 4.2672,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47773891269639e-05,
+      "loss": 4.2564,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476900317945338e-05,
+      "loss": 4.2631,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476061723194286e-05,
+      "loss": 4.2535,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475223128443234e-05,
+      "loss": 4.2627,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474384533692182e-05,
+      "loss": 4.2745,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473547576821504e-05,
+      "loss": 4.2635,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472710619950825e-05,
+      "loss": 4.2614,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471872025199773e-05,
+      "loss": 4.2652,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4710350683290936e-05,
+      "loss": 4.262,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701964735780416e-05,
+      "loss": 4.2643,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693578788269896e-05,
+      "loss": 4.244,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685192840759376e-05,
+      "loss": 4.2554,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676806893248856e-05,
+      "loss": 4.247,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668420945738336e-05,
+      "loss": 4.2524,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4660034998227816e-05,
+      "loss": 4.2542,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651649050717296e-05,
+      "loss": 4.2588,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643263103206776e-05,
+      "loss": 4.2511,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463489353449999e-05,
+      "loss": 4.2544,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462650758698947e-05,
+      "loss": 4.2569,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461812163947895e-05,
+      "loss": 4.2539,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609735691968425e-05,
+      "loss": 4.2574,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4601349744457905e-05,
+      "loss": 4.2505,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4592963796947384e-05,
+      "loss": 4.2358,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584577849436864e-05,
+      "loss": 4.2578,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576191901926344e-05,
+      "loss": 4.2469,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567822333219554e-05,
+      "loss": 4.2547,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455945276451277e-05,
+      "loss": 4.2415,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455106681700225e-05,
+      "loss": 4.2547,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.454268086949173e-05,
+      "loss": 4.2327,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453429492198121e-05,
+      "loss": 4.2635,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452590897447069e-05,
+      "loss": 4.2314,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451752302696017e-05,
+      "loss": 4.2429,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450913707944965e-05,
+      "loss": 4.2487,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.450075113193913e-05,
+      "loss": 4.2388,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.449236518442861e-05,
+      "loss": 4.2411,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.448399561572182e-05,
+      "loss": 4.2447,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.44756096682113e-05,
+      "loss": 4.2321,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.446724009950451e-05,
+      "loss": 4.2299,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445885415199399e-05,
+      "loss": 4.2432,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445046820448347e-05,
+      "loss": 4.2431,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444208225697295e-05,
+      "loss": 4.2334,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443369630946243e-05,
+      "loss": 4.24,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425310361951914e-05,
+      "loss": 4.2219,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416924414441394e-05,
+      "loss": 4.2398,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408538466930874e-05,
+      "loss": 4.2406,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440016889822408e-05,
+      "loss": 4.2409,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439178295071356e-05,
+      "loss": 4.2493,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438341338200677e-05,
+      "loss": 4.2477,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437502743449625e-05,
+      "loss": 4.253,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436664148698573e-05,
+      "loss": 4.2274,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435825553947521e-05,
+      "loss": 4.2364,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434986959196469e-05,
+      "loss": 4.2416,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43415000232579e-05,
+      "loss": 4.2201,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433311407574738e-05,
+      "loss": 4.2421,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432472812823687e-05,
+      "loss": 4.2354,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431634218072635e-05,
+      "loss": 4.2429,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.430795623321583e-05,
+      "loss": 4.2422,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4299586664509036e-05,
+      "loss": 4.2228,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4291200716998516e-05,
+      "loss": 4.2376,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.2344,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.244,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4266042874466956e-05,
+      "loss": 4.2363,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.425765692695643e-05,
+      "loss": 4.2438,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424927097944591e-05,
+      "loss": 4.2402,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424088503193539e-05,
+      "loss": 4.2267,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232515463228605e-05,
+      "loss": 4.2391,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224129515718085e-05,
+      "loss": 4.2259,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42157599470113e-05,
+      "loss": 4.2344,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420737399950078e-05,
+      "loss": 4.2442,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419898805199026e-05,
+      "loss": 4.2271,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4190602104479734e-05,
+      "loss": 4.2227,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4182216156969214e-05,
