diff --git "a/checkpoint-457914/trainer_state.json" "b/checkpoint-457914/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457914/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 4.168251037597656,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/1/checkpoints/checkpoint-457914",
+  "epoch": 0.025000278439663435,
+  "eval_steps": 10,
+  "global_step": 457914,
+  "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": 10.8213,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5645,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0515,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9831,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9548,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8802,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.736,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6253,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5411,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4495,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99161405248948e-05,
+      "loss": 6.4027,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990775457738428e-05,
+      "loss": 6.3385,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989936862987376e-05,
+      "loss": 6.2752,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989098268236324e-05,
+      "loss": 6.2086,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988259673485272e-05,
+      "loss": 6.1566,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98742107873422e-05,
+      "loss": 6.1122,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986582483983168e-05,
+      "loss": 6.0644,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985743889232116e-05,
+      "loss": 6.0115,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984905294481064e-05,
+      "loss": 5.9764,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984066699730012e-05,
+      "loss": 5.9421,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9832281049789595e-05,
+      "loss": 5.8975,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8621,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8402,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.8026,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.979875363855125e-05,
+      "loss": 5.7808,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 5.754,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 5.7313,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.7017,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6683,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6615,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6285,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6122,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9731715194878284e-05,
+      "loss": 5.586,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9723329247367764e-05,
+      "loss": 5.5823,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5499,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5409,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5269,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.5069,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4964,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4737,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4496,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4441,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4158,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.4276,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.4002,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3954,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96143610661422e-05,
+      "loss": 5.3797,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.3764,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3695,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3413,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3289,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957244770739333e-05,
+      "loss": 5.3212,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956406175988281e-05,
+      "loss": 5.3278,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9555692191176016e-05,
+      "loss": 5.3042,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9547306243665496e-05,
+      "loss": 5.3119,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9538920296154976e-05,
+      "loss": 5.2815,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530534348644456e-05,
+      "loss": 5.2791,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522148401133936e-05,
+      "loss": 5.2658,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951376245362342e-05,
+      "loss": 5.2593,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.2565,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2359,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.2316,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948023504238507e-05,
+      "loss": 5.2351,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947184909487455e-05,
+      "loss": 5.2133,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.946346314736403e-05,
+      "loss": 5.2125,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945509357865724e-05,
+      "loss": 5.2008,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944670763114672e-05,
+      "loss": 5.2003,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94383216836362e-05,
+      "loss": 5.1851,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942993573612568e-05,
+      "loss": 5.1879,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942154978861516e-05,
+      "loss": 5.1751,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1642,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1667,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1528,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9388022377376816e-05,
+      "loss": 5.1613,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1582,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371266861159505e-05,
+      "loss": 5.1489,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362880913648985e-05,
+      "loss": 5.1294,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1139,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1069,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.112,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1168,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.1102,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0849,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0786,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0875,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0727,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279021438543785e-05,
+      "loss": 5.0728,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0653,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0665,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.0551,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.047,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0335,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.047,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0329,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0162,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0157,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 5.0278,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 5.0184,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 5.0099,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.9974,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 5.0021,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 4.9813,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 4.9968,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 4.9991,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 4.9737,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911975395105883e-05,
+      "loss": 4.9676,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.9428,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910298205603779e-05,
+      "loss": 4.9555,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909459610852727e-05,
+      "loss": 4.9556,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9086226539820476e-05,
+      "loss": 4.9508,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9509,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9356,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906110145489638e-05,
+      "loss": 4.9361,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.937,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904432955987534e-05,
+      "loss": 4.9393,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903594361236482e-05,
+      "loss": 4.9269,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.90275576648543e-05,
+      "loss": 4.9157,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901917171734378e-05,
+      "loss": 4.9165,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901078576983326e-05,
+      "loss": 4.92,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900239982232274e-05,
+      "loss": 4.9154,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.91,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8992,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8977274737398646e-05,
+      "loss": 4.8913,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968888789888125e-05,
+      "loss": 4.8854,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.8927,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.8904,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8911,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8653,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8793,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8787,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.865,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8807,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8592,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8725,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8632,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8408,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8402,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8403,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.844,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8483,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882637681862047e-05,
+      "loss": 4.8431,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8381,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8284,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8163,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8157,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.8223,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.8298,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8215,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875930561734005e-05,
+      "loss": 4.804,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.8036,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.769171237945557,
+      "eval_runtime": 589.0702,
+      "eval_samples_per_second": 647.785,
+      "eval_steps_per_second": 20.244,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874253372231901e-05,
+      "loss": 4.8109,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873414777480849e-05,
