diff --git "a/checkpoint-534240/trainer_state.json" "b/checkpoint-534240/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-534240/trainer_state.json"
@@ -0,0 +1,6339 @@
+{
+  "best_metric": 4.1431803703308105,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/superlative-quantifier/lstm/0/checkpoints/checkpoint-534240",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 534240,
+  "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": 10.82,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5546,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0503,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9849,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9419,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8974,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7337,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994129836742636e-05,
+      "loss": 6.6213,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993291241991584e-05,
+      "loss": 6.5231,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992452647240532e-05,
+      "loss": 6.4488,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.387,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990778733499175e-05,
+      "loss": 6.3192,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989940138748123e-05,
+      "loss": 6.251,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989101543997071e-05,
+      "loss": 6.1877,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988262949246018e-05,
+      "loss": 6.1332,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987424354494966e-05,
+      "loss": 6.0853,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986585759743914e-05,
+      "loss": 6.0394,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985747164992862e-05,
+      "loss": 6.0014,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.98490857024181e-05,
+      "loss": 5.9521,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984069975490758e-05,
+      "loss": 5.9235,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983231380739706e-05,
+      "loss": 5.888,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9823927859886547e-05,
+      "loss": 5.8421,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9815541912376026e-05,
+      "loss": 5.8122,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807155964865506e-05,
+      "loss": 5.7768,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.757,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7237,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.6951,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6657,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6602,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6197,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6061,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5963,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5668,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5484,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5429,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.521,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5112,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4712,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.4662,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4373,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4341,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4212,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.4154,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.3845,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963110020355577e-05,
+      "loss": 5.3834,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962271425604525e-05,
+      "loss": 5.3702,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961432830853473e-05,
+      "loss": 5.3712,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3632,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3355,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3197,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.3291,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957241494978587e-05,
+      "loss": 5.3235,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.2975,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.2878,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.2851,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953888753854752e-05,
+      "loss": 5.2668,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.2751,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2395,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.2441,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2344,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2325,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2094,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948020228477761e-05,
+      "loss": 5.2163,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9471816337267086e-05,
+      "loss": 5.1898,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1954,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945507719985351e-05,
+      "loss": 5.1949,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.1876,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.1666,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.1504,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1431,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.941316384110464e-05,
+      "loss": 5.1588,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1467,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1317,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1423,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9379636429866296e-05,
+      "loss": 5.1407,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1246,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1141,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9354494966138465e-05,
+      "loss": 5.1025,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9346109018627945e-05,
+      "loss": 5.1008,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0848,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0875,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0759,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0886,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9304195659879074e-05,
+      "loss": 5.0761,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295809712368554e-05,
+      "loss": 5.0531,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.0553,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279037817347514e-05,
+      "loss": 5.0543,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0537,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0475,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0301,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0286,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0242,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.01,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0009,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921196661606708e-05,
+      "loss": 5.0005,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920358066855656e-05,
+      "loss": 5.0074,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919519472104604e-05,
+      "loss": 4.9963,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 4.9849,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 4.9757,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 4.978,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.91616673098077e-05,
+      "loss": 4.9804,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9153314119904636e-05,
+      "loss": 4.9733,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9144928172394116e-05,
+      "loss": 4.9661,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9136542224883596e-05,
+      "loss": 4.9545,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9128156277373076e-05,
+      "loss": 4.9626,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 4.9414,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111400761155765e-05,
+      "loss": 4.9456,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9103014813645245e-05,
+      "loss": 4.9375,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9295,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9266,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9311,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9175,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9165,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905271550738586e-05,
+      "loss": 4.9123,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904434593867907e-05,
+      "loss": 4.9078,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903595999116855e-05,
+      "loss": 4.9117,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.902757404365803e-05,
+      "loss": 4.9042,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901918809614751e-05,
+      "loss": 4.9018,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.901080214863699e-05,
+      "loss": 4.8986,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.900241620112647e-05,
+      "loss": 4.8912,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.899403025361595e-05,
+      "loss": 4.8912,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.898564430610543e-05,
+      "loss": 4.8787,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.8745,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.8747,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896048646357387e-05,
+      "loss": 4.8718,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895210051606335e-05,
+      "loss": 4.8669,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.8643,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8618,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.8489,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.8589,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8910203536118214e-05,
+      "loss": 4.8571,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890181758860769e-05,
+      "loss": 4.8379,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8257,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888504569358665e-05,
+      "loss": 4.8331,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8876659746076134e-05,
+      "loss": 4.8404,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8868273798565614e-05,
+      "loss": 4.8476,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8294,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.8286,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8286,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8279,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8146,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8077,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8158,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8058,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8023,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87844634598716e-05,
+      "loss": 4.8069,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877607751236109e-05,
+      "loss": 4.7918,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876769156485057e-05,
+      "loss": 4.8023,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875932199614378e-05,
+      "loss": 4.7977,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875093604863326e-05,
+      "loss": 4.7768,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.762511730194092,
+      "eval_runtime": 289.3295,
+      "eval_samples_per_second": 1318.88,
+      "eval_steps_per_second": 41.216,
+      "step": 76320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.79,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.7782,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.7894,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8717408637394906e-05,
+      "loss": 4.7777,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709022689884386e-05,
+      "loss": 4.7783,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700636742373866e-05,
+      "loss": 4.7634,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692250794863346e-05,
+      "loss": 4.7643,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683864847352826e-05,
+      "loss": 4.7456,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.7723,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.7645,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.7603,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7642,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7465,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7453,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7414,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7391,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7391,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7315,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7411,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.745,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7307,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856651071861674e-05,
+      "loss": 4.7233,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558124771106226e-05,
