diff --git "a/checkpoint-2442224/trainer_state.json" "b/checkpoint-2442224/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2442224/trainer_state.json"
@@ -0,0 +1,28895 @@
+{
+  "best_metric": 3.9891669750213623,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/rel-cl/lstm/3/checkpoints/checkpoint-2442224",
+  "epoch": 1.0250006060157382,
+  "eval_steps": 10,
+  "global_step": 2442224,
+  "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": 10.8198,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5697,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0565,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.9781,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.9507,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8456,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994968431493688e-05,
+      "loss": 6.7236,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.6208,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5339,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4398,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3912,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3226,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.2576,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.191,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1427,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.105,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0659,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 6.0125,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9768,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9418,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.9049,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8781,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8498,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9807172343669236e-05,
+      "loss": 5.8175,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798786396158716e-05,
+      "loss": 5.7889,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790400448648195e-05,
+      "loss": 5.7723,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9782014501137675e-05,
+      "loss": 5.7401,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773628553627155e-05,
+      "loss": 5.7195,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9765242606116635e-05,
+      "loss": 5.6842,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.6802,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6467,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.6311,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.6193,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.602,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5709,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5577,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5515,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.5266,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.5154,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4951,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4707,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4599,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964788847738054e-05,
+      "loss": 5.4448,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963950252987002e-05,
+      "loss": 5.4383,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.4263,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.4202,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3964,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960595873982795e-05,
+      "loss": 5.3959,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959757279231743e-05,
+      "loss": 5.3996,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958918684480691e-05,
+      "loss": 5.3719,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958080089729639e-05,
+      "loss": 5.3628,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.957241494978587e-05,
+      "loss": 5.3496,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956402900227535e-05,
+      "loss": 5.359,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.3201,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.3384,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953888753854752e-05,
+      "loss": 5.3088,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530501591037e-05,
+      "loss": 5.3052,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.952211564352648e-05,
+      "loss": 5.2998,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.951372969601596e-05,
+      "loss": 5.2898,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95053765061129e-05,
+      "loss": 5.277,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.949699055860238e-05,
+      "loss": 5.2662,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948860461109186e-05,
+      "loss": 5.251,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2581,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.245,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.2357,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.2293,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.944669125234299e-05,
+      "loss": 5.2185,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.943830530483247e-05,
+      "loss": 5.2008,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942991935732195e-05,
+      "loss": 5.2185,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1957,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.1875,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.940477789359412e-05,
+      "loss": 5.1983,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.93963919460836e-05,
+      "loss": 5.1737,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1873,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1776,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9371250482355776e-05,
+      "loss": 5.1662,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1671,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.141,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.1373,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.933770669231369e-05,
+      "loss": 5.1362,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.1354,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.1243,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.1112,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0992,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.105,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287423764858034e-05,
+      "loss": 5.1043,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927905419615125e-05,
+      "loss": 5.1007,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.927066824864073e-05,
+      "loss": 5.0981,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.92622823011302e-05,