+      "loss": 4.8026,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.872576182729797e-05,
+      "loss": 4.7949,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871737587978745e-05,
+      "loss": 4.7949,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870898993227693e-05,
+      "loss": 4.7846,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87006039847664e-05,
+      "loss": 4.7696,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869221803725588e-05,
+      "loss": 4.7767,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.868383208974536e-05,
+      "loss": 4.7869,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867544614223484e-05,
+      "loss": 4.7699,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8667076573528056e-05,
+      "loss": 4.7771,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8658690626017536e-05,
+      "loss": 4.7817,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8650304678507016e-05,
+      "loss": 4.7796,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8641918730996496e-05,
+      "loss": 4.7685,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8633532783485976e-05,
+      "loss": 4.7648,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8625146835975456e-05,
+      "loss": 4.7622,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8616760888464936e-05,
+      "loss": 4.7531,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8608374940954416e-05,
+      "loss": 4.7559,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8599988993443896e-05,
+      "loss": 4.7563,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8591603045933376e-05,
+      "loss": 4.7553,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8583217098422856e-05,
+      "loss": 4.7532,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8574831150912336e-05,
+      "loss": 4.7413,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8566445203401816e-05,
+      "loss": 4.7521,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.855805925589129e-05,
+      "loss": 4.7377,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854967330838077e-05,
+      "loss": 4.7476,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541287360870256e-05,
+      "loss": 4.737,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85329341709672e-05,
+      "loss": 4.7359,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.852454822345668e-05,
+      "loss": 4.7376,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.851616227594616e-05,
+      "loss": 4.7198,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850777632843564e-05,
+      "loss": 4.7336,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8499390380925114e-05,
+      "loss": 4.714,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8491004433414594e-05,
+      "loss": 4.7198,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8482618485904074e-05,
+      "loss": 4.7157,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.847424891719729e-05,
+      "loss": 4.7243,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84658793484905e-05,
+      "loss": 4.7119,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.845749340097998e-05,
+      "loss": 4.7053,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844910745346946e-05,
+      "loss": 4.7219,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.844072150595894e-05,
+      "loss": 4.7076,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843233555844842e-05,
+      "loss": 4.7151,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.84239496109379e-05,
+      "loss": 4.7019,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841556366342738e-05,
+      "loss": 4.6875,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840717771591686e-05,
+      "loss": 4.6919,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839880814721007e-05,
+      "loss": 4.6831,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839042219969955e-05,
+      "loss": 4.6987,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.838203625218903e-05,
+      "loss": 4.6872,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837365030467851e-05,
+      "loss": 4.6933,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836526435716799e-05,
+      "loss": 4.6852,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8356894788461196e-05,
+      "loss": 4.6899,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8348508840950676e-05,
+      "loss": 4.6865,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8340122893440156e-05,
+      "loss": 4.6644,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833173694592964e-05,
+      "loss": 4.668,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832336737722285e-05,
+      "loss": 4.6661,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.831498142971233e-05,
+      "loss": 4.6784,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.830659548220181e-05,
+      "loss": 4.6562,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829820953469129e-05,
+      "loss": 4.6805,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828982358718077e-05,
+      "loss": 4.6559,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828143763967025e-05,
+      "loss": 4.6552,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.827305169215973e-05,
+      "loss": 4.6579,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.826466574464921e-05,
+      "loss": 4.6531,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.825627979713869e-05,
+      "loss": 4.6541,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.824792660723563e-05,
+      "loss": 4.6459,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.823954065972511e-05,
+      "loss": 4.6447,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231171091018326e-05,
+      "loss": 4.6548,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8222785143507806e-05,
+      "loss": 4.6412,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8214399195997286e-05,
+      "loss": 4.6403,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8206013248486766e-05,
+      "loss": 4.6374,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8197627300976246e-05,
+      "loss": 4.643,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8189241353465726e-05,
+      "loss": 4.6266,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8180855405955206e-05,
+      "loss": 4.6451,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8172485837248415e-05,
+      "loss": 4.6328,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8164099889737895e-05,
+      "loss": 4.6304,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8155713942227375e-05,
+      "loss": 4.6366,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8147327994716854e-05,
+      "loss": 4.6245,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8138942047206334e-05,
+      "loss": 4.6417,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8130556099695814e-05,
+      "loss": 4.6454,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8122170152185294e-05,
+      "loss": 4.6396,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8113784204674774e-05,
+      "loss": 4.6219,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8105398257164254e-05,
+      "loss": 4.6139,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8097012309653734e-05,
+      "loss": 4.6101,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808864274094695e-05,
+      "loss": 4.6203,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.808027317224016e-05,
+      "loss": 4.6318,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.807188722472964e-05,
+      "loss": 4.6234,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.806350127721912e-05,
+      "loss": 4.6095,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.80551153297086e-05,
+      "loss": 4.5985,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.804672938219807e-05,
+      "loss": 4.6168,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.803834343468755e-05,
+      "loss": 4.6017,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802995748717703e-05,
+      "loss": 4.6049,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.802157153966652e-05,
+      "loss": 4.6079,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801320197095973e-05,
+      "loss": 4.6147,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800481602344921e-05,
+      "loss": 4.6023,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799643007593869e-05,
+      "loss": 4.5988,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798804412842817e-05,
+      "loss": 4.5903,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797965818091765e-05,
+      "loss": 4.6031,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797128861221086e-05,
+      "loss": 4.601,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796290266470034e-05,
+      "loss": 4.5775,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795451671718982e-05,
+      "loss": 4.5858,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.79461307696793e-05,
+      "loss": 4.6014,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793774482216878e-05,
+      "loss": 4.6017,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792935887465826e-05,
+      "loss": 4.588,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792098930595147e-05,
+      "loss": 4.5824,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791260335844095e-05,
+      "loss": 4.5923,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.790421741093043e-05,
+      "loss": 4.5731,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789583146341991e-05,
+      "loss": 4.5881,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788744551590939e-05,
+      "loss": 4.5934,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.78790759472026e-05,
+      "loss": 4.5784,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787068999969208e-05,
+      "loss": 4.5772,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786230405218156e-05,
+      "loss": 4.5558,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.785393448347477e-05,
+      "loss": 4.5621,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.784554853596425e-05,
+      "loss": 4.5681,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.783716258845373e-05,
+      "loss": 4.5696,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782877664094321e-05,
+      "loss": 4.5729,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782039069343269e-05,
+      "loss": 4.5607,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.781200474592217e-05,
+      "loss": 4.5587,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780361879841165e-05,
+      "loss": 4.5656,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.779523285090114e-05,
+      "loss": 4.5725,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778684690339061e-05,
+      "loss": 4.5609,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7778477334683826e-05,
+      "loss": 4.5613,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7770091387173306e-05,
+      "loss": 4.5575,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7761705439662786e-05,
+      "loss": 4.561,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7753335870955995e-05,
+      "loss": 4.5606,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7744949923445475e-05,