+      "loss": 4.2276,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417384658826243e-05,
+      "loss": 4.2348,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.2407,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415707469324138e-05,
+      "loss": 4.2234,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414868874573086e-05,
+      "loss": 4.2232,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414031917702408e-05,
+      "loss": 4.2193,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413193322951356e-05,
+      "loss": 4.2358,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412354728200304e-05,
+      "loss": 4.2086,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4115177713296255e-05,
+      "loss": 4.2219,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4106791765785735e-05,
+      "loss": 4.2296,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409840581827521e-05,
+      "loss": 4.2228,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409001987076469e-05,
+      "loss": 4.2199,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.2316,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407324797574365e-05,
+      "loss": 4.2199,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.406487840703686e-05,
+      "loss": 4.2172,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.405649245952634e-05,
+      "loss": 4.2266,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.404810651201582e-05,
+      "loss": 4.2251,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40397205645053e-05,
+      "loss": 4.232,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031334616994777e-05,
+      "loss": 4.229,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4022948669484257e-05,
+      "loss": 4.2274,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401456272197374e-05,
+      "loss": 4.2127,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400617677446322e-05,
+      "loss": 4.2227,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39977908269527e-05,
+      "loss": 4.2211,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398940487944218e-05,
+      "loss": 4.2076,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.2278,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397264936322487e-05,
+      "loss": 4.2105,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396426341571435e-05,
+      "loss": 4.2239,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395587746820383e-05,
+      "loss": 4.2171,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394750789949704e-05,
+      "loss": 4.2199,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393912195198652e-05,
+      "loss": 4.1988,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3930736004476e-05,
+      "loss": 4.2196,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392235005696548e-05,
+      "loss": 4.2158,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391396410945496e-05,
+      "loss": 4.2216,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390559454074818e-05,
+      "loss": 4.222,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389720859323766e-05,
+      "loss": 4.2268,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388882264572714e-05,
+      "loss": 4.2037,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388043669821662e-05,
+      "loss": 4.2205,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2278,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2257,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2197,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846909286978266e-05,
+      "loss": 4.2243,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838539718271475e-05,
+      "loss": 4.2167,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830153770760955e-05,
+      "loss": 4.2209,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.2017,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813381875739915e-05,
+      "loss": 4.2156,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.2084,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2109,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.2111,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.2109,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3771468516991044e-05,
+      "loss": 4.2102,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3763082569480524e-05,
+      "loss": 4.2088,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3754696621970004e-05,
+      "loss": 4.2205,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.202013969421387,
+      "eval_runtime": 284.1681,
+      "eval_samples_per_second": 1342.835,
+      "eval_steps_per_second": 41.965,
+      "step": 381600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.2062,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.1998,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2128,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372116921073165e-05,
+      "loss": 4.2115,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371278326322113e-05,
+      "loss": 4.2127,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370441369451435e-05,
+      "loss": 4.2021,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369602774700383e-05,
+      "loss": 4.2007,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368764179949331e-05,
+      "loss": 4.2024,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367925585198279e-05,
+      "loss": 4.206,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670886283276e-05,
+      "loss": 4.216,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.366250033576548e-05,
+      "loss": 4.204,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.365411438825496e-05,
+      "loss": 4.2159,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.364572844074444e-05,
+      "loss": 4.1941,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.363734249323392e-05,
+      "loss": 4.1992,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3628972924527126e-05,
+      "loss": 4.2062,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620586977016606e-05,
+      "loss": 4.1924,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361221740830982e-05,
+      "loss": 4.2009,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36038314607993e-05,
+      "loss": 4.2037,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359544551328878e-05,
+      "loss": 4.196,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358705956577826e-05,
+      "loss": 4.2245,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357867361826774e-05,
+      "loss": 4.2056,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357028767075722e-05,
+      "loss": 4.2022,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35619017232467e-05,
+      "loss": 4.2064,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355351577573618e-05,