+      "loss": 5.0934,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.925389635361968e-05,
+      "loss": 5.0793,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.924551040610916e-05,
+      "loss": 5.0627,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.923712445859864e-05,
+      "loss": 5.0591,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922873851108812e-05,
+      "loss": 5.0652,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.92203525635776e-05,
+      "loss": 5.0622,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921198299487081e-05,
+      "loss": 5.0404,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920359704736029e-05,
+      "loss": 5.0438,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919521109984978e-05,
+      "loss": 5.042,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918682515233926e-05,
+      "loss": 5.0424,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917843920482874e-05,
+      "loss": 5.0343,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917005325731822e-05,
+      "loss": 5.034,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916168368861143e-05,
+      "loss": 5.0269,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915329774110091e-05,
+      "loss": 5.0175,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914491179359039e-05,
+      "loss": 5.0164,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913652584607987e-05,
+      "loss": 5.0178,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912813989856935e-05,
+      "loss": 5.0013,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9119770329862556e-05,
+      "loss": 5.0064,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9111384382352036e-05,
+      "loss": 4.9835,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9102998434841516e-05,
+      "loss": 4.9785,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.909462886613473e-05,
+      "loss": 4.9778,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9824,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9744,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9516,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9571,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.9606,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9677,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9534,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.9526,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.9496,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.9506,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002383443519005e-05,
+      "loss": 4.9337,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8993997496008485e-05,
+      "loss": 4.9347,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.9256,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897725835859491e-05,
+      "loss": 4.9299,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8968872411084396e-05,
+      "loss": 4.9112,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8960502842377605e-05,
+      "loss": 4.9247,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8952116894867085e-05,
+      "loss": 4.9142,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8943730947356565e-05,
+      "loss": 4.9212,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893534499984604e-05,
+      "loss": 4.8942,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892695905233552e-05,
+      "loss": 4.9051,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8918573104825e-05,
+      "loss": 4.9138,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891018715731448e-05,
+      "loss": 4.8863,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890180120980396e-05,
+      "loss": 4.905,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889343164109717e-05,
+      "loss": 4.8891,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888506207239038e-05,
+      "loss": 4.8928,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.887667612487986e-05,
+      "loss": 4.8827,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886829017736934e-05,
+      "loss": 4.8781,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.885990422985882e-05,
+      "loss": 4.8776,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88515182823483e-05,
+      "loss": 4.877,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.884313233483778e-05,
+      "loss": 4.8761,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.883474638732726e-05,
+      "loss": 4.8701,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.882636043981674e-05,
+      "loss": 4.8734,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.881799087110995e-05,
+      "loss": 4.8711,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880960492359943e-05,
+      "loss": 4.8518,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.880121897608891e-05,
+      "loss": 4.8516,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879283302857839e-05,
+      "loss": 4.8425,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878444708106787e-05,
+      "loss": 4.843,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877606113355735e-05,
+      "loss": 4.8422,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.876767518604683e-05,
+      "loss": 4.8361,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875928923853632e-05,
+      "loss": 4.8333,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875091966982953e-05,
+      "loss": 4.8333,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.795548439025879,
+      "eval_runtime": 544.0651,
+      "eval_samples_per_second": 701.37,
+      "eval_steps_per_second": 21.918,
+      "step": 76319
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.874255010112274e-05,
+      "loss": 4.8402,
+      "step": 76800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.873416415361222e-05,
+      "loss": 4.8339,
+      "step": 77312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.87257782061017e-05,
+      "loss": 4.8225,
+      "step": 77824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.871739225859118e-05,
+      "loss": 4.8244,
+      "step": 78336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8709006311080657e-05,
+      "loss": 4.8184,
+      "step": 78848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8700620363570137e-05,
+      "loss": 4.7956,
+      "step": 79360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8692234416059616e-05,
+      "loss": 4.8082,
+      "step": 79872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8683848468549096e-05,
+      "loss": 4.8113,
+      "step": 80384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8675478899842306e-05,
+      "loss": 4.794,
+      "step": 80896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.866710933113552e-05,
+      "loss": 4.809,
+      "step": 81408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8658723383625e-05,