+      "loss": 4.5546,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7736563975934955e-05,
+      "loss": 4.5515,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7728178028424435e-05,
+      "loss": 4.5472,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771979208091391e-05,
+      "loss": 4.5424,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771140613340339e-05,
+      "loss": 4.5481,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7703020185892875e-05,
+      "loss": 4.5508,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7694634238382355e-05,
+      "loss": 4.5538,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686248290871835e-05,
+      "loss": 4.5356,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7677862343361315e-05,
+      "loss": 4.5436,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669476395850795e-05,
+      "loss": 4.5556,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661106827144004e-05,
+      "loss": 4.5395,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.765273725843721e-05,
+      "loss": 4.5519,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.764435131092669e-05,
+      "loss": 4.5363,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.763596536341617e-05,
+      "loss": 4.5547,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.762759579470938e-05,
+      "loss": 4.5418,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761920984719886e-05,
+      "loss": 4.5279,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761082389968834e-05,
+      "loss": 4.5299,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760243795217783e-05,
+      "loss": 4.525,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759405200466731e-05,
+      "loss": 4.5383,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758566605715679e-05,
+      "loss": 4.5384,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757728010964627e-05,
+      "loss": 4.5395,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756889416213575e-05,
+      "loss": 4.5332,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756050821462523e-05,
+      "loss": 4.5258,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755212226711471e-05,
+      "loss": 4.5207,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754373631960419e-05,
+      "loss": 4.5162,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.75353667508974e-05,
+      "loss": 4.5269,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752698080338688e-05,
+      "loss": 4.5398,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751859485587636e-05,
+      "loss": 4.5253,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751020890836584e-05,
+      "loss": 4.516,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.750182296085532e-05,
+      "loss": 4.5156,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.485103130340576,
+      "eval_runtime": 527.7119,
+      "eval_samples_per_second": 723.105,
+      "eval_steps_per_second": 22.598,
+      "step": 152638
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.74934370133448e-05,
+      "loss": 4.5263,
+      "step": 153088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.748505106583428e-05,
+      "loss": 4.5219,
+      "step": 153600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.747668149712749e-05,
+      "loss": 4.5095,
+      "step": 154112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.74683119284207e-05,
+      "loss": 4.5124,
+      "step": 154624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745992598091018e-05,
+      "loss": 4.5079,
+      "step": 155136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.745154003339966e-05,
+      "loss": 4.4944,
+      "step": 155648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.744315408588914e-05,
+      "loss": 4.5033,
+      "step": 156160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.5144,
+      "step": 156672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.4941,
+      "step": 157184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.741802900096504e-05,
+      "loss": 4.5058,
+      "step": 157696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.740964305345452e-05,
+      "loss": 4.5157,
+      "step": 158208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7401257105944e-05,
+      "loss": 4.5131,
+      "step": 158720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.5003,
+      "step": 159232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.5014,
+      "step": 159744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.4991,
+      "step": 160256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4937,
+      "step": 160768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4923,
+      "step": 161280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4985,
+      "step": 161792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.4993,
+      "step": 162304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.733416952585984e-05,
+      "loss": 4.4986,
+      "step": 162816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.732578357834932e-05,
+      "loss": 4.4894,
+      "step": 163328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.73173976308388e-05,
+      "loss": 4.5022,
+      "step": 163840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4888,
+      "step": 164352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4955,
+      "step": 164864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4897,
+      "step": 165376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4892,
+      "step": 165888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.494,
+      "step": 166400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7267098324579415e-05,
+      "loss": 4.4764,
+      "step": 166912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4913,
+      "step": 167424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4761,
+      "step": 167936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4792,
+      "step": 168448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4798,
+      "step": 168960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4851,
+      "step": 169472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.721681539712375e-05,
+      "loss": 4.4783,
+      "step": 169984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4698,
+      "step": 170496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4887,
+      "step": 171008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.719165755459219e-05,
+      "loss": 4.4752,
+      "step": 171520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.718327160708167e-05,
+      "loss": 4.4876,
+      "step": 172032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.717488565957115e-05,
+      "loss": 4.4736,
+      "step": 172544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.716649971206063e-05,
+      "loss": 4.4606,
+      "step": 173056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.715811376455011e-05,
+      "loss": 4.4608,
+      "step": 173568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714974419584333e-05,
+      "loss": 4.4621,
+      "step": 174080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.714137462713654e-05,
+      "loss": 4.4806,
+      "step": 174592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.713298867962602e-05,
+      "loss": 4.4587,
+      "step": 175104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.71246027321155e-05,
+      "loss": 4.4722,
+      "step": 175616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.711621678460498e-05,
+      "loss": 4.4669,
+      "step": 176128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4653,
+      "step": 176640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.4728,
+      "step": 177152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.709105894207341e-05,
+      "loss": 4.4475,
+      "step": 177664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4513,
+      "step": 178176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7074303425856106e-05,
+      "loss": 4.4476,
+      "step": 178688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7065917478345586e-05,
+      "loss": 4.4672,
+      "step": 179200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7057531530835066e-05,
+      "loss": 4.4459,
+      "step": 179712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7049145583324546e-05,
+      "loss": 4.4699,
+      "step": 180224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7040776014617755e-05,
+      "loss": 4.4456,
+      "step": 180736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7032390067107235e-05,
+      "loss": 4.4458,
+      "step": 181248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.4455,
+      "step": 181760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4435,
+      "step": 182272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4497,
+      "step": 182784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4447,
+      "step": 183296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4335,
+      "step": 183808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4508,
+      "step": 184320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.438,
+      "step": 184832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4411,
+      "step": 185344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.438,
+      "step": 185856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.4436,
+      "step": 186368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4244,
+      "step": 186880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4494,
+      "step": 187392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.692340550707794e-05,
+      "loss": 4.4339,
+      "step": 187904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.691501955956742e-05,
+      "loss": 4.4373,
+      "step": 188416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.69066336120569e-05,
+      "loss": 4.4395,
+      "step": 188928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.689826404335011e-05,
+      "loss": 4.4292,
+      "step": 189440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.688987809583959e-05,
+      "loss": 4.4479,
+      "step": 189952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4521,
+      "step": 190464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4497,
+      "step": 190976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6864753010915494e-05,
+      "loss": 4.4338,
+      "step": 191488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6856367063404974e-05,
+      "loss": 4.4249,
+      "step": 192000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6847981115894454e-05,
+      "loss": 4.4181,
+      "step": 192512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6839595168383934e-05,
+      "loss": 4.4363,
+      "step": 193024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6831209220873414e-05,
+      "loss": 4.4439,
+      "step": 193536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6822823273362894e-05,
+      "loss": 4.4378,
+      "step": 194048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6814437325852373e-05,
+      "loss": 4.4247,
+      "step": 194560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6806051378341853e-05,