+      "loss": 4.7216,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.85497388235957e-05,
+      "loss": 4.7274,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.854135287608518e-05,
+      "loss": 4.7189,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.853296692857466e-05,
+      "loss": 4.7055,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.852458098106414e-05,
+      "loss": 4.7062,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.851619503355362e-05,
+      "loss": 4.7205,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.6936,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8499455896140044e-05,
+      "loss": 4.6998,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7086,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.7048,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.7028,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.7012,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.845754253739117e-05,
+      "loss": 4.6982,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7013,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.6776,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.6857,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.6708,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.6779,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.6762,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839884090481753e-05,
+      "loss": 4.6776,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839045495730701e-05,
+      "loss": 4.6716,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838208538860022e-05,
+      "loss": 4.6702,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83736994410897e-05,
+      "loss": 4.6704,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836531349357918e-05,
+      "loss": 4.673,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.835692754606866e-05,
+      "loss": 4.6759,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834854159855814e-05,
+      "loss": 4.659,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834015565104763e-05,
+      "loss": 4.6493,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.6662,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.832340013483032e-05,
+      "loss": 4.6701,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83150141873198e-05,
+      "loss": 4.6532,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.830662823980928e-05,
+      "loss": 4.6497,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.829824229229876e-05,
+      "loss": 4.6499,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.828985634478824e-05,
+      "loss": 4.6444,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281486776081446e-05,
+      "loss": 4.653,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273100828570926e-05,
+      "loss": 4.6336,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264714881060406e-05,
+      "loss": 4.645,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256328933549886e-05,
+      "loss": 4.6406,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6426,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823957341733258e-05,
+      "loss": 4.6235,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823118746982206e-05,
+      "loss": 4.6439,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8222801522311535e-05,
+      "loss": 4.618,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8214415574801015e-05,
+      "loss": 4.634,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.6331,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.6327,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8189274111073184e-05,
+      "loss": 4.6174,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8180888163562664e-05,
+      "loss": 4.6131,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.612,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816413264734535e-05,
+      "loss": 4.6219,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815574669983483e-05,
+      "loss": 4.6238,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814736075232432e-05,
+      "loss": 4.6154,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.6226,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6282,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.6227,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8113849719889704e-05,
+      "loss": 4.6061,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8105463772379184e-05,
+      "loss": 4.6122,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809707782486866e-05,
+      "loss": 4.6127,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.5903,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.6122,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.5947,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.6159,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.805514808731606e-05,
+      "loss": 4.6035,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804676213980554e-05,
+      "loss": 4.5833,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803837619229502e-05,
+      "loss": 4.5997,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8029990244784504e-05,
+      "loss": 4.5898,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.6032,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801325110737092e-05,
+      "loss": 4.5941,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80048651598604e-05,
+      "loss": 4.5906,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799647921234988e-05,
+      "loss": 4.5882,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798809326483936e-05,
+      "loss": 4.5827,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797970731732884e-05,
+      "loss": 4.5771,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797132136981832e-05,
+      "loss": 4.5704,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796295180111153e-05,
+      "loss": 4.5749,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795456585360101e-05,
+      "loss": 4.5858,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794617990609049e-05,
+      "loss": 4.5763,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793779395857997e-05,
+      "loss": 4.5707,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792940801106946e-05,
+      "loss": 4.5656,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792102206355894e-05,
+      "loss": 4.5722,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791263611604842e-05,
+      "loss": 4.5791,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79042501685379e-05,
+      "loss": 4.572,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.789586422102737e-05,
+      "loss": 4.567,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.788747827351685e-05,
+      "loss": 4.5647,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.787909232600633e-05,
+      "loss": 4.5702,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870722757299546e-05,
+      "loss": 4.5517,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.786233680978902e-05,
+      "loss": 4.5606,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5589,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.552,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5526,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782884215615814e-05,
+      "loss": 4.5556,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782045620864762e-05,
+      "loss": 4.5506,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.5496,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5486,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.5464,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778691241860554e-05,
+      "loss": 4.5563,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777852647109502e-05,
+      "loss": 4.5524,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777014052358449e-05,
+      "loss": 4.5487,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776177095487771e-05,
+      "loss": 4.5466,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775338500736719e-05,
+      "loss": 4.5466,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.774499905985666e-05,
+      "loss": 4.5458,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773661311234615e-05,
+      "loss": 4.5367,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772822716483563e-05,
+      "loss": 4.534,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771984121732511e-05,
+      "loss": 4.539,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771145526981459e-05,
+      "loss": 4.5383,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.770306932230407e-05,
+      "loss": 4.536,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769468337479355e-05,
+      "loss": 4.537,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.768629742728303e-05,
+      "loss": 4.5314,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767791147977251e-05,
+      "loss": 4.5309,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766952553226199e-05,
+      "loss": 4.531,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76611559635552e-05,
+      "loss": 4.5396,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.765277001604468e-05,
+      "loss": 4.5212,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.764438406853416e-05,
+      "loss": 4.5103,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.763599812102364e-05,
+      "loss": 4.5176,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762761217351312e-05,
+      "loss": 4.531,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.76192262260026e-05,
+      "loss": 4.5328,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761084027849208e-05,
+      "loss": 4.5249,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7602470709785293e-05,
+      "loss": 4.5231,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7594084762274773e-05,
+      "loss": 4.5264,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758569881476425e-05,
+      "loss": 4.526,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757731286725373e-05,
+      "loss": 4.5176,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7568926919743206e-05,
+      "loss": 4.5138,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7560540972232686e-05,
+      "loss": 4.5188,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7552155024722166e-05,
+      "loss": 4.5155,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7543769077211646e-05,
+      "loss": 4.5066,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7535399508504855e-05,
+      "loss": 4.5192,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7527013560994335e-05,
+      "loss": 4.503,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7518627613483815e-05,
+      "loss": 4.5151,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751025804477703e-05,
+      "loss": 4.5112,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750187209726651e-05,
+      "loss": 4.4994,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.485108852386475,
+      "eval_runtime": 288.7078,
+      "eval_samples_per_second": 1321.721,
+      "eval_steps_per_second": 41.305,
+      "step": 152640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749348614975599e-05,
+      "loss": 4.5102,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748510020224547e-05,
+      "loss": 4.4999,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747671425473495e-05,
+      "loss": 4.5105,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746832830722443e-05,
+      "loss": 4.5022,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745994235971391e-05,
+      "loss": 4.5076,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745155641220339e-05,
+      "loss": 4.4938,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744317046469287e-05,
+      "loss": 4.4934,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743478451718235e-05,
+      "loss": 4.4792,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742639856967183e-05,
+      "loss": 4.5098,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741801262216131e-05,
+      "loss": 4.4999,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740962667465079e-05,
+      "loss": 4.4917,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401240727140264e-05,
+      "loss": 4.5064,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.739287115843348e-05,
+      "loss": 4.4829,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384485210922966e-05,
+      "loss": 4.4921,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376099263412446e-05,
+      "loss": 4.4831,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736771331590192e-05,
+      "loss": 4.4809,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.73593273683914e-05,
+      "loss": 4.4883,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.735094142088088e-05,
+      "loss": 4.4768,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.734255547337036e-05,
+      "loss": 4.4903,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.4933,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.4842,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.4815,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730902806213201e-05,