+      "loss": 4.8119,
+      "step": 81920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.865033743611448e-05,
+      "loss": 4.7967,
+      "step": 82432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.864195148860396e-05,
+      "loss": 4.7893,
+      "step": 82944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.863356554109344e-05,
+      "loss": 4.7912,
+      "step": 83456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.862517959358292e-05,
+      "loss": 4.7912,
+      "step": 83968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.86167936460724e-05,
+      "loss": 4.7857,
+      "step": 84480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8608407698561874e-05,
+      "loss": 4.7809,
+      "step": 84992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.860003812985509e-05,
+      "loss": 4.7789,
+      "step": 85504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.859165218234457e-05,
+      "loss": 4.7748,
+      "step": 86016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.858326623483405e-05,
+      "loss": 4.7777,
+      "step": 86528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.857488028732352e-05,
+      "loss": 4.7774,
+      "step": 87040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.856649433981301e-05,
+      "loss": 4.7764,
+      "step": 87552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8558141149909955e-05,
+      "loss": 4.7717,
+      "step": 88064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8549755202399435e-05,
+      "loss": 4.7699,
+      "step": 88576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8541369254888915e-05,
+      "loss": 4.7636,
+      "step": 89088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8532983307378395e-05,
+      "loss": 4.7623,
+      "step": 89600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8524597359867875e-05,
+      "loss": 4.7636,
+      "step": 90112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8516227791161084e-05,
+      "loss": 4.7462,
+      "step": 90624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.8507841843650564e-05,
+      "loss": 4.759,
+      "step": 91136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8499455896140044e-05,
+      "loss": 4.7402,
+      "step": 91648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8491069948629524e-05,
+      "loss": 4.7451,
+      "step": 92160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8482684001119e-05,
+      "loss": 4.7549,
+      "step": 92672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.847429805360848e-05,
+      "loss": 4.7518,
+      "step": 93184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8465912106097964e-05,
+      "loss": 4.7378,
+      "step": 93696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8457526158587444e-05,
+      "loss": 4.73,
+      "step": 94208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844915658988065e-05,
+      "loss": 4.7457,
+      "step": 94720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.844077064237013e-05,
+      "loss": 4.7342,
+      "step": 95232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.843238469485961e-05,
+      "loss": 4.7334,
+      "step": 95744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.842399874734909e-05,
+      "loss": 4.7285,
+      "step": 96256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.841561279983857e-05,
+      "loss": 4.7155,
+      "step": 96768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.840722685232805e-05,
+      "loss": 4.7124,
+      "step": 97280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839885728362126e-05,
+      "loss": 4.7141,
+      "step": 97792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.839048771491447e-05,
+      "loss": 4.7171,
+      "step": 98304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.838210176740395e-05,
+      "loss": 4.7114,
+      "step": 98816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.837371581989343e-05,
+      "loss": 4.7143,
+      "step": 99328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.836532987238292e-05,
+      "loss": 4.7056,
+      "step": 99840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.83569439248724e-05,
+      "loss": 4.7093,
+      "step": 100352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834855797736188e-05,
+      "loss": 4.7218,
+      "step": 100864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.834017202985136e-05,
+      "loss": 4.7005,
+      "step": 101376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.833178608234084e-05,
+      "loss": 4.7018,
+      "step": 101888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8323416513634046e-05,
+      "loss": 4.6924,
+      "step": 102400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8315030566123526e-05,
+      "loss": 4.7133,
+      "step": 102912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8306644618613006e-05,
+      "loss": 4.6825,
+      "step": 103424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8298258671102486e-05,
+      "loss": 4.7046,
+      "step": 103936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8289889102395695e-05,
+      "loss": 4.6823,
+      "step": 104448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8281503154885175e-05,
+      "loss": 4.6899,
+      "step": 104960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8273117207374655e-05,
+      "loss": 4.6916,
+      "step": 105472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8264731259864135e-05,
+      "loss": 4.6847,
+      "step": 105984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8256345312353615e-05,
+      "loss": 4.6776,
+      "step": 106496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8247959364843095e-05,
+      "loss": 4.6765,
+      "step": 107008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823958979613631e-05,
+      "loss": 4.665,
+      "step": 107520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.823120384862579e-05,
+      "loss": 4.6812,
+      "step": 108032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.822281790111527e-05,
+      "loss": 4.6725,
+      "step": 108544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.821443195360475e-05,
+      "loss": 4.6666,
+      "step": 109056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.820604600609423e-05,
+      "loss": 4.665,
+      "step": 109568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.819766005858371e-05,
+      "loss": 4.6632,
+      "step": 110080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.818929048987692e-05,
+      "loss": 4.6467,
+      "step": 110592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.81809045423664e-05,
+      "loss": 4.6717,
+      "step": 111104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.817251859485588e-05,