+      "loss": 4.4157,
+      "step": 195072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.6797665430831333e-05,
+      "loss": 4.4293,
+      "step": 195584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678927948332081e-05,
+      "loss": 4.4159,
+      "step": 196096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.421,
+      "step": 196608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.67725239671035e-05,
+      "loss": 4.429,
+      "step": 197120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.676413801959298e-05,
+      "loss": 4.4286,
+      "step": 197632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.675575207208246e-05,
+      "loss": 4.4235,
+      "step": 198144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.674736612457194e-05,
+      "loss": 4.4243,
+      "step": 198656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4129,
+      "step": 199168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4243,
+      "step": 199680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.4212,
+      "step": 200192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.671383871333359e-05,
+      "loss": 4.406,
+      "step": 200704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.670546914462681e-05,
+      "loss": 4.4093,
+      "step": 201216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.669708319711629e-05,
+      "loss": 4.4255,
+      "step": 201728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668869724960576e-05,
+      "loss": 4.4273,
+      "step": 202240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.668031130209524e-05,
+      "loss": 4.4122,
+      "step": 202752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6671941733388456e-05,
+      "loss": 4.4134,
+      "step": 203264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.4193,
+      "step": 203776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6655169838367416e-05,
+      "loss": 4.4022,
+      "step": 204288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6646783890856896e-05,
+      "loss": 4.4186,
+      "step": 204800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6638397943346376e-05,
+      "loss": 4.4239,
+      "step": 205312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.4078,
+      "step": 205824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.4144,
+      "step": 206336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.3884,
+      "step": 206848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.3918,
+      "step": 207360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.4015,
+      "step": 207872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6588098637086985e-05,
+      "loss": 4.4042,
+      "step": 208384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.4107,
+      "step": 208896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6571326742065945e-05,
+      "loss": 4.3957,
+      "step": 209408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6562957173359154e-05,
+      "loss": 4.3936,
+      "step": 209920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.655458760465237e-05,
+      "loss": 4.3993,
+      "step": 210432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.654620165714185e-05,
+      "loss": 4.4114,
+      "step": 210944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.4021,
+      "step": 211456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.4048,
+      "step": 211968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.3946,
+      "step": 212480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.3957,
+      "step": 212992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.650428829839298e-05,
+      "loss": 4.4068,
+      "step": 213504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.649590235088246e-05,
+      "loss": 4.3947,
+      "step": 214016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.648751640337194e-05,
+      "loss": 4.3941,
+      "step": 214528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.647913045586142e-05,
+      "loss": 4.3924,
+      "step": 215040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.64707445083509e-05,
+      "loss": 4.3842,
+      "step": 215552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.646235856084038e-05,
+      "loss": 4.3928,
+      "step": 216064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.645397261332986e-05,
+      "loss": 4.3928,
+      "step": 216576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.644558666581934e-05,
+      "loss": 4.4017,
+      "step": 217088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.3809,
+      "step": 217600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.6428831149602034e-05,
+      "loss": 4.3875,
+      "step": 218112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.4029,
+      "step": 218624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.385,
+      "step": 219136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.4004,
+      "step": 219648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.3869,
+      "step": 220160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.638691779085316e-05,
+      "loss": 4.4033,
+      "step": 220672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637853184334264e-05,
+      "loss": 4.3887,
+      "step": 221184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.637014589583212e-05,
+      "loss": 4.3809,
+      "step": 221696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.636179270592906e-05,
+      "loss": 4.3798,
+      "step": 222208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.635340675841854e-05,
+      "loss": 4.3751,
+      "step": 222720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.634502081090802e-05,
+      "loss": 4.3877,
+      "step": 223232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.3948,
+      "step": 223744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.632824891588699e-05,
+      "loss": 4.3892,
+      "step": 224256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.3858,
+      "step": 224768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.3795,
+      "step": 225280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.630309107335542e-05,
+      "loss": 4.3773,
+      "step": 225792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.62947051258449e-05,
+      "loss": 4.3714,
+      "step": 226304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.628631917833438e-05,
+      "loss": 4.384,
+      "step": 226816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.3931,
+      "step": 227328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626956366211707e-05,
+      "loss": 4.3819,
+      "step": 227840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.626117771460655e-05,
+      "loss": 4.3746,
+      "step": 228352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.3728,
+      "step": 228864
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.3495965003967285,
+      "eval_runtime": 528.0293,
+      "eval_samples_per_second": 722.67,
+      "eval_steps_per_second": 22.584,
+      "step": 228957
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.3835,
+      "step": 229376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.3841,
+      "step": 229888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3702,
+      "step": 230400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.3685,
+      "step": 230912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6210862029543436e-05,
+      "loss": 4.3712,
+      "step": 231424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6202476082032916e-05,
+      "loss": 4.3529,
+      "step": 231936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6194090134522396e-05,
+      "loss": 4.3629,
+      "step": 232448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6185704187011876e-05,
+      "loss": 4.3723,
+      "step": 232960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6177318239501356e-05,
+      "loss": 4.3615,
+      "step": 233472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6168932291990836e-05,
+      "loss": 4.3695,
+      "step": 233984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616054634448031e-05,
+      "loss": 4.3765,
+      "step": 234496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.615216039696979e-05,
+      "loss": 4.373,
+      "step": 235008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.614377444945927e-05,
+      "loss": 4.3642,
+      "step": 235520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.613538850194875e-05,
+      "loss": 4.3695,
+      "step": 236032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.612700255443823e-05,
+      "loss": 4.3618,
+      "step": 236544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611861660692771e-05,
+      "loss": 4.3621,
+      "step": 237056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.611023065941719e-05,
+      "loss": 4.3548,
+      "step": 237568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6101861090710405e-05,
+      "loss": 4.3633,
+      "step": 238080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6093475143199885e-05,
+      "loss": 4.3644,
+      "step": 238592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6085089195689365e-05,
+      "loss": 4.3678,
+      "step": 239104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6076703248178845e-05,
+      "loss": 4.3579,
+      "step": 239616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6068317300668325e-05,
+      "loss": 4.3697,
+      "step": 240128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6059931353157805e-05,
+      "loss": 4.3595,
+      "step": 240640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6051545405647285e-05,
+      "loss": 4.3603,
+      "step": 241152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6043159458136764e-05,
+      "loss": 4.36,
+      "step": 241664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6034789889429974e-05,
+      "loss": 4.3574,
+      "step": 242176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.6026403941919454e-05,
+      "loss": 4.3632,
+      "step": 242688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.601803437321266e-05,
+      "loss": 4.3485,
+      "step": 243200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.600964842570214e-05,
+      "loss": 4.3636,
+      "step": 243712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.600126247819162e-05,
+      "loss": 4.3482,
+      "step": 244224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.599287653068111e-05,
+      "loss": 4.3511,
+      "step": 244736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.598449058317059e-05,
+      "loss": 4.3541,
+      "step": 245248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.59761210144638e-05,
+      "loss": 4.3607,
+      "step": 245760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.596773506695328e-05,
+      "loss": 4.3465,
+      "step": 246272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595934911944276e-05,
+      "loss": 4.3492,
+      "step": 246784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3564,
+      "step": 247296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.594259360322545e-05,
+      "loss": 4.351,