+      "loss": 4.2151,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354512982822566e-05,
+      "loss": 4.1943,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353674388071514e-05,
+      "loss": 4.2038,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352835793320462e-05,
+      "loss": 4.1997,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35199719856941e-05,
+      "loss": 4.1973,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351160241698731e-05,
+      "loss": 4.1901,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350321646947679e-05,
+      "loss": 4.2038,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349483052196627e-05,
+      "loss": 4.2133,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348644457445575e-05,
+      "loss": 4.2025,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3478075005748966e-05,
+      "loss": 4.2051,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469705437042175e-05,
+      "loss": 4.2018,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461335868335385e-05,
+      "loss": 4.1992,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3452949920824865e-05,
+      "loss": 4.2032,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444563973314344e-05,
+      "loss": 4.1896,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436178025803824e-05,
+      "loss": 4.1919,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427792078293304e-05,
+      "loss": 4.1855,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419406130782784e-05,
+      "loss": 4.1912,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3411020183272264e-05,
+      "loss": 4.1955,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402634235761744e-05,
+      "loss": 4.1986,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394248288251224e-05,
+      "loss": 4.1891,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385862340740704e-05,
+      "loss": 4.194,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377476393230184e-05,
+      "loss": 4.1998,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3369090445719664e-05,
+      "loss": 4.195,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.336072087701287e-05,
+      "loss": 4.1979,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.335233492950235e-05,
+      "loss": 4.1919,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334394898199183e-05,
+      "loss": 4.1756,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333557941328505e-05,
+      "loss": 4.201,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332719346577452e-05,
+      "loss": 4.1876,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3318807518264e-05,
+      "loss": 4.1956,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331042157075348e-05,
+      "loss": 4.1835,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33020520020467e-05,
+      "loss": 4.1938,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329366605453618e-05,
+      "loss": 4.1757,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328528010702566e-05,
+      "loss": 4.2015,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3276910538318874e-05,
+      "loss": 4.1742,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326852459080835e-05,
+      "loss": 4.1867,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326013864329783e-05,
+      "loss": 4.1921,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325175269578731e-05,
+      "loss": 4.1803,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.324336674827679e-05,
+      "loss": 4.1797,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.323498080076627e-05,
+      "loss": 4.1934,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322659485325575e-05,
+      "loss": 4.1717,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321820890574523e-05,
+      "loss": 4.1702,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209822958234707e-05,
+      "loss": 4.186,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3201437010724187e-05,
+      "loss": 4.189,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3193051063213666e-05,
+      "loss": 4.175,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3184665115703146e-05,
+      "loss": 4.1859,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317631192580009e-05,
+      "loss": 4.1617,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316792597828957e-05,
+      "loss": 4.1789,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315954003077905e-05,
+      "loss": 4.185,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315115408326853e-05,
+      "loss": 4.1861,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314276813575801e-05,
+      "loss": 4.1906,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313438218824749e-05,
+      "loss": 4.1896,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312599624073697e-05,
+      "loss": 4.1995,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311761029322645e-05,
+      "loss": 4.1686,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310922434571593e-05,
+      "loss": 4.1831,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310087115581287e-05,
+      "loss": 4.1878,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.309248520830235e-05,
+      "loss": 4.1634,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308409926079183e-05,
+      "loss": 4.1853,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3075713313281316e-05,
+      "loss": 4.1795,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3067327365770796e-05,
+      "loss": 4.1867,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058957797064005e-05,
+      "loss": 4.1871,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050571849553485e-05,
+      "loss": 4.1706,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042185902042965e-05,
+      "loss": 4.1775,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033799954532445e-05,
+      "loss": 4.1779,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025414007021925e-05,
+      "loss": 4.1913,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3017028059511405e-05,
+      "loss": 4.177,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008642112000885e-05,
+      "loss": 4.1896,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300025616449036e-05,
+      "loss": 4.1888,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991886595783574e-05,
+      "loss": 4.1728,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983500648273054e-05,
+      "loss": 4.1821,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975114700762534e-05,
+      "loss": 4.1709,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296674513205575e-05,
+      "loss": 4.1816,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295835918454523e-05,
+      "loss": 4.1915,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294997323703471e-05,