+      "loss": 4.4787,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730064211462149e-05,
+      "loss": 4.4829,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.729225616711097e-05,
+      "loss": 4.4727,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.728387021960045e-05,
+      "loss": 4.4712,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.727548427208993e-05,
+      "loss": 4.4619,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.4831,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.4572,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250342808362104e-05,
+      "loss": 4.4676,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7241956860851584e-05,
+      "loss": 4.4711,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7233570913341064e-05,
+      "loss": 4.4732,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7225184965830544e-05,
+      "loss": 4.4703,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7216799018320024e-05,
+      "loss": 4.4719,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720842944961323e-05,
+      "loss": 4.4684,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720004350210271e-05,
+      "loss": 4.4737,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719165755459219e-05,
+      "loss": 4.4549,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718327160708167e-05,
+      "loss": 4.4595,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717488565957115e-05,
+      "loss": 4.4463,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716649971206063e-05,
+      "loss": 4.4542,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715811376455011e-05,
+      "loss": 4.4504,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714972781703959e-05,
+      "loss": 4.4612,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714135824833281e-05,
+      "loss": 4.4513,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713297230082228e-05,
+      "loss": 4.453,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712458635331176e-05,
+      "loss": 4.4539,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711620040580124e-05,
+      "loss": 4.4567,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710783083709446e-05,
+      "loss": 4.4599,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709944488958393e-05,
+      "loss": 4.444,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709105894207341e-05,
+      "loss": 4.4336,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708267299456289e-05,
+      "loss": 4.4579,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707428704705237e-05,
+      "loss": 4.4541,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706590109954185e-05,
+      "loss": 4.4471,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7057531530835066e-05,
+      "loss": 4.4445,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7049145583324546e-05,
+      "loss": 4.4381,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040759635814026e-05,
+      "loss": 4.4406,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032373688303506e-05,
+      "loss": 4.4468,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7023987740792986e-05,
+      "loss": 4.43,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4415,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007232224575675e-05,
+      "loss": 4.4375,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998846277065155e-05,
+      "loss": 4.4367,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990460329554635e-05,
+      "loss": 4.4236,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982074382044115e-05,
+      "loss": 4.4453,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973688434533595e-05,
+      "loss": 4.4225,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6965318865826804e-05,
+      "loss": 4.4341,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695693291831629e-05,
+      "loss": 4.4315,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.438,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4207,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693179145458846e-05,
+      "loss": 4.4234,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692340550707794e-05,
+      "loss": 4.4127,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691501955956742e-05,
+      "loss": 4.4316,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69066336120569e-05,
+      "loss": 4.4326,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689824766454638e-05,
+      "loss": 4.4258,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688986171703586e-05,
+      "loss": 4.4294,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688147576952534e-05,
+      "loss": 4.437,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687308982201481e-05,
+      "loss": 4.4366,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686472025330803e-05,
+      "loss": 4.4149,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.685633430579751e-05,
+      "loss": 4.4241,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.4311,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839578789580204e-05,
+      "loss": 4.4075,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6831192842069684e-05,
+      "loss": 4.4246,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6822806894559164e-05,
+      "loss": 4.4106,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6814420947048644e-05,
+      "loss": 4.4344,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680603499953812e-05,
+      "loss": 4.423,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67976490520276e-05,
+      "loss": 4.4024,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678926310451708e-05,
+      "loss": 4.4146,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678087715700656e-05,
+      "loss": 4.4135,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677249120949604e-05,
+      "loss": 4.4215,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6764121640789246e-05,
+      "loss": 4.4168,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6755735693278726e-05,
+      "loss": 4.4133,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6747349745768206e-05,
+      "loss": 4.4115,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673896379825769e-05,
+      "loss": 4.4068,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673057785074717e-05,
+      "loss": 4.404,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672219190323665e-05,
+      "loss": 4.3934,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671380595572613e-05,
+      "loss": 4.4052,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670543638701934e-05,
+      "loss": 4.4133,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669705043950882e-05,
+      "loss": 4.4054,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66886644919983e-05,
+      "loss": 4.398,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668027854448778e-05,
+      "loss": 4.3964,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667190897578099e-05,
+      "loss": 4.3998,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.666352302827047e-05,
+      "loss": 4.4127,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.665513708075995e-05,
+      "loss": 4.4025,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664675113324943e-05,
+      "loss": 4.3995,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663836518573891e-05,
+      "loss": 4.3935,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662997923822839e-05,
+      "loss": 4.4057,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662159329071788e-05,
+      "loss": 4.3833,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661320734320736e-05,
+      "loss": 4.3973,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604854153304296e-05,
+      "loss": 4.3905,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596468205793776e-05,
+      "loss": 4.3925,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588082258283256e-05,
+      "loss": 4.388,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579696310772736e-05,
+      "loss": 4.3922,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6571310363262216e-05,
+      "loss": 4.3894,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562924415751695e-05,
+      "loss": 4.3864,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554554847044905e-05,
+      "loss": 4.3922,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6546168899534385e-05,
+      "loss": 4.3865,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6537782952023865e-05,
+      "loss": 4.3945,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529397004513344e-05,
+      "loss": 4.3938,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652102743580656e-05,
+      "loss": 4.3904,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651264148829604e-05,
+      "loss": 4.3908,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650425554078552e-05,
+      "loss": 4.3882,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495869593275e-05,
+      "loss": 4.3839,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648748364576447e-05,
+      "loss": 4.3828,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647909769825395e-05,
+      "loss": 4.3828,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647071175074343e-05,
+      "loss": 4.3791,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646234218203665e-05,
+      "loss": 4.3891,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645395623452612e-05,
+      "loss": 4.3796,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64455702870156e-05,
+      "loss": 4.3847,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643720071830882e-05,
+      "loss": 4.38,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64288147707983e-05,
+      "loss": 4.378,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642042882328778e-05,
+      "loss": 4.3759,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641204287577726e-05,
+      "loss": 4.391,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640365692826674e-05,
+      "loss": 4.3687,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639527098075622e-05,
+      "loss": 4.3665,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63868850332457e-05,
+      "loss": 4.3649,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637849908573518e-05,
+      "loss": 4.3783,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637011313822466e-05,
+      "loss": 4.385,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636172719071414e-05,
+      "loss": 4.3809,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635334124320362e-05,
+      "loss": 4.3751,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63449552956931e-05,
+      "loss": 4.3808,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633658572698631e-05,
+      "loss": 4.375,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632819977947579e-05,
+      "loss": 4.3753,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631981383196527e-05,
+      "loss": 4.3669,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631142788445475e-05,
+      "loss": 4.3764,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630307469455169e-05,
+      "loss": 4.3713,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629468874704117e-05,
+      "loss": 4.3608,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628631917833438e-05,
+      "loss": 4.3757,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.3664,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626954728331334e-05,
+      "loss": 4.3683,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626116133580282e-05,
+      "loss": 4.3688,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62527753882923e-05,
+      "loss": 4.3586,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.35234260559082,
+      "eval_runtime": 281.6215,
+      "eval_samples_per_second": 1354.978,
+      "eval_steps_per_second": 42.344,
+      "step": 228960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624438944078178e-05,
+      "loss": 4.371,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623600349327126e-05,
+      "loss": 4.3603,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.369,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.3615,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210862029543436e-05,
+      "loss": 4.373,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202476082032916e-05,
+      "loss": 4.3574,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194090134522396e-05,
+      "loss": 4.353,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185704187011876e-05,
+      "loss": 4.3446,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177318239501356e-05,
+      "loss": 4.3704,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168932291990836e-05,
+      "loss": 4.3665,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616054634448031e-05,
+      "loss": 4.352,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6152176775773525e-05,
+      "loss": 4.3683,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143807207066734e-05,
+      "loss": 4.35,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135421259556214e-05,
+      "loss": 4.3565,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6127035312045694e-05,