+      "loss": 4.6558,
+      "step": 111616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.816414902614909e-05,
+      "loss": 4.6558,
+      "step": 112128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.815576307863857e-05,
+      "loss": 4.6644,
+      "step": 112640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.814737713112805e-05,
+      "loss": 4.6513,
+      "step": 113152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8138991183617535e-05,
+      "loss": 4.6639,
+      "step": 113664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.813060523610701e-05,
+      "loss": 4.6671,
+      "step": 114176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.812221928859649e-05,
+      "loss": 4.657,
+      "step": 114688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.811383334108597e-05,
+      "loss": 4.6585,
+      "step": 115200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.810544739357545e-05,
+      "loss": 4.6408,
+      "step": 115712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.809706144606493e-05,
+      "loss": 4.6366,
+      "step": 116224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808869187735814e-05,
+      "loss": 4.6467,
+      "step": 116736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.808030592984762e-05,
+      "loss": 4.6494,
+      "step": 117248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.80719199823371e-05,
+      "loss": 4.6395,
+      "step": 117760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.806353403482658e-05,
+      "loss": 4.6359,
+      "step": 118272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.8055164466119786e-05,
+      "loss": 4.621,
+      "step": 118784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.804677851860927e-05,
+      "loss": 4.6298,
+      "step": 119296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803839257109875e-05,
+      "loss": 4.6342,
+      "step": 119808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.803000662358823e-05,
+      "loss": 4.6347,
+      "step": 120320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.802163705488144e-05,
+      "loss": 4.6401,
+      "step": 120832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.801326748617466e-05,
+      "loss": 4.6386,
+      "step": 121344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.800489791746787e-05,
+      "loss": 4.6276,
+      "step": 121856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.799651196995735e-05,
+      "loss": 4.6186,
+      "step": 122368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.798812602244683e-05,
+      "loss": 4.6131,
+      "step": 122880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.79797400749363e-05,
+      "loss": 4.6204,
+      "step": 123392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.797135412742578e-05,
+      "loss": 4.6225,
+      "step": 123904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.796296817991526e-05,
+      "loss": 4.6103,
+      "step": 124416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.795458223240474e-05,
+      "loss": 4.6102,
+      "step": 124928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.794619628489423e-05,
+      "loss": 4.617,
+      "step": 125440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.793781033738371e-05,
+      "loss": 4.6213,
+      "step": 125952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792942438987319e-05,
+      "loss": 4.611,
+      "step": 126464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.792103844236267e-05,
+      "loss": 4.6169,
+      "step": 126976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.791265249485215e-05,
+      "loss": 4.6176,
+      "step": 127488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.790426654734163e-05,
+      "loss": 4.6106,
+      "step": 128000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7895880599831107e-05,
+      "loss": 4.6048,
+      "step": 128512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7887511031124316e-05,
+      "loss": 4.6155,
+      "step": 129024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7879125083613796e-05,
+      "loss": 4.6014,
+      "step": 129536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7870739136103276e-05,
+      "loss": 4.6116,
+      "step": 130048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7862353188592756e-05,
+      "loss": 4.5934,
+      "step": 130560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7853967241082236e-05,
+      "loss": 4.5825,
+      "step": 131072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7845581293571715e-05,
+      "loss": 4.5912,
+      "step": 131584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7837195346061195e-05,
+      "loss": 4.5983,
+      "step": 132096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782882577735441e-05,
+      "loss": 4.5906,
+      "step": 132608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.782045620864762e-05,
+      "loss": 4.5767,
+      "step": 133120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.78120702611371e-05,
+      "loss": 4.5782,
+      "step": 133632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.780368431362658e-05,
+      "loss": 4.5868,
+      "step": 134144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.779529836611606e-05,
+      "loss": 4.5985,
+      "step": 134656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.778692879740927e-05,
+      "loss": 4.5848,
+      "step": 135168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777855922870248e-05,
+      "loss": 4.5926,
+      "step": 135680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.777017328119196e-05,
+      "loss": 4.5869,
+      "step": 136192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.776178733368144e-05,
+      "loss": 4.5843,
+      "step": 136704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.775340138617092e-05,
+      "loss": 4.5801,
+      "step": 137216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77450154386604e-05,
+      "loss": 4.578,
+      "step": 137728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.773662949114988e-05,
+      "loss": 4.5731,
+      "step": 138240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.772824354363936e-05,
+      "loss": 4.5836,
+      "step": 138752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7719857596128845e-05,
+      "loss": 4.5684,
+      "step": 139264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.771147164861832e-05,
+      "loss": 4.5709,
+      "step": 139776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.77030857011078e-05,
+      "loss": 4.5757,
+      "step": 140288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.769469975359728e-05,
+      "loss": 4.5813,
+      "step": 140800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7686330184890494e-05,