+      "step": 247808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.593420765571493e-05,
+      "loss": 4.3661,
+      "step": 248320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.592582170820441e-05,
+      "loss": 4.3496,
+      "step": 248832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.591743576069389e-05,
+      "loss": 4.3374,
+      "step": 249344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590904981318337e-05,
+      "loss": 4.3399,
+      "step": 249856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.590066386567285e-05,
+      "loss": 4.3353,
+      "step": 250368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.589227791816233e-05,
+      "loss": 4.3579,
+      "step": 250880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.588389197065181e-05,
+      "loss": 4.3389,
+      "step": 251392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.587550602314129e-05,
+      "loss": 4.3493,
+      "step": 251904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.586712007563077e-05,
+      "loss": 4.3448,
+      "step": 252416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585873412812025e-05,
+      "loss": 4.3442,
+      "step": 252928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.585038093821719e-05,
+      "loss": 4.3525,
+      "step": 253440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.584199499070667e-05,
+      "loss": 4.3274,
+      "step": 253952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5833609043196145e-05,
+      "loss": 4.3281,
+      "step": 254464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5825223095685625e-05,
+      "loss": 4.3346,
+      "step": 254976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5816837148175105e-05,
+      "loss": 4.3477,
+      "step": 255488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.580846757946832e-05,
+      "loss": 4.3264,
+      "step": 256000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.58000816319578e-05,
+      "loss": 4.3472,
+      "step": 256512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5791712063251017e-05,
+      "loss": 4.3282,
+      "step": 257024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5783326115740497e-05,
+      "loss": 4.3255,
+      "step": 257536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.577494016822997e-05,
+      "loss": 4.3309,
+      "step": 258048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.576655422071945e-05,
+      "loss": 4.3202,
+      "step": 258560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.575816827320893e-05,
+      "loss": 4.3388,
+      "step": 259072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5749798704502146e-05,
+      "loss": 4.3295,
+      "step": 259584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.574141275699162e-05,
+      "loss": 4.3151,
+      "step": 260096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.57330268094811e-05,
+      "loss": 4.3348,
+      "step": 260608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.572464086197058e-05,
+      "loss": 4.3234,
+      "step": 261120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.571625491446006e-05,
+      "loss": 4.3292,
+      "step": 261632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.570786896694954e-05,
+      "loss": 4.3206,
+      "step": 262144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.569948301943902e-05,
+      "loss": 4.3318,
+      "step": 262656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.56910970719285e-05,
+      "loss": 4.3128,
+      "step": 263168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.568271112441798e-05,
+      "loss": 4.3357,
+      "step": 263680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5674325176907465e-05,
+      "loss": 4.3195,
+      "step": 264192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5665939229396945e-05,
+      "loss": 4.3199,
+      "step": 264704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5657553281886425e-05,
+      "loss": 4.3279,
+      "step": 265216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5649183713179634e-05,
+      "loss": 4.3235,
+      "step": 265728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5640797765669114e-05,
+      "loss": 4.331,
+      "step": 266240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5632411818158594e-05,
+      "loss": 4.342,
+      "step": 266752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5624025870648074e-05,
+      "loss": 4.3382,
+      "step": 267264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.561565630194128e-05,
+      "loss": 4.3233,
+      "step": 267776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.560727035443076e-05,
+      "loss": 4.317,
+      "step": 268288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559888440692024e-05,
+      "loss": 4.3097,
+      "step": 268800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.559049845940972e-05,
+      "loss": 4.327,
+      "step": 269312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.558212889070293e-05,
+      "loss": 4.3345,
+      "step": 269824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.557374294319242e-05,
+      "loss": 4.3293,
+      "step": 270336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.55653569956819e-05,
+      "loss": 4.3153,
+      "step": 270848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.555697104817138e-05,
+      "loss": 4.3061,
+      "step": 271360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554860147946459e-05,
+      "loss": 4.3241,
+      "step": 271872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.554021553195407e-05,
+      "loss": 4.3027,
+      "step": 272384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.553182958444355e-05,
+      "loss": 4.3173,
+      "step": 272896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.552344363693303e-05,
+      "loss": 4.3184,
+      "step": 273408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.551505768942251e-05,
+      "loss": 4.3222,
+      "step": 273920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.550667174191198e-05,
+      "loss": 4.3176,
+      "step": 274432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.549828579440146e-05,
+      "loss": 4.3239,
+      "step": 274944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548989984689094e-05,
+      "loss": 4.3044,
+      "step": 275456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.548151389938042e-05,
+      "loss": 4.3166,
+      "step": 275968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.54731279518699e-05,
+      "loss": 4.3141,
+      "step": 276480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5464758383163116e-05,
+      "loss": 4.3054,
+      "step": 276992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5456372435652596e-05,
+      "loss": 4.3035,
+      "step": 277504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5447986488142076e-05,
+      "loss": 4.3211,
+      "step": 278016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5439600540631556e-05,
+      "loss": 4.319,
+      "step": 278528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5431214593121036e-05,
+      "loss": 4.3152,
+      "step": 279040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5422828645610516e-05,
+      "loss": 4.3086,
+      "step": 279552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5414459076903725e-05,
+      "loss": 4.319,
+      "step": 280064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5406073129393205e-05,
+      "loss": 4.3022,
+      "step": 280576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5397703560686414e-05,
+      "loss": 4.3152,
+      "step": 281088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5389317613175894e-05,
+      "loss": 4.3196,
+      "step": 281600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5380931665665374e-05,
+      "loss": 4.3021,
+      "step": 282112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5372545718154854e-05,
+      "loss": 4.3186,
+      "step": 282624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.536415977064434e-05,
+      "loss": 4.2845,
+      "step": 283136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.535577382313382e-05,
+      "loss": 4.292,
+      "step": 283648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.53473878756233e-05,
+      "loss": 4.3014,
+      "step": 284160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533900192811278e-05,
+      "loss": 4.3046,
+      "step": 284672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533061598060226e-05,
+      "loss": 4.3121,
+      "step": 285184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.532224641189547e-05,
+      "loss": 4.2942,
+      "step": 285696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.531386046438495e-05,
+      "loss": 4.2948,
+      "step": 286208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.530547451687443e-05,
+      "loss": 4.3008,
+      "step": 286720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.529708856936391e-05,
+      "loss": 4.3112,
+      "step": 287232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528870262185339e-05,
+      "loss": 4.3074,
+      "step": 287744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.528031667434287e-05,
+      "loss": 4.3069,
+      "step": 288256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.527193072683234e-05,
+      "loss": 4.2959,
+      "step": 288768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.526354477932182e-05,
+      "loss": 4.2976,
+      "step": 289280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.525517521061504e-05,
+      "loss": 4.3112,
+      "step": 289792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5246805641908255e-05,
+      "loss": 4.2964,
+      "step": 290304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5238419694397735e-05,
+      "loss": 4.2952,
+      "step": 290816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5230033746887215e-05,
+      "loss": 4.2996,
+      "step": 291328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5221647799376694e-05,
+      "loss": 4.2857,
+      "step": 291840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.521326185186617e-05,
+      "loss": 4.2995,
+      "step": 292352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5204892283159384e-05,
+      "loss": 4.2975,
+      "step": 292864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5196506335648863e-05,
+      "loss": 4.308,
+      "step": 293376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5188120388138343e-05,
+      "loss": 4.2858,
+      "step": 293888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5179734440627817e-05,
+      "loss": 4.2902,
+      "step": 294400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5171348493117297e-05,
+      "loss": 4.3105,
+      "step": 294912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.516297892441051e-05,
+      "loss": 4.2899,
+      "step": 295424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.515459297689999e-05,
+      "loss": 4.301,
+      "step": 295936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.514620702938947e-05,
+      "loss": 4.2964,
+      "step": 296448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.513782108187895e-05,
+      "loss": 4.3076,