+      "loss": 4.3536,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.3455,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3534,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.347,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609350790080735e-05,
+      "loss": 4.3569,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608512195329683e-05,
+      "loss": 4.3589,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607673600578631e-05,
+      "loss": 4.3587,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606835005827578e-05,
+      "loss": 4.3504,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605996411076526e-05,
+      "loss": 4.3438,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605157816325474e-05,
+      "loss": 4.3591,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604319221574422e-05,
+      "loss": 4.3392,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60348062682337e-05,
+      "loss": 4.3435,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602643669952691e-05,
+      "loss": 4.3342,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601805075201639e-05,
+      "loss": 4.3539,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600966480450588e-05,
+      "loss": 4.3296,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600127885699536e-05,
+      "loss": 4.34,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599289290948484e-05,
+      "loss": 4.3474,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598450696197432e-05,
+      "loss": 4.3444,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.59761210144638e-05,
+      "loss": 4.344,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596773506695328e-05,
+      "loss": 4.3473,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.3441,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3466,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594259360322545e-05,
+      "loss": 4.3361,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593420765571493e-05,
+      "loss": 4.3283,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592582170820441e-05,
+      "loss": 4.3246,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.591743576069389e-05,
+      "loss": 4.3325,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590904981318337e-05,
+      "loss": 4.3286,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900680244476576e-05,
+      "loss": 4.3406,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589229429696606e-05,
+      "loss": 4.329,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588390834945554e-05,
+      "loss": 4.3318,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587552240194502e-05,
+      "loss": 4.3366,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5867136454434496e-05,
+      "loss": 4.3299,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5858750506923976e-05,
+      "loss": 4.3423,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585038093821719e-05,
+      "loss": 4.324,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584199499070667e-05,
+      "loss": 4.3092,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5833609043196145e-05,
+      "loss": 4.3419,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5825223095685625e-05,
+      "loss": 4.3341,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5816837148175105e-05,
+      "loss": 4.3336,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808451200664585e-05,
+      "loss": 4.3235,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58000816319578e-05,
+      "loss": 4.3214,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.579169568444728e-05,
+      "loss": 4.3222,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.578330973693676e-05,
+      "loss": 4.3275,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577492378942624e-05,
+      "loss": 4.3146,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576655422071945e-05,
+      "loss": 4.3191,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575816827320893e-05,
+      "loss": 4.3271,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574978232569841e-05,
+      "loss": 4.3177,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574139637818789e-05,
+      "loss": 4.3076,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573301043067737e-05,
+      "loss": 4.3327,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572462448316685e-05,
+      "loss": 4.3081,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571623853565633e-05,
+      "loss": 4.3134,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570785258814581e-05,
+      "loss": 4.3233,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569948301943902e-05,
+      "loss": 4.3248,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56910970719285e-05,
+      "loss": 4.3076,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682727503221714e-05,
+      "loss": 4.3127,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674341555711194e-05,
+      "loss": 4.2959,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5665955608200674e-05,
+      "loss": 4.324,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5657569660690154e-05,
+      "loss": 4.3178,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5649183713179634e-05,
+      "loss": 4.3166,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5640797765669114e-05,
+      "loss": 4.3146,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5632411818158594e-05,
+      "loss": 4.3268,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5624025870648074e-05,
+      "loss": 4.3264,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561567268074501e-05,
+      "loss": 4.3051,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560728673323449e-05,
+      "loss": 4.314,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559890078572397e-05,
+      "loss": 4.323,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559051483821345e-05,
+      "loss": 4.2972,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3174,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557375932199615e-05,
+      "loss": 4.3023,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3206,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3172,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.2979,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55402319107578e-05,
+      "loss": 4.309,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553184596324728e-05,
+      "loss": 4.3042,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552346001573676e-05,
+      "loss": 4.3181,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551507406822624e-05,
+      "loss": 4.3069,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550668812071572e-05,
+      "loss": 4.3075,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54983021732052e-05,
+      "loss": 4.3047,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548991622569468e-05,
+      "loss": 4.3045,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481530278184157e-05,
+      "loss": 4.2991,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5473144330673637e-05,
+      "loss": 4.2878,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5464758383163116e-05,
+      "loss": 4.3,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5456372435652596e-05,
+      "loss": 4.3084,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3081,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543963329823902e-05,
+      "loss": 4.2879,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54312473507285e-05,
+      "loss": 4.2953,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542286140321798e-05,
+      "loss": 4.2974,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5414475455707455e-05,
+      "loss": 4.3097,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5406089508196934e-05,
+      "loss": 4.2996,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5397703560686414e-05,
+      "loss": 4.2983,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5389317613175894e-05,
+      "loss": 4.2904,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5380931665665374e-05,
+      "loss": 4.3062,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5372578475762326e-05,
+      "loss": 4.2807,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5364192528251806e-05,
+      "loss": 4.298,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535580658074128e-05,
+      "loss": 4.2889,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534742063323076e-05,
+      "loss": 4.2939,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533903468572024e-05,
+      "loss": 4.2876,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533064873820972e-05,
+      "loss": 4.2968,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53222627906992e-05,
+      "loss": 4.2863,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531387684318868e-05,
+      "loss": 4.2872,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530549089567816e-05,
+      "loss": 4.2918,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.2889,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528871900065712e-05,
+      "loss": 4.2977,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.52803330531466e-05,
+      "loss": 4.298,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527196348443981e-05,
+      "loss": 4.2927,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263577536929295e-05,
+      "loss": 4.2957,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255191589418775e-05,
+      "loss": 4.2902,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246805641908255e-05,
+      "loss": 4.2851,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238436073201464e-05,
+      "loss": 4.285,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.523006650449467e-05,
+      "loss": 4.285,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2876,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.2934,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520490866196311e-05,
+      "loss": 4.2834,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519653909325632e-05,
+      "loss": 4.2907,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.2802,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.2838,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517138125072476e-05,
+      "loss": 4.282,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516301168201798e-05,
+      "loss": 4.2969,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515462573450746e-05,
+      "loss": 4.2783,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514623978699694e-05,
+      "loss": 4.2715,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513785383948642e-05,
+      "loss": 4.2695,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51294678919759e-05,
+      "loss": 4.2847,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512108194446538e-05,
+      "loss": 4.2894,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511269599695486e-05,
+      "loss": 4.2894,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510431004944434e-05,
+      "loss": 4.2836,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095940480737546e-05,
+      "loss": 4.2885,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087554533227026e-05,
+      "loss": 4.282,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079168585716506e-05,
+      "loss": 4.2808,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070782638205986e-05,
+      "loss": 4.2762,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062413069499195e-05,
+      "loss": 4.2826,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505404350079241e-05,
+      "loss": 4.2809,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504565755328189e-05,
+      "loss": 4.2684,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.50372879845751e-05,
+      "loss": 4.2854,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502890203706458e-05,
+      "loss": 4.2796,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.502051608955406e-05,
+      "loss": 4.2736,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.501213014204354e-05,
+      "loss": 4.2762,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.500374419453302e-05,
+      "loss": 4.2722,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.270402908325195,
+      "eval_runtime": 282.2519,
+      "eval_samples_per_second": 1351.952,
+      "eval_steps_per_second": 42.249,
+      "step": 305280
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.49953582470225e-05,
+      "loss": 4.2767,
+      "step": 305664
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.498697229951198e-05,
+      "loss": 4.2722,
+      "step": 306176
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.497858635200146e-05,
+      "loss": 4.2811,
+      "step": 306688
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.497020040449093e-05,
+      "loss": 4.2712,
+      "step": 307200
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.496181445698042e-05,
+      "loss": 4.2828,
+      "step": 307712
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.49534285094699e-05,
+      "loss": 4.2721,
+      "step": 308224