+      "loss": 4.5595,
+      "step": 141312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.767794423737997e-05,
+      "loss": 4.5631,
+      "step": 141824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766955828986945e-05,
+      "loss": 4.5845,
+      "step": 142336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.766118872116266e-05,
+      "loss": 4.5585,
+      "step": 142848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7652802773652136e-05,
+      "loss": 4.5777,
+      "step": 143360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7644416826141616e-05,
+      "loss": 4.5616,
+      "step": 143872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.7636030878631096e-05,
+      "loss": 4.5737,
+      "step": 144384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.762764493112058e-05,
+      "loss": 4.5616,
+      "step": 144896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761925898361006e-05,
+      "loss": 4.5569,
+      "step": 145408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.761087303609954e-05,
+      "loss": 4.5642,
+      "step": 145920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.760248708858902e-05,
+      "loss": 4.5605,
+      "step": 146432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75941011410785e-05,
+      "loss": 4.5677,
+      "step": 146944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.758571519356798e-05,
+      "loss": 4.5636,
+      "step": 147456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.757732924605746e-05,
+      "loss": 4.5637,
+      "step": 147968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756894329854694e-05,
+      "loss": 4.5606,
+      "step": 148480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.756055735103642e-05,
+      "loss": 4.5457,
+      "step": 148992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75521714035259e-05,
+      "loss": 4.5547,
+      "step": 149504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.754380183481911e-05,
+      "loss": 4.5427,
+      "step": 150016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.753543226611232e-05,
+      "loss": 4.5473,
+      "step": 150528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.75270463186018e-05,
+      "loss": 4.5498,
+      "step": 151040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751866037109128e-05,
+      "loss": 4.5373,
+      "step": 151552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.751027442358077e-05,
+      "loss": 4.5428,
+      "step": 152064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.750188847607025e-05,
+      "loss": 4.5414,
+      "step": 152576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.506901264190674,
+      "eval_runtime": 550.7692,
+      "eval_samples_per_second": 692.833,
+      "eval_steps_per_second": 21.652,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.749350252855973e-05,
+      "loss": 4.5539,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748511658104921e-05,
+      "loss": 4.5484,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747673063353868e-05,
+      "loss": 4.5387,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746834468602816e-05,
+      "loss": 4.5383,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745995873851764e-05,
+      "loss": 4.537,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745157279100712e-05,
+      "loss": 4.5196,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74431868434966e-05,
+      "loss": 4.5351,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743480089598608e-05,
+      "loss": 4.5328,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.742641494847556e-05,
+      "loss": 4.5191,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.741804537976877e-05,
+      "loss": 4.5386,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.740965943225825e-05,
+      "loss": 4.5434,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401273484747736e-05,
+      "loss": 4.527,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392887537237216e-05,
+      "loss": 4.52,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384501589726696e-05,
+      "loss": 4.5234,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7376115642216176e-05,
+      "loss": 4.5271,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7367729694705656e-05,
+      "loss": 4.522,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359343747195135e-05,
+      "loss": 4.5184,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350957799684615e-05,
+      "loss": 4.5169,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342571852174095e-05,
+      "loss": 4.5184,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.733418590466357e-05,
+      "loss": 4.5209,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.732579995715305e-05,
+      "loss": 4.5215,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.731741400964253e-05,
+      "loss": 4.5245,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730904444093574e-05,
+      "loss": 4.523,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.730065849342522e-05,
+      "loss": 4.5148,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292272545914704e-05,
+      "loss": 4.5182,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283886598404184e-05,
+      "loss": 4.5145,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275500650893664e-05,
+      "loss": 4.5158,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267114703383144e-05,
+      "loss": 4.5015,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258728755872624e-05,
+      "loss": 4.5128,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.725035918716583e-05,
+      "loss": 4.4983,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724197323965531e-05,
+      "loss": 4.504,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.723358729214479e-05,
+      "loss": 4.515,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722520134463427e-05,
+      "loss": 4.5129,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721683177592748e-05,
+      "loss": 4.4999,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720844582841696e-05,
+      "loss": 4.4943,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720005988090644e-05,
+      "loss": 4.5145,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719169031219965e-05,
+      "loss": 4.4952,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718330436468914e-05,
+      "loss": 4.5038,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717491841717862e-05,
+      "loss": 4.5005,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.71665324696681e-05,
+      "loss": 4.485,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715814652215758e-05,
+      "loss": 4.4808,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714976057464706e-05,