+      "loss": 4.1695,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294158728952418e-05,
+      "loss": 4.1734,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293320134201366e-05,
+      "loss": 4.1695,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292483177330688e-05,
+      "loss": 4.1827,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291644582579636e-05,
+      "loss": 4.1908,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290805987828583e-05,
+      "loss": 4.1718,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289969030957905e-05,
+      "loss": 4.1698,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289130436206853e-05,
+      "loss": 4.1626,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288291841455801e-05,
+      "loss": 4.1846,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287453246704749e-05,
+      "loss": 4.1538,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286614651953697e-05,
+      "loss": 4.1709,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285776057202645e-05,
+      "loss": 4.1711,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284937462451593e-05,
+      "loss": 4.1745,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284098867700541e-05,
+      "loss": 4.1679,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283260272949489e-05,
+      "loss": 4.1751,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282421678198437e-05,
+      "loss": 4.1704,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281583083447385e-05,
+      "loss": 4.1647,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280744488696333e-05,
+      "loss": 4.1739,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2799075318256536e-05,
+      "loss": 4.1736,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790689370746016e-05,
+      "loss": 4.1786,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2782303423235496e-05,
+      "loss": 4.1795,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2773933854528705e-05,
+      "loss": 4.1719,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.276554790701819e-05,
+      "loss": 4.1627,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275716195950767e-05,
+      "loss": 4.1698,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274877601199715e-05,
+      "loss": 4.1715,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274039006448663e-05,
+      "loss": 4.1595,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273202049577984e-05,
+      "loss": 4.1723,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272363454826932e-05,
+      "loss": 4.1595,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27152486007588e-05,
+      "loss": 4.1711,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270686265324828e-05,
+      "loss": 4.1668,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269847670573776e-05,
+      "loss": 4.1668,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269009075822724e-05,
+      "loss": 4.1534,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2681704810716714e-05,
+      "loss": 4.1648,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2673318863206194e-05,
+      "loss": 4.1654,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2664932915695674e-05,
+      "loss": 4.1726,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265656334698889e-05,
+      "loss": 4.1667,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264817739947837e-05,
+      "loss": 4.1743,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263979145196785e-05,
+      "loss": 4.1559,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263140550445733e-05,
+      "loss": 4.1687,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2623035935750545e-05,
+      "loss": 4.1771,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261464998824002e-05,
+      "loss": 4.173,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26062640407295e-05,
+      "loss": 4.1712,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259787809321898e-05,
+      "loss": 4.1757,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258950852451219e-05,
+      "loss": 4.1634,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258112257700167e-05,
+      "loss": 4.1721,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.257273662949115e-05,
+      "loss": 4.154,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256435068198063e-05,
+      "loss": 4.1665,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255598111327384e-05,
+      "loss": 4.1582,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254759516576332e-05,
+      "loss": 4.16,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25392092182528e-05,
+      "loss": 4.1625,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253082327074228e-05,
+      "loss": 4.1581,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252243732323176e-05,
+      "loss": 4.1686,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251408413332871e-05,
+      "loss": 4.1513,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250569818581819e-05,
+      "loss": 4.174,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.160501956939697,
+      "eval_runtime": 290.4595,
+      "eval_samples_per_second": 1313.749,
+      "eval_steps_per_second": 41.056,
+      "step": 457920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249731223830766e-05,
+      "loss": 4.1538,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248892629079714e-05,
+      "loss": 4.1509,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248054034328662e-05,
+      "loss": 4.1652,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24721543957761e-05,
+      "loss": 4.1635,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246376844826558e-05,
+      "loss": 4.1646,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245538250075506e-05,
+      "loss": 4.1545,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244699655324455e-05,
+      "loss": 4.1478,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243861060573403e-05,
+      "loss": 4.156,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243022465822351e-05,
+      "loss": 4.1585,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242183871071299e-05,
+      "loss": 4.1664,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241345276320247e-05,
+      "loss": 4.1584,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240506681569195e-05,
+      "loss": 4.1641,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239668086818143e-05,
+      "loss": 4.1498,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238829492067091e-05,
+      "loss": 4.1511,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237990897316038e-05,
+      "loss": 4.1576,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237152302564986e-05,
+      "loss": 4.1435,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236313707813934e-05,