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.494504256195938e-05,
+      "loss": 4.2615,
+      "step": 308736
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.493667299325259e-05,
+      "loss": 4.2606,
+      "step": 309248
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4928303424545805e-05,
+      "loss": 4.2757,
+      "step": 309760
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4919917477035285e-05,
+      "loss": 4.2788,
+      "step": 310272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4911531529524765e-05,
+      "loss": 4.2676,
+      "step": 310784
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.490314558201424e-05,
+      "loss": 4.2784,
+      "step": 311296
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4894776013307454e-05,
+      "loss": 4.2659,
+      "step": 311808
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4886390065796934e-05,
+      "loss": 4.2709,
+      "step": 312320
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.487800411828641e-05,
+      "loss": 4.2648,
+      "step": 312832
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.486961817077589e-05,
+      "loss": 4.2598,
+      "step": 313344
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4861232223265373e-05,
+      "loss": 4.2638,
+      "step": 313856
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.4852846275754853e-05,
+      "loss": 4.2604,
+      "step": 314368
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.484446032824433e-05,
+      "loss": 4.2738,
+      "step": 314880
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.483607438073381e-05,
+      "loss": 4.274,
+      "step": 315392
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.482768843322329e-05,
+      "loss": 4.275,
+      "step": 315904
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.481930248571277e-05,
+      "loss": 4.2664,
+      "step": 316416
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.481091653820225e-05,
+      "loss": 4.2598,
+      "step": 316928
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.480253059069173e-05,
+      "loss": 4.2706,
+      "step": 317440
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.479416102198494e-05,
+      "loss": 4.2582,
+      "step": 317952
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.478577507447442e-05,
+      "loss": 4.2588,
+      "step": 318464
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.47773891269639e-05,
+      "loss": 4.2494,
+      "step": 318976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.476900317945338e-05,
+      "loss": 4.265,
+      "step": 319488
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.476063361074659e-05,
+      "loss": 4.2529,
+      "step": 320000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.475224766323607e-05,
+      "loss": 4.2513,
+      "step": 320512
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.474386171572556e-05,
+      "loss": 4.2658,
+      "step": 321024
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.473547576821504e-05,
+      "loss": 4.2597,
+      "step": 321536
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.472710619950825e-05,
+      "loss": 4.2659,
+      "step": 322048
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.471872025199773e-05,
+      "loss": 4.2669,
+      "step": 322560
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.471033430448721e-05,
+      "loss": 4.2562,
+      "step": 323072
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.470194835697669e-05,
+      "loss": 4.265,
+      "step": 323584
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.469356240946617e-05,
+      "loss": 4.2576,
+      "step": 324096
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.468517646195565e-05,
+      "loss": 4.2438,
+      "step": 324608
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.467679051444512e-05,
+      "loss": 4.241,
+      "step": 325120
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.46684045669346e-05,
+      "loss": 4.2484,
+      "step": 325632
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4660034998227816e-05,
+      "loss": 4.2473,
+      "step": 326144
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4651649050717296e-05,
+      "loss": 4.2601,
+      "step": 326656
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.464327948201051e-05,
+      "loss": 4.2504,
+      "step": 327168
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.463489353449999e-05,
+      "loss": 4.2475,
+      "step": 327680
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.462650758698947e-05,
+      "loss": 4.2562,
+      "step": 328192
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.461812163947895e-05,
+      "loss": 4.2521,
+      "step": 328704
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4609735691968425e-05,
+      "loss": 4.2589,
+      "step": 329216
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4601349744457905e-05,
+      "loss": 4.2431,
+      "step": 329728
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.459298017575112e-05,
+      "loss": 4.2281,
+      "step": 330240
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4584594228240594e-05,
+      "loss": 4.2609,
+      "step": 330752
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4576208280730074e-05,
+      "loss": 4.2594,
+      "step": 331264
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.456783871202329e-05,
+      "loss": 4.2527,
+      "step": 331776
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.455945276451277e-05,
+      "loss": 4.2443,
+      "step": 332288
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.455106681700225e-05,
+      "loss": 4.2422,
+      "step": 332800
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.454268086949173e-05,
+      "loss": 4.2437,
+      "step": 333312
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.453429492198121e-05,
+      "loss": 4.2465,
+      "step": 333824
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.452590897447069e-05,
+      "loss": 4.2373,
+      "step": 334336
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.451752302696017e-05,
+      "loss": 4.2421,
+      "step": 334848
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.450913707944965e-05,
+      "loss": 4.2489,
+      "step": 335360
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.450075113193913e-05,
+      "loss": 4.2425,
+      "step": 335872
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.449236518442861e-05,
+      "loss": 4.2303,
+      "step": 336384
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.448397923691809e-05,
+      "loss": 4.2553,
+      "step": 336896
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.447559328940757e-05,
+      "loss": 4.2302,
+      "step": 337408
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.446724009950451e-05,
+      "loss": 4.2334,
+      "step": 337920
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.445885415199399e-05,
+      "loss": 4.2488,
+      "step": 338432
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.445046820448347e-05,
+      "loss": 4.2517,
+      "step": 338944
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.444208225697295e-05,
+      "loss": 4.227,
+      "step": 339456
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.443369630946243e-05,
+      "loss": 4.2367,
+      "step": 339968
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4425310361951914e-05,
+      "loss": 4.2198,
+      "step": 340480
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4416924414441394e-05,
+      "loss": 4.2465,
+      "step": 340992
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4408538466930874e-05,
+      "loss": 4.245,
+      "step": 341504
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.440016889822408e-05,
+      "loss": 4.2357,
+      "step": 342016
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.439178295071356e-05,
+      "loss": 4.2421,
+      "step": 342528
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.438339700320304e-05,
+      "loss": 4.2481,
+      "step": 343040
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.437501105569252e-05,
+      "loss": 4.2566,
+      "step": 343552
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.436665786578946e-05,
+      "loss": 4.2294,
+      "step": 344064
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.435828829708267e-05,
+      "loss": 4.2419,
+      "step": 344576
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.434990234957216e-05,
+      "loss": 4.2472,
+      "step": 345088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.434151640206164e-05,
+      "loss": 4.2217,
+      "step": 345600
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4333130454551117e-05,
+      "loss": 4.2447,
+      "step": 346112
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4324744507040597e-05,
+      "loss": 4.23,
+      "step": 346624
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4316358559530076e-05,
+      "loss": 4.2463,
+      "step": 347136
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4307972612019556e-05,
+      "loss": 4.2443,
+      "step": 347648
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4299586664509036e-05,
+      "loss": 4.2238,
+      "step": 348160
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4291200716998516e-05,
+      "loss": 4.2307,
+      "step": 348672
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4282814769487996e-05,
+      "loss": 4.2317,
+      "step": 349184
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4274428821977476e-05,
+      "loss": 4.2463,
+      "step": 349696
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4266042874466956e-05,
+      "loss": 4.2333,
+      "step": 350208
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 4.4257673305760165e-05,
+      "loss": 4.2313,
+      "step": 350720
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4249287358249645e-05,
+      "loss": 4.2293,
+      "step": 351232
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4240901410739125e-05,
+      "loss": 4.2332,
+      "step": 351744
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4232515463228605e-05,
+      "loss": 4.2271,
+      "step": 352256
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4224129515718085e-05,
+      "loss": 4.2138,
+      "step": 352768
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.421577632581503e-05,
+      "loss": 4.2302,
+      "step": 353280
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.420739037830451e-05,
+      "loss": 4.2357,
+      "step": 353792
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.419900443079399e-05,
+      "loss": 4.2366,
+      "step": 354304
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.419061848328347e-05,
+      "loss": 4.2158,
+      "step": 354816
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.418223253577295e-05,
+      "loss": 4.2224,
+      "step": 355328
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.417384658826243e-05,
+      "loss": 4.2274,
+      "step": 355840
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.41654606407519e-05,
+      "loss": 4.2334,
+      "step": 356352
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.415709107204512e-05,
+      "loss": 4.2319,
+      "step": 356864
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.41487051245346e-05,
+      "loss": 4.2239,
+      "step": 357376
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.414031917702408e-05,
+      "loss": 4.2204,
+      "step": 357888
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.413193322951356e-05,
+      "loss": 4.2335,
+      "step": 358400
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.412354728200304e-05,
+      "loss": 4.2112,
+      "step": 358912
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.411516133449252e-05,
+      "loss": 4.2266,
+      "step": 359424
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.4106775386982e-05,
+      "loss": 4.2152,
+      "step": 359936
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.409838943947148e-05,
+      "loss": 4.2272,
+      "step": 360448
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.409001987076469e-05,
+      "loss": 4.2131,
+      "step": 360960
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.408163392325417e-05,
+      "loss": 4.228,
+      "step": 361472
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.407324797574365e-05,
+      "loss": 4.2179,
+      "step": 361984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.406487840703686e-05,
+      "loss": 4.2187,
+      "step": 362496
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.405649245952634e-05,
+      "loss": 4.2214,
+      "step": 363008
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.404810651201582e-05,
+      "loss": 4.2191,
+      "step": 363520
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.40397205645053e-05,
+      "loss": 4.225,
+      "step": 364032
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.403135099579851e-05,
+      "loss": 4.2318,
+      "step": 364544
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.402296504828799e-05,
+      "loss": 4.2256,
+      "step": 365056