+      "loss": 4.4846,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714139100594027e-05,
+      "loss": 4.4987,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.713300505842975e-05,
+      "loss": 4.4889,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.712461911091923e-05,
+      "loss": 4.4898,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711623316340871e-05,
+      "loss": 4.4817,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710784721589819e-05,
+      "loss": 4.486,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7099477647191396e-05,
+      "loss": 4.5017,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7091091699680876e-05,
+      "loss": 4.4805,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7082705752170356e-05,
+      "loss": 4.4806,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7074319804659836e-05,
+      "loss": 4.4764,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706593385714932e-05,
+      "loss": 4.4985,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70575479096388e-05,
+      "loss": 4.4686,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704916196212828e-05,
+      "loss": 4.4907,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7040776014617755e-05,
+      "loss": 4.4668,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7032390067107235e-05,
+      "loss": 4.4801,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7024004119596715e-05,
+      "loss": 4.482,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7015618172086195e-05,
+      "loss": 4.4706,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7007248603379404e-05,
+      "loss": 4.4755,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6998862655868884e-05,
+      "loss": 4.4693,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990476708358364e-05,
+      "loss": 4.4533,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982090760847844e-05,
+      "loss": 4.4732,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973704813337324e-05,
+      "loss": 4.4721,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696533524463054e-05,
+      "loss": 4.4612,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.695694929712002e-05,
+      "loss": 4.4633,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69485633496095e-05,
+      "loss": 4.4634,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.694017740209898e-05,
+      "loss": 4.4451,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.693180783339219e-05,
+      "loss": 4.4735,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.692342188588167e-05,
+      "loss": 4.4524,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.691503593837115e-05,
+      "loss": 4.4575,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.690664999086063e-05,
+      "loss": 4.4682,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.689828042215384e-05,
+      "loss": 4.4574,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.688989447464332e-05,
+      "loss": 4.4666,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68815085271328e-05,
+      "loss": 4.4752,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.687312257962228e-05,
+      "loss": 4.4617,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.686473663211176e-05,
+      "loss": 4.4696,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856350684601245e-05,
+      "loss": 4.451,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6847964737090725e-05,
+      "loss": 4.4434,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6839578789580204e-05,
+      "loss": 4.4552,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683122559967714e-05,
+      "loss": 4.4634,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.682283965216662e-05,
+      "loss": 4.4534,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68144537046561e-05,
+      "loss": 4.4454,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680606775714558e-05,
+      "loss": 4.4374,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679768180963506e-05,
+      "loss": 4.4429,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678929586212454e-05,
+      "loss": 4.444,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678090991461402e-05,
+      "loss": 4.4483,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.67725239671035e-05,
+      "loss": 4.4559,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676413801959298e-05,
+      "loss": 4.4545,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675576845088619e-05,
+      "loss": 4.4447,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674738250337568e-05,
+      "loss": 4.439,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673899655586516e-05,
+      "loss": 4.4342,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673061060835464e-05,
+      "loss": 4.4382,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672222466084412e-05,
+      "loss": 4.4392,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671385509213733e-05,
+      "loss": 4.4351,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670546914462681e-05,
+      "loss": 4.4324,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669708319711629e-05,
+      "loss": 4.4403,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6688713628409496e-05,
+      "loss": 4.4413,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6680327680898976e-05,
+      "loss": 4.4411,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6671941733388456e-05,
+      "loss": 4.4408,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663555785877936e-05,
+      "loss": 4.4442,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655169838367416e-05,
+      "loss": 4.4374,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.664680026966063e-05,
+      "loss": 4.4297,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663841432215011e-05,
+      "loss": 4.4464,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.663002837463959e-05,
+      "loss": 4.4251,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6621642427129065e-05,
+      "loss": 4.4462,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6613256479618545e-05,
+      "loss": 4.42,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6604870532108025e-05,
+      "loss": 4.4131,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6596484584597505e-05,
+      "loss": 4.4216,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6588098637086985e-05,
+      "loss": 4.4361,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579712689576465e-05,
+      "loss": 4.4252,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657135949967341e-05,
+      "loss": 4.4115,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.656297355216288e-05,
+      "loss": 4.4123,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.655458760465237e-05,