+      "loss": 4.1521,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235476750943255e-05,
+      "loss": 4.1572,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234638156192203e-05,
+      "loss": 4.1497,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337995614411516e-05,
+      "loss": 4.1736,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329609666900996e-05,
+      "loss": 4.1635,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321223719390476e-05,
+      "loss": 4.152,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312837771879956e-05,
+      "loss": 4.1597,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2304451824369436e-05,
+      "loss": 4.1701,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2296065876858916e-05,
+      "loss": 4.1446,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2287696308152125e-05,
+      "loss": 4.159,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2279326739445334e-05,
+      "loss": 4.149,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2270940791934814e-05,
+      "loss": 4.1508,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2262554844424294e-05,
+      "loss": 4.1453,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254168896913774e-05,
+      "loss": 4.1545,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245782949403254e-05,
+      "loss": 4.1627,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223741338069647e-05,
+      "loss": 4.1556,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222902743318595e-05,
+      "loss": 4.1607,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222064148567543e-05,
+      "loss": 4.1581,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221227191696864e-05,
+      "loss": 4.1485,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220390234826185e-05,
+      "loss": 4.1562,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219551640075133e-05,
+      "loss": 4.1463,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218713045324081e-05,
+      "loss": 4.144,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217874450573029e-05,
+      "loss": 4.1407,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217035855821977e-05,
+      "loss": 4.1431,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216197261070925e-05,
+      "loss": 4.1489,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215358666319873e-05,
+      "loss": 4.154,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214520071568821e-05,
+      "loss": 4.1449,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213681476817769e-05,
+      "loss": 4.1468,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212842882066717e-05,
+      "loss": 4.154,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212004287315665e-05,
+      "loss": 4.1484,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111656925646134e-05,
+      "loss": 4.1534,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.1459,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094901409428823e-05,
+      "loss": 4.127,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20865154619183e-05,
+      "loss": 4.1582,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207812951440778e-05,
+      "loss": 4.1416,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206974356689726e-05,
+      "loss": 4.153,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061357619386736e-05,
+      "loss": 4.1425,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052971671876216e-05,
+      "loss": 4.1428,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204460210316943e-05,
+      "loss": 4.1371,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.203621615565891e-05,
+      "loss": 4.1491,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027830208148385e-05,
+      "loss": 4.1342,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201944426063787e-05,
+      "loss": 4.142,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201105831312735e-05,
+      "loss": 4.1526,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200267236561683e-05,
+      "loss": 4.1315,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199428641810631e-05,
+      "loss": 4.1339,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198590047059579e-05,
+      "loss": 4.1523,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1977530901889e-05,
+      "loss": 4.1226,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196914495437848e-05,
+      "loss": 4.126,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196075900686796e-05,
+      "loss": 4.1405,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195237305935744e-05,
+      "loss": 4.144,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194398711184692e-05,
+      "loss": 4.1333,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193561754314013e-05,
+      "loss": 4.1446,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192723159562961e-05,
+      "loss": 4.1202,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1918862026922826e-05,
+      "loss": 4.1352,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1910476079412306e-05,
+      "loss": 4.143,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1902090131901786e-05,
+      "loss": 4.1414,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1893704184391266e-05,
+      "loss": 4.1466,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1477,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876948668173955e-05,
+      "loss": 4.1539,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1868579099467164e-05,
+      "loss": 4.1278,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1860193151956644e-05,
+      "loss": 4.1389,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185182358324986e-05,
+      "loss": 4.1442,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184343763573933e-05,
+      "loss": 4.1222,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183505168822881e-05,
+      "loss": 4.1412,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182666574071829e-05,
+      "loss": 4.1371,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.181827979320778e-05,
+      "loss": 4.1453,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180989384569726e-05,
+      "loss": 4.1428,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180150789818674e-05,
+      "loss": 4.1299,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179312195067622e-05,
+      "loss": 4.1323,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.17847360031657e-05,
+      "loss": 4.1381,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177635005565518e-05,
+      "loss": 4.1431,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176796410814466e-05,
+      "loss": 4.1393,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175957816063414e-05,