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.401457910077747e-05,
+      "loss": 4.2248,
+      "step": 365568
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.400619315326695e-05,
+      "loss": 4.219,
+      "step": 366080
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.399780720575643e-05,
+      "loss": 4.2166,
+      "step": 366592
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.398942125824591e-05,
+      "loss": 4.2129,
+      "step": 367104
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.398103531073539e-05,
+      "loss": 4.2235,
+      "step": 367616
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.397264936322487e-05,
+      "loss": 4.2156,
+      "step": 368128
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.396427979451808e-05,
+      "loss": 4.2232,
+      "step": 368640
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.395589384700756e-05,
+      "loss": 4.2159,
+      "step": 369152
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.394752427830077e-05,
+      "loss": 4.2227,
+      "step": 369664
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.393913833079025e-05,
+      "loss": 4.2143,
+      "step": 370176
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.393075238327973e-05,
+      "loss": 4.217,
+      "step": 370688
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.392236643576921e-05,
+      "loss": 4.2121,
+      "step": 371200
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.39139804882587e-05,
+      "loss": 4.2271,
+      "step": 371712
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.390559454074818e-05,
+      "loss": 4.2143,
+      "step": 372224
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3897224972041386e-05,
+      "loss": 4.2071,
+      "step": 372736
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3888839024530866e-05,
+      "loss": 4.1996,
+      "step": 373248
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3880453077020346e-05,
+      "loss": 4.2185,
+      "step": 373760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3872067129509826e-05,
+      "loss": 4.2239,
+      "step": 374272
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3863681181999306e-05,
+      "loss": 4.2206,
+      "step": 374784
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3855295234488786e-05,
+      "loss": 4.2205,
+      "step": 375296
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3846909286978266e-05,
+      "loss": 4.2193,
+      "step": 375808
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.383852333946774e-05,
+      "loss": 4.2153,
+      "step": 376320
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3830153770760955e-05,
+      "loss": 4.2214,
+      "step": 376832
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3821767823250435e-05,
+      "loss": 4.2074,
+      "step": 377344
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3813381875739915e-05,
+      "loss": 4.216,
+      "step": 377856
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3804995928229395e-05,
+      "loss": 4.2128,
+      "step": 378368
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.379662635952261e-05,
+      "loss": 4.2074,
+      "step": 378880
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.378824041201209e-05,
+      "loss": 4.2186,
+      "step": 379392
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.3779854464501564e-05,
+      "loss": 4.2143,
+      "step": 379904
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.377148489579478e-05,
+      "loss": 4.2106,
+      "step": 380416
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.376309894828426e-05,
+      "loss": 4.2111,
+      "step": 380928
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 4.375471300077374e-05,
+      "loss": 4.2128,
+      "step": 381440
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 4.214416027069092,
+      "eval_runtime": 281.5365,
+      "eval_samples_per_second": 1355.387,
+      "eval_steps_per_second": 42.357,
+      "step": 381600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.374632705326321e-05,
+      "loss": 4.213,
+      "step": 381952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.373794110575269e-05,
+      "loss": 4.2076,
+      "step": 382464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372955515824217e-05,
+      "loss": 4.2163,
+      "step": 382976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.372116921073165e-05,
+      "loss": 4.2061,
+      "step": 383488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.371278326322113e-05,
+      "loss": 4.2196,
+      "step": 384000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.370439731571061e-05,
+      "loss": 4.2071,
+      "step": 384512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36960113682001e-05,
+      "loss": 4.199,
+      "step": 385024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.368762542068958e-05,
+      "loss": 4.1959,
+      "step": 385536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367923947317906e-05,
+      "loss": 4.2133,
+      "step": 386048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.367085352566854e-05,
+      "loss": 4.2135,
+      "step": 386560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.366246757815802e-05,
+      "loss": 4.2077,
+      "step": 387072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36540816306475e-05,
+      "loss": 4.2137,
+      "step": 387584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.364571206194071e-05,
+      "loss": 4.2048,
+      "step": 388096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.363732611443019e-05,
+      "loss": 4.208,
+      "step": 388608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362894016691967e-05,
+      "loss": 4.2027,
+      "step": 389120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.362055421940915e-05,
+      "loss": 4.1946,
+      "step": 389632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.361216827189863e-05,
+      "loss": 4.2046,
+      "step": 390144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.36037823243881e-05,
+      "loss": 4.1962,
+      "step": 390656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.359541275568132e-05,
+      "loss": 4.2088,
+      "step": 391168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35870268081708e-05,
+      "loss": 4.2123,
+      "step": 391680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3578640860660284e-05,
+      "loss": 4.2163,
+      "step": 392192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.357025491314976e-05,
+      "loss": 4.2038,
+      "step": 392704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.356186896563924e-05,
+      "loss": 4.2002,
+      "step": 393216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.355348301812872e-05,
+      "loss": 4.2054,
+      "step": 393728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.35450970706182e-05,
+      "loss": 4.1949,
+      "step": 394240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3536711123107677e-05,
+      "loss": 4.2005,
+      "step": 394752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3528341554400886e-05,
+      "loss": 4.1853,
+      "step": 395264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3519955606890366e-05,
+      "loss": 4.2082,
+      "step": 395776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3511569659379846e-05,
+      "loss": 4.1943,
+      "step": 396288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.3503183711869326e-05,
+      "loss": 4.1892,
+      "step": 396800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3494797764358805e-05,
+      "loss": 4.1988,
+      "step": 397312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3486411816848285e-05,
+      "loss": 4.2029,
+      "step": 397824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3478025869337765e-05,
+      "loss": 4.2062,
+      "step": 398336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346963992182725e-05,
+      "loss": 4.2032,
+      "step": 398848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.346128673192419e-05,
+      "loss": 4.1959,
+      "step": 399360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.345290078441367e-05,
+      "loss": 4.2044,
+      "step": 399872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.344451483690315e-05,
+      "loss": 4.2007,
+      "step": 400384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.343612888939263e-05,
+      "loss": 4.1836,
+      "step": 400896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.342774294188211e-05,
+      "loss": 4.1859,
+      "step": 401408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.341937337317532e-05,
+      "loss": 4.1862,
+      "step": 401920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.34109874256648e-05,
+      "loss": 4.1917,
+      "step": 402432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.340260147815428e-05,
+      "loss": 4.1989,
+      "step": 402944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.339421553064376e-05,
+      "loss": 4.1915,
+      "step": 403456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3385845961936975e-05,
+      "loss": 4.1848,
+      "step": 403968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3377460014426455e-05,
+      "loss": 4.1978,
+      "step": 404480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3369074066915935e-05,
+      "loss": 4.19,
+      "step": 404992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3360688119405415e-05,
+      "loss": 4.206,
+      "step": 405504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3352318550698624e-05,
+      "loss": 4.186,
+      "step": 406016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3343932603188104e-05,
+      "loss": 4.1662,
+      "step": 406528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3335546655677584e-05,
+      "loss": 4.2046,
+      "step": 407040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3327160708167064e-05,
+      "loss": 4.197,
+      "step": 407552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3318774760656544e-05,
+      "loss": 4.2001,
+      "step": 408064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3310388813146024e-05,
+      "loss": 4.1895,
+      "step": 408576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3302002865635504e-05,
+      "loss": 4.1851,
+      "step": 409088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3293616918124984e-05,
+      "loss": 4.1842,
+      "step": 409600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.328524734941819e-05,
+      "loss": 4.188,
+      "step": 410112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.327686140190767e-05,
+      "loss": 4.1808,
+      "step": 410624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326849183320089e-05,
+      "loss": 4.1836,
+      "step": 411136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.326010588569037e-05,
+      "loss": 4.1911,
+      "step": 411648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.325171993817985e-05,
+      "loss": 4.1855,
+      "step": 412160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.324333399066933e-05,
+      "loss": 4.1761,
+      "step": 412672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.323494804315881e-05,
+      "loss": 4.1948,
+      "step": 413184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.322657847445202e-05,
+      "loss": 4.1759,
+      "step": 413696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.32181925269415e-05,
+      "loss": 4.1774,
+      "step": 414208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320980657943098e-05,
+      "loss": 4.1915,
+      "step": 414720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.320142063192046e-05,
+      "loss": 4.1924,
+      "step": 415232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.319303468440994e-05,
+      "loss": 4.173,
+      "step": 415744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.318464873689941e-05,
+      "loss": 4.1822,
+      "step": 416256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.317626278938889e-05,
+      "loss": 4.1651,
+      "step": 416768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3167893220682106e-05,
+      "loss": 4.1861,
+      "step": 417280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3159507273171586e-05,
+      "loss": 4.187,
+      "step": 417792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3151121325661066e-05,
+      "loss": 4.1846,
+      "step": 418304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3142735378150546e-05,
+      "loss": 4.1847,
+      "step": 418816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3134349430640026e-05,
+      "loss": 4.1923,
+      "step": 419328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3125963483129506e-05,
+      "loss": 4.198,
+      "step": 419840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3117577535618986e-05,
+      "loss": 4.1733,
+      "step": 420352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3109191588108466e-05,
+      "loss": 4.1893,
+      "step": 420864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3100822019401675e-05,
+      "loss": 4.1911,
+      "step": 421376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3092436071891155e-05,