+      "step": 296960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512943513436843e-05,
+      "loss": 4.2978,
+      "step": 297472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512106556566164e-05,
+      "loss": 4.2899,
+      "step": 297984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.511267961815112e-05,
+      "loss": 4.2849,
+      "step": 298496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.51042936706406e-05,
+      "loss": 4.2817,
+      "step": 299008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.509590772313008e-05,
+      "loss": 4.2937,
+      "step": 299520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.508752177561956e-05,
+      "loss": 4.3045,
+      "step": 300032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507915220691277e-05,
+      "loss": 4.2951,
+      "step": 300544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.507076625940225e-05,
+      "loss": 4.2965,
+      "step": 301056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.506238031189173e-05,
+      "loss": 4.2861,
+      "step": 301568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.505399436438121e-05,
+      "loss": 4.2868,
+      "step": 302080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.50456084168707e-05,
+      "loss": 4.2778,
+      "step": 302592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.503722246936018e-05,
+      "loss": 4.2948,
+      "step": 303104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5028852900653386e-05,
+      "loss": 4.2961,
+      "step": 303616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5020466953142866e-05,
+      "loss": 4.2997,
+      "step": 304128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5012081005632346e-05,
+      "loss": 4.2811,
+      "step": 304640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5003695058121826e-05,
+      "loss": 4.2801,
+      "step": 305152
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.2668843269348145,
+      "eval_runtime": 552.3769,
+      "eval_samples_per_second": 690.816,
+      "eval_steps_per_second": 21.589,
+      "step": 305276
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4995309110611306e-05,
+      "loss": 4.2939,
+      "step": 305664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.2912,
+      "step": 306176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2805,
+      "step": 306688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2807,
+      "step": 307200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2817,
+      "step": 307712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2675,
+      "step": 308224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.2752,
+      "step": 308736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2789,
+      "step": 309248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2784,
+      "step": 309760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.491988471942782e-05,
+      "loss": 4.2846,
+      "step": 310272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.49114987719173e-05,
+      "loss": 4.2869,
+      "step": 310784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.490311282440678e-05,
+      "loss": 4.2808,
+      "step": 311296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.489472687689626e-05,
+      "loss": 4.2796,
+      "step": 311808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.488635730818947e-05,
+      "loss": 4.2831,
+      "step": 312320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.487797136067895e-05,
+      "loss": 4.2765,
+      "step": 312832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486958541316843e-05,
+      "loss": 4.2713,
+      "step": 313344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.486119946565791e-05,
+      "loss": 4.2683,
+      "step": 313856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.485281351814739e-05,
+      "loss": 4.2802,
+      "step": 314368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.484442757063687e-05,
+      "loss": 4.2767,
+      "step": 314880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.483604162312635e-05,
+      "loss": 4.287,
+      "step": 315392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.482765567561583e-05,
+      "loss": 4.2717,
+      "step": 315904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.481926972810531e-05,
+      "loss": 4.2833,
+      "step": 316416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4810900159398524e-05,
+      "loss": 4.2768,
+      "step": 316928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.4802514211888004e-05,
+      "loss": 4.2754,
+      "step": 317440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.479412826437748e-05,
+      "loss": 4.2737,
+      "step": 317952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.478574231686696e-05,
+      "loss": 4.2717,
+      "step": 318464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.477735636935644e-05,
+      "loss": 4.2797,
+      "step": 318976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.476897042184592e-05,
+      "loss": 4.2626,
+      "step": 319488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.47605844743354e-05,
+      "loss": 4.2796,
+      "step": 320000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.475219852682488e-05,
+      "loss": 4.266,
+      "step": 320512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4743828958118086e-05,
+      "loss": 4.2644,
+      "step": 321024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.473544301060757e-05,
+      "loss": 4.2728,
+      "step": 321536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.472705706309705e-05,
+      "loss": 4.2765,
+      "step": 322048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471867111558653e-05,
+      "loss": 4.2683,
+      "step": 322560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.471030154687974e-05,
+      "loss": 4.2648,
+      "step": 323072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.470193197817295e-05,
+      "loss": 4.2694,
+      "step": 323584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.469354603066243e-05,
+      "loss": 4.2718,
+      "step": 324096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.468516008315191e-05,
+      "loss": 4.2832,
+      "step": 324608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.467677413564139e-05,
+      "loss": 4.267,
+      "step": 325120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466838818813087e-05,
+      "loss": 4.2587,
+      "step": 325632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.466000224062035e-05,
+      "loss": 4.2575,
+      "step": 326144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.465161629310983e-05,
+      "loss": 4.2557,
+      "step": 326656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.464323034559931e-05,
+      "loss": 4.2711,
+      "step": 327168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4634860776892526e-05,
+      "loss": 4.2594,
+      "step": 327680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4626474829382006e-05,
+      "loss": 4.2705,
+      "step": 328192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4618088881871486e-05,
+      "loss": 4.2665,
+      "step": 328704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4609702934360966e-05,
+      "loss": 4.2618,
+      "step": 329216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4601316986850446e-05,
+      "loss": 4.2723,
+      "step": 329728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4592947418143655e-05,
+      "loss": 4.2515,
+      "step": 330240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4584561470633135e-05,
+      "loss": 4.2455,
+      "step": 330752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4576175523122615e-05,
+      "loss": 4.2568,
+      "step": 331264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4567789575612095e-05,
+      "loss": 4.2652,
+      "step": 331776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4559420006905304e-05,
+      "loss": 4.2446,
+      "step": 332288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4551034059394784e-05,
+      "loss": 4.2693,
+      "step": 332800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.454266449068799e-05,
+      "loss": 4.2493,
+      "step": 333312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.453429492198121e-05,
+      "loss": 4.245,
+      "step": 333824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.452590897447069e-05,
+      "loss": 4.2511,
+      "step": 334336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.451752302696017e-05,
+      "loss": 4.2444,
+      "step": 334848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450913707944965e-05,
+      "loss": 4.2635,
+      "step": 335360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.450075113193913e-05,
+      "loss": 4.2543,
+      "step": 335872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.449236518442861e-05,
+      "loss": 4.2376,
+      "step": 336384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.448397923691809e-05,
+      "loss": 4.2531,
+      "step": 336896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.447559328940757e-05,
+      "loss": 4.2463,
+      "step": 337408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.446720734189705e-05,
+      "loss": 4.2572,
+      "step": 337920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445883777319026e-05,
+      "loss": 4.2415,
+      "step": 338432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.445045182567974e-05,
+      "loss": 4.2536,
+      "step": 338944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.444206587816922e-05,
+      "loss": 4.2358,
+      "step": 339456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.44336799306587e-05,
+      "loss": 4.2546,
+      "step": 339968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.442529398314818e-05,
+      "loss": 4.2435,
+      "step": 340480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.441690803563766e-05,
+      "loss": 4.2444,
+      "step": 340992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440852208812714e-05,
+      "loss": 4.2534,
+      "step": 341504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.440013614061662e-05,
+      "loss": 4.2463,
+      "step": 342016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4391766571909834e-05,
+      "loss": 4.2555,
+      "step": 342528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4383380624399314e-05,
+      "loss": 4.2638,
+      "step": 343040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.437499467688879e-05,
+      "loss": 4.2626,
+      "step": 343552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4366625108182e-05,
+      "loss": 4.2468,
+      "step": 344064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.435823916067148e-05,
+      "loss": 4.2424,
+      "step": 344576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4349853213160956e-05,
+      "loss": 4.2357,
+      "step": 345088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4341467265650436e-05,
+      "loss": 4.249,
+      "step": 345600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4333081318139916e-05,
+      "loss": 4.2633,
+      "step": 346112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.432471174943313e-05,