+      "loss": 4.1463,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175119221312362e-05,
+      "loss": 4.1455,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.17428062656131e-05,
+      "loss": 4.1337,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173443669690631e-05,
+      "loss": 4.1372,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172605074939579e-05,
+      "loss": 4.1257,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171766480188527e-05,
+      "loss": 4.1459,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170927885437475e-05,
+      "loss": 4.1454,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170089290686423e-05,
+      "loss": 4.131,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169250695935371e-05,
+      "loss": 4.1298,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.168412101184319e-05,
+      "loss": 4.1257,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167573506433267e-05,
+      "loss": 4.141,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166734911682215e-05,
+      "loss": 4.1474,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165897954811536e-05,
+      "loss": 4.1297,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165060997940857e-05,
+      "loss": 4.1294,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642224031898046e-05,
+      "loss": 4.1262,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1633838084387526e-05,
+      "loss": 4.1383,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625452136877006e-05,
+      "loss": 4.1167,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1617066189366486e-05,
+      "loss": 4.1281,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608680241855966e-05,
+      "loss": 4.1225,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160031067314918e-05,
+      "loss": 4.1384,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159192472563866e-05,
+      "loss": 4.1274,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158353877812814e-05,
+      "loss": 4.1363,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157515283061762e-05,
+      "loss": 4.129,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15667668831071e-05,
+      "loss": 4.1218,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155838093559658e-05,
+      "loss": 4.1345,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155001136688979e-05,
+      "loss": 4.1323,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154162541937927e-05,
+      "loss": 4.1365,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153323947186875e-05,
+      "loss": 4.1375,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152486990316196e-05,
+      "loss": 4.1261,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151648395565144e-05,
+      "loss": 4.1294,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150809800814092e-05,
+      "loss": 4.1225,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14997120606304e-05,
+      "loss": 4.1318,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149132611311988e-05,
+      "loss": 4.1273,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1482940165609366e-05,
+      "loss": 4.1276,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1474554218098846e-05,
+      "loss": 4.115,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1466168270588326e-05,
+      "loss": 4.1298,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457782323077806e-05,
+      "loss": 4.1247,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449396375567286e-05,
+      "loss": 4.1295,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144101042805676e-05,
+      "loss": 4.1115,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.143262448054624e-05,
+      "loss": 4.1287,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424254911839455e-05,
+      "loss": 4.1228,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415885343132664e-05,
+      "loss": 4.137,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140751577442587e-05,
+      "loss": 4.1263,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139912982691535e-05,
+      "loss": 4.1335,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139074387940483e-05,
+      "loss": 4.1156,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138235793189432e-05,
+      "loss": 4.1312,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137398836318753e-05,
+      "loss": 4.1358,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136560241567701e-05,
+      "loss": 4.1326,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135721646816649e-05,
+      "loss": 4.1314,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134883052065597e-05,
+      "loss": 4.1358,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134044457314545e-05,
+      "loss": 4.1243,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133205862563493e-05,
+      "loss": 4.1358,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132367267812441e-05,
+      "loss": 4.1152,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.131528673061388e-05,
+      "loss": 4.1276,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1196,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129851483559284e-05,
+      "loss": 4.123,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129012888808232e-05,
+      "loss": 4.1189,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128175931937554e-05,
+      "loss": 4.1162,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127337337186502e-05,
+      "loss": 4.1278,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.12649874243545e-05,
+      "loss": 4.1159,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125660147684398e-05,
+      "loss": 4.137,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.129661560058594,
+      "eval_runtime": 300.887,
+      "eval_samples_per_second": 1268.22,
+      "eval_steps_per_second": 39.633,
+      "step": 534240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124821552933346e-05,
+      "loss": 4.1159,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123982958182294e-05,
+      "loss": 4.1115,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123144363431242e-05,
+      "loss": 4.1226,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.12230576868019e-05,
+      "loss": 4.123,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1214688118095106e-05,
+      "loss": 4.1218,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1206318549388315e-05,
+      "loss": 4.1192,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1197932601877795e-05,
+      "loss": 4.1127,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1189546654367275e-05,
+      "loss": 4.1161,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1181160706856755e-05,
+      "loss": 4.1217,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117277475934624e-05,