+      "loss": 4.4192,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654620165714185e-05,
+      "loss": 4.4349,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653781570963133e-05,
+      "loss": 4.4194,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652942976212081e-05,
+      "loss": 4.4293,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.652104381461029e-05,
+      "loss": 4.4242,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.651265786709977e-05,
+      "loss": 4.4205,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.650427191958925e-05,
+      "loss": 4.4241,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.649590235088246e-05,
+      "loss": 4.4128,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.648751640337194e-05,
+      "loss": 4.4142,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647913045586142e-05,
+      "loss": 4.4231,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64707445083509e-05,
+      "loss": 4.4087,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646237493964411e-05,
+      "loss": 4.4145,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645398899213359e-05,
+      "loss": 4.4168,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644560304462307e-05,
+      "loss": 4.4216,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6437217097112554e-05,
+      "loss": 4.4043,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6428831149602034e-05,
+      "loss": 4.4042,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642046158089524e-05,
+      "loss": 4.4295,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641207563338472e-05,
+      "loss": 4.4058,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.64036896858742e-05,
+      "loss": 4.4195,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639530373836368e-05,
+      "loss": 4.4124,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.638691779085316e-05,
+      "loss": 4.4183,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637853184334264e-05,
+      "loss": 4.4079,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637014589583212e-05,
+      "loss": 4.406,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6361759948321596e-05,
+      "loss": 4.4151,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6353374000811076e-05,
+      "loss": 4.4072,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634500443210429e-05,
+      "loss": 4.4176,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633663486339751e-05,
+      "loss": 4.4143,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632824891588699e-05,
+      "loss": 4.4118,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631986296837647e-05,
+      "loss": 4.407,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631147702086595e-05,
+      "loss": 4.399,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630309107335542e-05,
+      "loss": 4.4059,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62947051258449e-05,
+      "loss": 4.3965,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628631917833438e-05,
+      "loss": 4.4007,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.627793323082386e-05,
+      "loss": 4.4019,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626956366211707e-05,
+      "loss": 4.3932,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626117771460655e-05,
+      "loss": 4.397,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625279176709603e-05,
+      "loss": 4.3958,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.369274139404297,
+      "eval_runtime": 539.1561,
+      "eval_samples_per_second": 707.756,
+      "eval_steps_per_second": 22.118,
+      "step": 228957
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624440581958551e-05,
+      "loss": 4.4044,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623601987207499e-05,
+      "loss": 4.408,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6227633924564476e-05,
+      "loss": 4.3969,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219247977053956e-05,
+      "loss": 4.3936,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6210862029543436e-05,
+      "loss": 4.396,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6202492460836645e-05,
+      "loss": 4.3762,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6194122892129854e-05,
+      "loss": 4.3931,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6185736944619334e-05,
+      "loss": 4.3933,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6177350997108814e-05,
+      "loss": 4.3818,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168981428402023e-05,
+      "loss": 4.3985,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.61605954808915e-05,
+      "loss": 4.4037,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615220953338098e-05,
+      "loss": 4.384,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.614382358587046e-05,
+      "loss": 4.3837,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.613543763835994e-05,
+      "loss": 4.3852,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612705169084942e-05,
+      "loss": 4.3912,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611866574333891e-05,
+      "loss": 4.383,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611027979582839e-05,
+      "loss": 4.3809,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.610189384831787e-05,
+      "loss": 4.3807,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.609352427961108e-05,
+      "loss": 4.3827,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.608513833210056e-05,
+      "loss": 4.39,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.607675238459004e-05,
+      "loss": 4.3848,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.606838281588325e-05,
+      "loss": 4.3905,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605999686837273e-05,
+      "loss": 4.3876,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.605161092086221e-05,
+      "loss": 4.3813,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604322497335169e-05,
+      "loss": 4.3802,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.603483902584117e-05,
+      "loss": 4.3822,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602645307833065e-05,
+      "loss": 4.3863,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601806713082013e-05,
+      "loss": 4.3736,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600968118330961e-05,
+      "loss": 4.382,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600129523579909e-05,
+      "loss": 4.3671,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599290928828857e-05,
+      "loss": 4.3719,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3862,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3814,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.373,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5959381877050217e-05,