+      "loss": 4.1652,
+      "step": 421888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3084050124380635e-05,
+      "loss": 4.1904,
+      "step": 422400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.3075664176870115e-05,
+      "loss": 4.1739,
+      "step": 422912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.306729460816333e-05,
+      "loss": 4.1916,
+      "step": 423424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305890866065281e-05,
+      "loss": 4.1905,
+      "step": 423936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.305052271314229e-05,
+      "loss": 4.1737,
+      "step": 424448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.304213676563177e-05,
+      "loss": 4.1727,
+      "step": 424960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.303375081812125e-05,
+      "loss": 4.1779,
+      "step": 425472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.302536487061073e-05,
+      "loss": 4.1916,
+      "step": 425984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.301697892310021e-05,
+      "loss": 4.1776,
+      "step": 426496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.300859297558969e-05,
+      "loss": 4.1786,
+      "step": 427008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.300020702807917e-05,
+      "loss": 4.175,
+      "step": 427520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.299182108056865e-05,
+      "loss": 4.1789,
+      "step": 428032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2983435133058124e-05,
+      "loss": 4.1737,
+      "step": 428544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2975049185547604e-05,
+      "loss": 4.1574,
+      "step": 429056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.296667961684082e-05,
+      "loss": 4.1741,
+      "step": 429568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.29582936693303e-05,
+      "loss": 4.1842,
+      "step": 430080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294990772181978e-05,
+      "loss": 4.1818,
+      "step": 430592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.294152177430926e-05,
+      "loss": 4.1667,
+      "step": 431104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2933168584406204e-05,
+      "loss": 4.1686,
+      "step": 431616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2924799015699413e-05,
+      "loss": 4.1736,
+      "step": 432128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2916413068188893e-05,
+      "loss": 4.178,
+      "step": 432640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.290802712067837e-05,
+      "loss": 4.1837,
+      "step": 433152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289964117316785e-05,
+      "loss": 4.1663,
+      "step": 433664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.289125522565733e-05,
+      "loss": 4.1727,
+      "step": 434176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.288286927814681e-05,
+      "loss": 4.1791,
+      "step": 434688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.287448333063629e-05,
+      "loss": 4.1604,
+      "step": 435200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2866097383125766e-05,
+      "loss": 4.1716,
+      "step": 435712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.285771143561525e-05,
+      "loss": 4.162,
+      "step": 436224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284934186690847e-05,
+      "loss": 4.1758,
+      "step": 436736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.284095591939795e-05,
+      "loss": 4.1616,
+      "step": 437248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.283256997188742e-05,
+      "loss": 4.1757,
+      "step": 437760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.28241840243769e-05,
+      "loss": 4.1652,
+      "step": 438272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281579807686638e-05,
+      "loss": 4.1668,
+      "step": 438784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.280741212935586e-05,
+      "loss": 4.1673,
+      "step": 439296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279902618184534e-05,
+      "loss": 4.17,
+      "step": 439808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.279064023433482e-05,
+      "loss": 4.1685,
+      "step": 440320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.278227066562803e-05,
+      "loss": 4.1836,
+      "step": 440832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.277390109692124e-05,
+      "loss": 4.1742,
+      "step": 441344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.276551514941072e-05,
+      "loss": 4.175,
+      "step": 441856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.275712920190021e-05,
+      "loss": 4.1647,
+      "step": 442368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.274874325438969e-05,
+      "loss": 4.1705,
+      "step": 442880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2740373685682896e-05,
+      "loss": 4.1617,
+      "step": 443392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2731987738172376e-05,
+      "loss": 4.1684,
+      "step": 443904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2723601790661856e-05,
+      "loss": 4.1664,
+      "step": 444416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2715215843151336e-05,
+      "loss": 4.1704,
+      "step": 444928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2706829895640816e-05,
+      "loss": 4.1683,
+      "step": 445440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2698443948130296e-05,
+      "loss": 4.1668,
+      "step": 445952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2690058000619776e-05,
+      "loss": 4.1649,
+      "step": 446464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2681672053109256e-05,
+      "loss": 4.1706,
+      "step": 446976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2673286105598735e-05,
+      "loss": 4.1606,
+      "step": 447488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2664900158088215e-05,
+      "loss": 4.1709,
+      "step": 448000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2656514210577695e-05,
+      "loss": 4.1651,
+      "step": 448512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2648128263067175e-05,
+      "loss": 4.1602,
+      "step": 449024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263977507316412e-05,
+      "loss": 4.1475,
+      "step": 449536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.263140550445733e-05,
+      "loss": 4.1712,
+      "step": 450048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.262301955694681e-05,
+      "loss": 4.1739,
+      "step": 450560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.261463360943629e-05,
+      "loss": 4.1706,
+      "step": 451072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260624766192577e-05,
+      "loss": 4.1687,
+      "step": 451584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.259787809321898e-05,
+      "loss": 4.1687,
+      "step": 452096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258949214570846e-05,
+      "loss": 4.1631,
+      "step": 452608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.258110619819794e-05,
+      "loss": 4.1769,
+      "step": 453120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.257272025068742e-05,
+      "loss": 4.1553,
+      "step": 453632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.25643343031769e-05,
+      "loss": 4.1656,
+      "step": 454144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.255594835566638e-05,
+      "loss": 4.1617,
+      "step": 454656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.254756240815586e-05,
+      "loss": 4.1605,
+      "step": 455168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.253917646064534e-05,
+      "loss": 4.1709,
+      "step": 455680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2530806891938554e-05,
+      "loss": 4.1609,
+      "step": 456192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2522420944428034e-05,
+      "loss": 4.1675,
+      "step": 456704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2514034996917514e-05,
+      "loss": 4.1573,
+      "step": 457216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.2505649049406994e-05,
+      "loss": 4.1637,
+      "step": 457728
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.173742294311523,
+      "eval_runtime": 300.335,
+      "eval_samples_per_second": 1270.551,
+      "eval_steps_per_second": 39.706,
+      "step": 457920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2497263101896474e-05,
+      "loss": 4.1633,
+      "step": 458240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2488877154385954e-05,
+      "loss": 4.1597,
+      "step": 458752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2480491206875434e-05,
+      "loss": 4.1653,
+      "step": 459264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.247210525936491e-05,
+      "loss": 4.1575,
+      "step": 459776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.246371931185439e-05,
+      "loss": 4.1714,
+      "step": 460288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.245533336434387e-05,
+      "loss": 4.1594,
+      "step": 460800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.244694741683335e-05,
+      "loss": 4.1546,
+      "step": 461312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243857784812656e-05,
+      "loss": 4.1477,
+      "step": 461824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.243019190061604e-05,
+      "loss": 4.1664,
+      "step": 462336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.242180595310552e-05,
+      "loss": 4.1663,
+      "step": 462848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2413420005595e-05,
+      "loss": 4.1587,
+      "step": 463360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.240503405808448e-05,
+      "loss": 4.1665,
+      "step": 463872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.239666448937769e-05,
+      "loss": 4.1576,
+      "step": 464384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.238827854186717e-05,
+      "loss": 4.1598,
+      "step": 464896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237989259435665e-05,
+      "loss": 4.1507,
+      "step": 465408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.237150664684613e-05,
+      "loss": 4.1486,
+      "step": 465920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.236312069933561e-05,
+      "loss": 4.1555,
+      "step": 466432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.235473475182509e-05,
+      "loss": 4.1517,
+      "step": 466944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.234634880431457e-05,
+      "loss": 4.1584,
+      "step": 467456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.233796285680405e-05,
+      "loss": 4.1644,
+      "step": 467968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232959328809726e-05,
+      "loss": 4.1701,
+      "step": 468480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.232120734058675e-05,
+      "loss": 4.1577,
+      "step": 468992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.231282139307623e-05,
+      "loss": 4.1543,
+      "step": 469504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.230443544556571e-05,
+      "loss": 4.1556,
+      "step": 470016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2296065876858916e-05,
+      "loss": 4.1516,
+      "step": 470528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2287679929348396e-05,
+      "loss": 4.1527,
+      "step": 471040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2279293981837876e-05,
+      "loss": 4.136,
+      "step": 471552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2270908034327356e-05,
+      "loss": 4.1638,
+      "step": 472064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2262538465620565e-05,
+      "loss": 4.1505,
+      "step": 472576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.2254152518110045e-05,
+      "loss": 4.145,
+      "step": 473088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2245766570599525e-05,
+      "loss": 4.1505,
+      "step": 473600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2237380623089005e-05,
+      "loss": 4.1576,
+      "step": 474112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2228994675578485e-05,
+      "loss": 4.1546,
+      "step": 474624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.22206251068717e-05,
+      "loss": 4.1594,
+      "step": 475136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.221223915936118e-05,
+      "loss": 4.1488,
+      "step": 475648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.220386959065439e-05,
+      "loss": 4.1588,
+      "step": 476160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.219548364314387e-05,
+      "loss": 4.1537,
+      "step": 476672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.218709769563335e-05,
+      "loss": 4.1379,
+      "step": 477184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217871174812283e-05,
+      "loss": 4.1408,
+      "step": 477696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.217032580061231e-05,
+      "loss": 4.1413,