+      "loss": 4.1281,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.116440519063945e-05,
+      "loss": 4.1229,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115601924312893e-05,
+      "loss": 4.1204,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114763329561841e-05,
+      "loss": 4.1166,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113924734810789e-05,
+      "loss": 4.1112,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113086140059737e-05,
+      "loss": 4.1185,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.112247545308685e-05,
+      "loss": 4.1066,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111408950557633e-05,
+      "loss": 4.1155,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110571993686954e-05,
+      "loss": 4.1178,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109733398935902e-05,
+      "loss": 4.1107,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10889480418485e-05,
+      "loss": 4.1334,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108056209433798e-05,
+      "loss": 4.1285,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107217614682746e-05,
+      "loss": 4.1153,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106379019931694e-05,
+      "loss": 4.1186,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105540425180642e-05,
+      "loss": 4.1311,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10470183042959e-05,
+      "loss": 4.1082,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038665114392845e-05,
+      "loss": 4.1215,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030279166882325e-05,
+      "loss": 4.1139,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021893219371805e-05,
+      "loss": 4.1127,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1013507271861285e-05,
+      "loss": 4.107,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1005137703154494e-05,
+      "loss": 4.1216,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0996751755643974e-05,
+      "loss": 4.1216,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0988365808133454e-05,
+      "loss": 4.1208,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0979979860622933e-05,
+      "loss": 4.1233,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0971593913112413e-05,
+      "loss": 4.1145,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096320796560189e-05,
+      "loss": 4.1128,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095482201809137e-05,
+      "loss": 4.1206,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094643607058085e-05,
+      "loss": 4.1113,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093805012307033e-05,
+      "loss": 4.1017,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092966417555981e-05,
+      "loss": 4.1056,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092127822804929e-05,
+      "loss": 4.1063,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091289228053877e-05,
+      "loss": 4.1104,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.090452271183198e-05,
+      "loss": 4.1171,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089613676432146e-05,
+      "loss": 4.1099,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088775081681094e-05,
+      "loss": 4.1083,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087936486930042e-05,
+      "loss": 4.1185,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.087099530059363e-05,
+      "loss": 4.1138,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.086260935308311e-05,
+      "loss": 4.1161,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.08542234055726e-05,
+      "loss": 4.1123,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.084583745806208e-05,
+      "loss": 4.0848,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083745151055156e-05,
+      "loss": 4.125,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082908194184477e-05,
+      "loss": 4.1034,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.082069599433425e-05,
+      "loss": 4.1174,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081231004682373e-05,
+      "loss": 4.1037,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080392409931321e-05,
+      "loss": 4.1087,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079553815180269e-05,
+      "loss": 4.1014,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078715220429217e-05,
+      "loss": 4.1127,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0778766256781647e-05,
+      "loss": 4.1002,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0770380309271127e-05,
+      "loss": 4.1042,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0762010740564336e-05,
+      "loss": 4.1167,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0753624793053816e-05,
+      "loss": 4.0993,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0745238845543296e-05,
+      "loss": 4.0959,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073686927683651e-05,
+      "loss": 4.1149,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072848332932599e-05,
+      "loss": 4.0914,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.072009738181547e-05,
+      "loss": 4.0896,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071171143430495e-05,
+      "loss": 4.106,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070334186559816e-05,
+      "loss": 4.1075,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069495591808764e-05,
+      "loss": 4.0985,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.068656997057712e-05,
+      "loss": 4.1074,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06781840230666e-05,
+      "loss": 4.0827,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066979807555608e-05,
+      "loss": 4.1005,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0661412128045553e-05,
+      "loss": 4.1087,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065302618053503e-05,
+      "loss": 4.1051,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064464023302452e-05,
+      "loss": 4.113,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636254285514e-05,
+      "loss": 4.1102,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.062788471680721e-05,
+      "loss": 4.1187,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619515148100425e-05,
+      "loss": 4.0912,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0611129200589905e-05,
+      "loss": 4.107,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.060274325307938e-05,
+      "loss": 4.1065,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.059435730556886e-05,
+      "loss": 4.0915,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.058597135805834e-05,
+      "loss": 4.1015,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577601789351554e-05,