+      "loss": 4.2518,
+      "step": 346624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.431632580192261e-05,
+      "loss": 4.2416,
+      "step": 347136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.430795623321583e-05,
+      "loss": 4.2285,
+      "step": 347648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429957028570531e-05,
+      "loss": 4.2545,
+      "step": 348160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.429118433819479e-05,
+      "loss": 4.2301,
+      "step": 348672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.2406,
+      "step": 349184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.2422,
+      "step": 349696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.4266042874466956e-05,
+      "loss": 4.2467,
+      "step": 350208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.425765692695643e-05,
+      "loss": 4.2461,
+      "step": 350720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424927097944591e-05,
+      "loss": 4.2505,
+      "step": 351232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.424088503193539e-05,
+      "loss": 4.2314,
+      "step": 351744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.423249908442487e-05,
+      "loss": 4.2441,
+      "step": 352256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.422411313691435e-05,
+      "loss": 4.2405,
+      "step": 352768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4215727189403836e-05,
+      "loss": 4.2343,
+      "step": 353280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4207341241893316e-05,
+      "loss": 4.2297,
+      "step": 353792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4198955294382796e-05,
+      "loss": 4.2526,
+      "step": 354304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4190569346872276e-05,
+      "loss": 4.2454,
+      "step": 354816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4182183399361756e-05,
+      "loss": 4.244,
+      "step": 355328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4173797451851236e-05,
+      "loss": 4.2359,
+      "step": 355840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4165411504340716e-05,
+      "loss": 4.2424,
+      "step": 356352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4157025556830196e-05,
+      "loss": 4.2304,
+      "step": 356864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4148655988123405e-05,
+      "loss": 4.2452,
+      "step": 357376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4140270040612885e-05,
+      "loss": 4.2448,
+      "step": 357888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4131900471906094e-05,
+      "loss": 4.231,
+      "step": 358400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4123514524395574e-05,
+      "loss": 4.2478,
+      "step": 358912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4115128576885054e-05,
+      "loss": 4.2148,
+      "step": 359424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4106742629374534e-05,
+      "loss": 4.2219,
+      "step": 359936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.409835668186402e-05,
+      "loss": 4.229,
+      "step": 360448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.40899707343535e-05,
+      "loss": 4.2327,
+      "step": 360960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.4081584786842974e-05,
+      "loss": 4.2422,
+      "step": 361472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.407321521813619e-05,
+      "loss": 4.2302,
+      "step": 361984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.406482927062567e-05,
+      "loss": 4.2193,
+      "step": 362496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.405644332311514e-05,
+      "loss": 4.2283,
+      "step": 363008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.404805737560462e-05,
+      "loss": 4.2396,
+      "step": 363520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403968780689784e-05,
+      "loss": 4.2368,
+      "step": 364032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.403130185938732e-05,
+      "loss": 4.2402,
+      "step": 364544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.402291591187679e-05,
+      "loss": 4.2264,
+      "step": 365056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.401452996436627e-05,
+      "loss": 4.2252,
+      "step": 365568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.400614401685576e-05,
+      "loss": 4.2421,
+      "step": 366080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.399777444814897e-05,
+      "loss": 4.2293,
+      "step": 366592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398938850063845e-05,
+      "loss": 4.2244,
+      "step": 367104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.398100255312793e-05,
+      "loss": 4.2292,
+      "step": 367616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.397261660561741e-05,
+      "loss": 4.2178,
+      "step": 368128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.396423065810689e-05,
+      "loss": 4.2304,
+      "step": 368640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.395584471059637e-05,
+      "loss": 4.2314,
+      "step": 369152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.394745876308585e-05,
+      "loss": 4.2375,
+      "step": 369664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3939089194379056e-05,
+      "loss": 4.2181,
+      "step": 370176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3930703246868536e-05,
+      "loss": 4.2235,
+      "step": 370688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3922317299358016e-05,
+      "loss": 4.2384,
+      "step": 371200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3913931351847496e-05,
+      "loss": 4.2273,
+      "step": 371712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3905545404336976e-05,
+      "loss": 4.2315,
+      "step": 372224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3897159456826456e-05,
+      "loss": 4.2322,
+      "step": 372736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3888773509315936e-05,
+      "loss": 4.2368,
+      "step": 373248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.388038756180542e-05,
+      "loss": 4.235,
+      "step": 373760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.387201799309863e-05,
+      "loss": 4.2204,
+      "step": 374272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.386363204558811e-05,
+      "loss": 4.2184,
+      "step": 374784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.385524609807759e-05,
+      "loss": 4.2202,
+      "step": 375296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.384686015056707e-05,
+      "loss": 4.2283,
+      "step": 375808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383847420305655e-05,
+      "loss": 4.2312,
+      "step": 376320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.383010463434976e-05,
+      "loss": 4.2327,
+      "step": 376832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.382173506564297e-05,
+      "loss": 4.2304,
+      "step": 377344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.381334911813245e-05,
+      "loss": 4.2221,
+      "step": 377856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.380496317062193e-05,
+      "loss": 4.2214,
+      "step": 378368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.379657722311141e-05,
+      "loss": 4.212,
+      "step": 378880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.378819127560089e-05,
+      "loss": 4.2287,
+      "step": 379392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3779805328090376e-05,
+      "loss": 4.2309,
+      "step": 379904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3771419380579856e-05,
+      "loss": 4.2346,
+      "step": 380416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3763033433069336e-05,
+      "loss": 4.2136,
+      "step": 380928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3754663864362545e-05,
+      "loss": 4.216,
+      "step": 381440
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.209681034088135,
+      "eval_runtime": 546.9996,
+      "eval_samples_per_second": 697.607,
+      "eval_steps_per_second": 21.801,
+      "step": 381595
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3746277916852025e-05,
+      "loss": 4.2385,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3737891969341505e-05,
+      "loss": 4.223,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372950602183098e-05,
+      "loss": 4.2188,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372112007432046e-05,
+      "loss": 4.2132,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371273412680994e-05,
+      "loss": 4.2169,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370434817929942e-05,
+      "loss": 4.206,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36959622317889e-05,
+      "loss": 4.2118,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368757628427838e-05,
+      "loss": 4.21,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367919033676786e-05,
+      "loss": 4.2146,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3670804389257345e-05,
+      "loss": 4.2242,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3662418441746825e-05,
+      "loss": 4.2179,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3654032494236305e-05,
+      "loss": 4.2164,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3645646546725785e-05,
+      "loss": 4.2206,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3637260599215265e-05,
+      "loss": 4.2192,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3628874651704745e-05,
+      "loss": 4.2126,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3620488704194225e-05,
+      "loss": 4.2078,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3612119135487434e-05,
+      "loss": 4.2057,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3603733187976914e-05,
+      "loss": 4.2136,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3595347240466394e-05,
+      "loss": 4.2142,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.358696129295587e-05,
+      "loss": 4.2238,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357857534544535e-05,
+      "loss": 4.2102,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357018939793483e-05,
+      "loss": 4.2208,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356180345042431e-05,
+      "loss": 4.2169,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355341750291379e-05,
+      "loss": 4.2107,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3545047934207e-05,
+      "loss": 4.2116,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.353666198669648e-05,
+      "loss": 4.2096,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.352827603918596e-05,
+      "loss": 4.2189,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351989009167544e-05,
+      "loss": 4.1993,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.351152052296865e-05,
+      "loss": 4.2214,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.350313457545813e-05,
+      "loss": 4.2033,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.349474862794761e-05,
+      "loss": 4.1994,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.348636268043709e-05,