+      "step": 478208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.216195623190552e-05,
+      "loss": 4.1466,
+      "step": 478720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.215358666319873e-05,
+      "loss": 4.1563,
+      "step": 479232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.214520071568821e-05,
+      "loss": 4.1444,
+      "step": 479744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.213681476817769e-05,
+      "loss": 4.1438,
+      "step": 480256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212842882066717e-05,
+      "loss": 4.1489,
+      "step": 480768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.212004287315665e-05,
+      "loss": 4.1441,
+      "step": 481280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2111656925646134e-05,
+      "loss": 4.1609,
+      "step": 481792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2103270978135614e-05,
+      "loss": 4.143,
+      "step": 482304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2094885030625094e-05,
+      "loss": 4.1177,
+      "step": 482816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.208649908311457e-05,
+      "loss": 4.1589,
+      "step": 483328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.207811313560405e-05,
+      "loss": 4.154,
+      "step": 483840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.206974356689726e-05,
+      "loss": 4.1531,
+      "step": 484352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2061357619386736e-05,
+      "loss": 4.1426,
+      "step": 484864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2052971671876216e-05,
+      "loss": 4.1408,
+      "step": 485376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2044585724365696e-05,
+      "loss": 4.1396,
+      "step": 485888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2036199776855176e-05,
+      "loss": 4.1451,
+      "step": 486400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2027813829344656e-05,
+      "loss": 4.137,
+      "step": 486912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2019427881834136e-05,
+      "loss": 4.1361,
+      "step": 487424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.2011041934323616e-05,
+      "loss": 4.1458,
+      "step": 487936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.20026559868131e-05,
+      "loss": 4.1437,
+      "step": 488448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.199428641810631e-05,
+      "loss": 4.1297,
+      "step": 488960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.198590047059579e-05,
+      "loss": 4.1519,
+      "step": 489472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.197751452308527e-05,
+      "loss": 4.1297,
+      "step": 489984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196912857557475e-05,
+      "loss": 4.1365,
+      "step": 490496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.196075900686796e-05,
+      "loss": 4.1477,
+      "step": 491008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.195238943816117e-05,
+      "loss": 4.1472,
+      "step": 491520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.194400349065065e-05,
+      "loss": 4.1279,
+      "step": 492032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.193561754314013e-05,
+      "loss": 4.1369,
+      "step": 492544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.192723159562961e-05,
+      "loss": 4.1221,
+      "step": 493056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191884564811909e-05,
+      "loss": 4.1432,
+      "step": 493568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.191045970060857e-05,
+      "loss": 4.1441,
+      "step": 494080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1902073753098057e-05,
+      "loss": 4.1426,
+      "step": 494592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1893687805587537e-05,
+      "loss": 4.1383,
+      "step": 495104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1885318236880746e-05,
+      "loss": 4.1457,
+      "step": 495616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1876948668173955e-05,
+      "loss": 4.1557,
+      "step": 496128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1868562720663435e-05,
+      "loss": 4.1324,
+      "step": 496640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1860176773152915e-05,
+      "loss": 4.1436,
+      "step": 497152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1851790825642395e-05,
+      "loss": 4.1472,
+      "step": 497664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1843404878131875e-05,
+      "loss": 4.1292,
+      "step": 498176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1835018930621355e-05,
+      "loss": 4.1428,
+      "step": 498688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1826649361914564e-05,
+      "loss": 4.1379,
+      "step": 499200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1818263414404044e-05,
+      "loss": 4.1447,
+      "step": 499712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.1809877466893524e-05,
+      "loss": 4.1472,
+      "step": 500224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.180149151938301e-05,
+      "loss": 4.1305,
+      "step": 500736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.179310557187249e-05,
+      "loss": 4.1301,
+      "step": 501248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.178471962436197e-05,
+      "loss": 4.136,
+      "step": 501760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.177635005565518e-05,
+      "loss": 4.1433,
+      "step": 502272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.176796410814466e-05,
+      "loss": 4.1364,
+      "step": 502784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.175957816063414e-05,
+      "loss": 4.1345,
+      "step": 503296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.175119221312362e-05,
+      "loss": 4.1351,
+      "step": 503808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.17428062656131e-05,
+      "loss": 4.1392,
+      "step": 504320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.173442031810257e-05,
+      "loss": 4.1305,
+      "step": 504832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.172603437059205e-05,
+      "loss": 4.116,
+      "step": 505344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.171764842308153e-05,
+      "loss": 4.1342,
+      "step": 505856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170929523317848e-05,
+      "loss": 4.1379,
+      "step": 506368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.170092566447169e-05,
+      "loss": 4.1389,
+      "step": 506880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.169253971696117e-05,
+      "loss": 4.1252,
+      "step": 507392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.168415376945065e-05,
+      "loss": 4.1245,
+      "step": 507904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.167576782194013e-05,
+      "loss": 4.133,
+      "step": 508416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.166738187442961e-05,
+      "loss": 4.1335,
+      "step": 508928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165899592691909e-05,
+      "loss": 4.1437,
+      "step": 509440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.165060997940857e-05,
+      "loss": 4.1263,
+      "step": 509952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.164224041070178e-05,
+      "loss": 4.1319,
+      "step": 510464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.163385446319126e-05,
+      "loss": 4.135,
+      "step": 510976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.162546851568074e-05,
+      "loss": 4.1188,
+      "step": 511488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.161708256817022e-05,
+      "loss": 4.1323,
+      "step": 512000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.16086966206597e-05,
+      "loss": 4.1164,
+      "step": 512512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.160031067314918e-05,
+      "loss": 4.1392,
+      "step": 513024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.159192472563866e-05,
+      "loss": 4.1189,
+      "step": 513536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.158353877812814e-05,
+      "loss": 4.1336,
+      "step": 514048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.157516920942135e-05,
+      "loss": 4.1261,
+      "step": 514560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.156678326191083e-05,
+      "loss": 4.1253,
+      "step": 515072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155839731440031e-05,
+      "loss": 4.1287,
+      "step": 515584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.155001136688979e-05,
+      "loss": 4.13,
+      "step": 516096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.154162541937927e-05,
+      "loss": 4.127,
+      "step": 516608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.153325585067248e-05,
+      "loss": 4.142,
+      "step": 517120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1524886281965695e-05,
+      "loss": 4.1274,
+      "step": 517632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.151650033445517e-05,
+      "loss": 4.1404,
+      "step": 518144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1508114386944655e-05,
+      "loss": 4.1253,
+      "step": 518656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1499728439434135e-05,
+      "loss": 4.126,
+      "step": 519168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1491342491923615e-05,
+      "loss": 4.1265,
+      "step": 519680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1482956544413095e-05,
+      "loss": 4.1247,
+      "step": 520192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1474570596902575e-05,
+      "loss": 4.1244,
+      "step": 520704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1466184649392055e-05,
+      "loss": 4.1315,
+      "step": 521216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1457798701881535e-05,
+      "loss": 4.126,
+      "step": 521728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1449412754371015e-05,
+      "loss": 4.1318,
+      "step": 522240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1441026806860495e-05,
+      "loss": 4.125,
+      "step": 522752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1432657238153704e-05,
+      "loss": 4.1301,
+      "step": 523264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1424271290643184e-05,
+      "loss": 4.1238,
+      "step": 523776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1415885343132664e-05,
+      "loss": 4.127,
+      "step": 524288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1407499395622144e-05,
+      "loss": 4.1227,
+      "step": 524800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1399113448111624e-05,
+      "loss": 4.124,
+      "step": 525312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1390727500601104e-05,
+      "loss": 4.1059,
+      "step": 525824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.1382341553090584e-05,
+      "loss": 4.1322,
+      "step": 526336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.13739719843838e-05,
+      "loss": 4.1317,
+      "step": 526848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.136558603687328e-05,
+      "loss": 4.1331,
+      "step": 527360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.135720008936276e-05,
+      "loss": 4.1267,
+      "step": 527872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134881414185223e-05,
+      "loss": 4.1331,
+      "step": 528384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.134042819434171e-05,
+      "loss": 4.1262,
+      "step": 528896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.133204224683119e-05,
+      "loss": 4.1349,
+      "step": 529408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.132365629932067e-05,
+      "loss": 4.1152,
+      "step": 529920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.131527035181015e-05,
+      "loss": 4.1223,
+      "step": 530432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.130690078310336e-05,
+      "loss": 4.1243,
+      "step": 530944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129851483559284e-05,
+      "loss": 4.1242,
+      "step": 531456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.129012888808232e-05,
+      "loss": 4.1276,
+      "step": 531968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.12817429405718e-05,
+      "loss": 4.1247,
+      "step": 532480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.127335699306129e-05,
+      "loss": 4.1258,
+      "step": 532992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.126497104555077e-05,
+      "loss": 4.1193,
+      "step": 533504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.125658509804025e-05,
+      "loss": 4.1288,
+      "step": 534016
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.1431803703308105,
+      "eval_runtime": 295.3123,
+      "eval_samples_per_second": 1292.161,
+      "eval_steps_per_second": 40.381,
+      "step": 534240
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 2.1183549595024416e+17,
+  "trial_name": null,
+  "trial_params": null
+}