+      "loss": 4.1046,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056921584184103e-05,
+      "loss": 4.1074,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056082989433051e-05,
+      "loss": 4.1134,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055244394681999e-05,
+      "loss": 4.092,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0544057999309474e-05,
+      "loss": 4.0967,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0535672051798954e-05,
+      "loss": 4.1035,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.052730248309216e-05,
+      "loss": 4.1079,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051891653558164e-05,
+      "loss": 4.1065,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051053058807112e-05,
+      "loss": 4.1077,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.05021446405606e-05,
+      "loss": 4.1097,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.049375869305008e-05,
+      "loss": 4.1056,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048538912434329e-05,
+      "loss": 4.1014,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047700317683277e-05,
+      "loss": 4.0945,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046861722932225e-05,
+      "loss": 4.1092,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046023128181173e-05,
+      "loss": 4.1111,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.045186171310494e-05,
+      "loss": 4.1004,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044347576559443e-05,
+      "loss": 4.0931,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043508981808391e-05,
+      "loss": 4.0922,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042670387057339e-05,
+      "loss": 4.1042,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0418334301866596e-05,
+      "loss": 4.1159,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0409948354356076e-05,
+      "loss": 4.0967,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0401562406845556e-05,
+      "loss": 4.0978,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0393176459335036e-05,
+      "loss": 4.0943,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0384790511824516e-05,
+      "loss": 4.0994,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0376420943117725e-05,
+      "loss": 4.0866,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0368034995607205e-05,
+      "loss": 4.0966,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359649048096685e-05,
+      "loss": 4.0843,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351263100586165e-05,
+      "loss": 4.1022,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342877153075645e-05,
+      "loss": 4.0969,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033450758436886e-05,
+      "loss": 4.0991,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032613801566207e-05,
+      "loss": 4.0958,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.031775206815155e-05,
+      "loss": 4.0959,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030936612064103e-05,
+      "loss": 4.096,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030098017313051e-05,
+      "loss": 4.0982,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029259422561999e-05,
+      "loss": 4.1015,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028420827810947e-05,
+      "loss": 4.1052,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027582233059895e-05,
+      "loss": 4.0933,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026743638308843e-05,
+      "loss": 4.0988,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025905043557791e-05,
+      "loss": 4.0914,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025066448806738e-05,
+      "loss": 4.0965,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.02422949193606e-05,
+      "loss": 4.0962,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0233925350653815e-05,
+      "loss": 4.0938,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0225539403143295e-05,
+      "loss": 4.0854,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0217153455632775e-05,
+      "loss": 4.0944,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0208767508122255e-05,
+      "loss": 4.0975,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200381560611735e-05,
+      "loss": 4.0933,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0191995613101214e-05,
+      "loss": 4.0751,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.018360966559069e-05,
+      "loss": 4.1002,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.017522371808017e-05,
+      "loss": 4.0879,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.016683777056965e-05,
+      "loss": 4.1027,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015845182305913e-05,
+      "loss": 4.0949,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.015006587554861e-05,
+      "loss": 4.1001,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0141696306841817e-05,
+      "loss": 4.0851,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0133310359331297e-05,
+      "loss": 4.1004,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.012492441182078e-05,
+      "loss": 4.1021,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.011653846431026e-05,
+      "loss": 4.1,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.010815251679974e-05,
+      "loss": 4.1003,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009976656928922e-05,
+      "loss": 4.1003,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.009139700058243e-05,
+      "loss": 4.0935,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.008301105307191e-05,
+      "loss": 4.1013,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.007462510556139e-05,
+      "loss": 4.082,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006623915805087e-05,
+      "loss": 4.0977,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005785321054035e-05,
+      "loss": 4.0855,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004946726302983e-05,
+      "loss": 4.0945,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004108131551931e-05,
+      "loss": 4.0833,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003269536800879e-05,
+      "loss": 4.0861,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0024325799302e-05,
+      "loss": 4.0918,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001595623059522e-05,
+      "loss": 4.0882,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.00075702830847e-05,
+      "loss": 4.1041,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.105152130126953,
+      "eval_runtime": 293.6825,
+      "eval_samples_per_second": 1299.332,
+      "eval_steps_per_second": 40.605,
+      "step": 610560
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.506204051808901e+17,
+  "trial_name": null,
+  "trial_params": null
+}