+      "loss": 4.2146,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34779931117303e-05,
+      "loss": 4.2114,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3469623543023516e-05,
+      "loss": 4.2141,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3461237595512996e-05,
+      "loss": 4.2059,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3452851648002476e-05,
+      "loss": 4.2053,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344448207929569e-05,
+      "loss": 4.2109,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3436096131785165e-05,
+      "loss": 4.2248,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3427710184274645e-05,
+      "loss": 4.2069,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3419324236764125e-05,
+      "loss": 4.1988,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3410938289253605e-05,
+      "loss": 4.197,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3402552341743085e-05,
+      "loss": 4.1953,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3394166394232565e-05,
+      "loss": 4.2089,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385780446722045e-05,
+      "loss": 4.2018,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377394499211525e-05,
+      "loss": 4.2077,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369008551701005e-05,
+      "loss": 4.2098,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360622604190485e-05,
+      "loss": 4.2019,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33522530354837e-05,
+      "loss": 4.2146,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.334386708797318e-05,
+      "loss": 4.1925,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.333549751926639e-05,
+      "loss": 4.1841,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.33271279505596e-05,
+      "loss": 4.2021,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331874200304908e-05,
+      "loss": 4.1995,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.331035605553856e-05,
+      "loss": 4.191,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.330197010802804e-05,
+      "loss": 4.2095,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.329360053932125e-05,
+      "loss": 4.1891,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328521459181073e-05,
+      "loss": 4.1906,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327682864430021e-05,
+      "loss": 4.1929,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326844269678969e-05,
+      "loss": 4.1879,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326005674927917e-05,
+      "loss": 4.2034,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3251670801768654e-05,
+      "loss": 4.1946,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3243284854258134e-05,
+      "loss": 4.1888,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3234898906747614e-05,
+      "loss": 4.1914,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3226512959237094e-05,
+      "loss": 4.185,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3218127011726574e-05,
+      "loss": 4.2034,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3209741064216054e-05,
+      "loss": 4.187,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320135511670553e-05,
+      "loss": 4.201,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319298554799874e-05,
+      "loss": 4.1758,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.1942,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.31762136529777e-05,
+      "loss": 4.1887,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.1864,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315945813676039e-05,
+      "loss": 4.1974,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.315108856805361e-05,
+      "loss": 4.1929,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.314270262054309e-05,
+      "loss": 4.1956,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.313431667303257e-05,
+      "loss": 4.2061,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.312593072552205e-05,
+      "loss": 4.2069,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.311756115681526e-05,
+      "loss": 4.1948,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310917520930474e-05,
+      "loss": 4.1879,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.310078926179422e-05,
+      "loss": 4.1817,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.30924033142837e-05,
+      "loss": 4.192,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.308401736677318e-05,
+      "loss": 4.2048,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.307563141926265e-05,
+      "loss": 4.196,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306724547175213e-05,
+      "loss": 4.1901,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305885952424161e-05,
+      "loss": 4.1738,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3050489955534826e-05,
+      "loss": 4.2003,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3042104008024306e-05,
+      "loss": 4.1731,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3033718060513786e-05,
+      "loss": 4.1871,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3025332113003266e-05,
+      "loss": 4.1884,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3016946165492746e-05,
+      "loss": 4.1923,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.1904,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3000174270471705e-05,
+      "loss": 4.1993,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2991788322961185e-05,
+      "loss": 4.1745,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983418754254395e-05,
+      "loss": 4.1861,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975032806743874e-05,
+      "loss": 4.1902,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2966646859233354e-05,
+      "loss": 4.1805,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.1762,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2949891343016043e-05,
+      "loss": 4.1959,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294152177430926e-05,
+      "loss": 4.19,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.293313582679874e-05,
+      "loss": 4.197,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.292474987928822e-05,
+      "loss": 4.1811,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29163639317777e-05,
+      "loss": 4.188,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290797798426718e-05,
+      "loss": 4.1771,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289960841556039e-05,
+      "loss": 4.1914,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289122246804987e-05,
+      "loss": 4.1946,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288283652053935e-05,
+      "loss": 4.178,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287445057302883e-05,
+      "loss": 4.1948,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.286606462551831e-05,
+      "loss": 4.1623,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285767867800779e-05,
+      "loss": 4.1725,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284929273049727e-05,
+      "loss": 4.1719,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284090678298675e-05,
+      "loss": 4.1812,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283252083547623e-05,
+      "loss": 4.1905,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.1836,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281574894045519e-05,
+      "loss": 4.1593,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280736299294467e-05,
+      "loss": 4.1801,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2798993424237884e-05,
+      "loss": 4.1855,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2790607476727364e-05,
+      "loss": 4.1882,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278222152921684e-05,
+      "loss": 4.1846,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277383558170632e-05,
+      "loss": 4.1753,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276546601299953e-05,
+      "loss": 4.1739,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275709644429274e-05,
+      "loss": 4.1929,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274871049678222e-05,
+      "loss": 4.1775,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27403245492717e-05,
+      "loss": 4.1752,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.1727,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.27235690330544e-05,
+      "loss": 4.1699,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.271518308554388e-05,
+      "loss": 4.1783,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.270679713803336e-05,
+      "loss": 4.1811,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269841119052284e-05,
+      "loss": 4.1865,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.269002524301231e-05,
+      "loss": 4.1705,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.268163929550179e-05,
+      "loss": 4.1706,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.267325334799127e-05,
+      "loss": 4.1865,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.266486740048075e-05,
+      "loss": 4.1726,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.265649783177396e-05,
+      "loss": 4.1804,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.264811188426344e-05,
+      "loss": 4.1829,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263972593675292e-05,
+      "loss": 4.1881,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.26313399892424e-05,
+      "loss": 4.1844,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2622970420535615e-05,
+      "loss": 4.1725,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2614584473025095e-05,
+      "loss": 4.1635,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2606198525514575e-05,
+      "loss": 4.1709,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2597812578004055e-05,
+      "loss": 4.1769,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2589426630493535e-05,
+      "loss": 4.1859,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2581040682983015e-05,
+      "loss": 4.1807,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2572654735472495e-05,
+      "loss": 4.1762,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2564268787961975e-05,
+      "loss": 4.1764,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2555899219255184e-05,
+      "loss": 4.1747,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2547513271744664e-05,
+      "loss": 4.1639,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253914370303787e-05,
+      "loss": 4.1714,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253075775552735e-05,
+      "loss": 4.1847,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.252237180801684e-05,
+      "loss": 4.1864,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.251398586050632e-05,
+      "loss": 4.1653,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25055999129958e-05,
+      "loss": 4.1643,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.168251037597656,
+      "eval_runtime": 545.9187,
+      "eval_samples_per_second": 698.989,
+      "eval_steps_per_second": 21.844,
+      "step": 457914
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.889853492561904e+17,
+  "trial_name": null,
+  "trial_params": null
+}