+      "loss": 4.3696,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5950995929539696e-05,
+      "loss": 4.3811,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5942609982029176e-05,
+      "loss": 4.3679,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5934240413322386e-05,
+      "loss": 4.38,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5925854465811865e-05,
+      "loss": 4.3721,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917468518301345e-05,
+      "loss": 4.36,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590908257079083e-05,
+      "loss": 4.3595,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.590071300208404e-05,
+      "loss": 4.3603,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589232705457352e-05,
+      "loss": 4.3708,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5883941107063e-05,
+      "loss": 4.3628,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587555515955248e-05,
+      "loss": 4.3663,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586716921204196e-05,
+      "loss": 4.3598,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585879964333517e-05,
+      "loss": 4.3635,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585041369582465e-05,
+      "loss": 4.3769,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584204412711786e-05,
+      "loss": 4.3575,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583365817960734e-05,
+      "loss": 4.3579,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582527223209682e-05,
+      "loss": 4.3609,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.58168862845863e-05,
+      "loss": 4.3706,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5808500337075786e-05,
+      "loss": 4.3498,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5800114389565266e-05,
+      "loss": 4.3672,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791728442054746e-05,
+      "loss": 4.3486,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783358873347955e-05,
+      "loss": 4.3556,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5774972925837435e-05,
+      "loss": 4.3648,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5766586978326915e-05,
+      "loss": 4.3487,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5758201030816395e-05,
+      "loss": 4.3602,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5749815083305875e-05,
+      "loss": 4.3539,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5741445514599084e-05,
+      "loss": 4.3357,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5733059567088564e-05,
+      "loss": 4.3513,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5724673619578044e-05,
+      "loss": 4.3543,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5716287672067524e-05,
+      "loss": 4.3504,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5707901724557004e-05,
+      "loss": 4.3462,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5699515777046484e-05,
+      "loss": 4.3478,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5691129829535963e-05,
+      "loss": 4.3301,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5682743882025443e-05,
+      "loss": 4.3566,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674357934514923e-05,
+      "loss": 4.3387,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56659719870044e-05,
+      "loss": 4.3431,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565758603949388e-05,
+      "loss": 4.3537,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564921647078709e-05,
+      "loss": 4.344,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564084690208031e-05,
+      "loss": 4.3525,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563246095456979e-05,
+      "loss": 4.3586,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562407500705927e-05,
+      "loss": 4.3525,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561568905954874e-05,
+      "loss": 4.3568,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560731949084196e-05,
+      "loss": 4.3374,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559893354333144e-05,
+      "loss": 4.334,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559054759582092e-05,
+      "loss": 4.3413,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55821616483104e-05,
+      "loss": 4.3549,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557377570079988e-05,
+      "loss": 4.3421,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556538975328936e-05,
+      "loss": 4.3342,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555700380577884e-05,
+      "loss": 4.3243,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.3356,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55402319107578e-05,
+      "loss": 4.3338,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553184596324728e-05,
+      "loss": 4.3379,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552346001573676e-05,
+      "loss": 4.3445,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551507406822624e-05,
+      "loss": 4.3473,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.550668812071572e-05,
+      "loss": 4.3375,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.54983021732052e-05,
+      "loss": 4.3346,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.548991622569468e-05,
+      "loss": 4.3243,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481530278184157e-05,
+      "loss": 4.3289,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547316070947737e-05,
+      "loss": 4.3347,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546479114077058e-05,
+      "loss": 4.3285,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.3237,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3362,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.3306,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.3422,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.3318,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.3391,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540610588700067e-05,
+      "loss": 4.3321,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539771993949015e-05,
+      "loss": 4.3245,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538933399197963e-05,
+      "loss": 4.3429,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538094804446911e-05,
+      "loss": 4.3213,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537256209695859e-05,
+      "loss": 4.3454,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536417614944807e-05,
+      "loss": 4.3151,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535580658074128e-05,
+      "loss": 4.313,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534742063323076e-05,
+      "loss": 4.3183,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533903468572024e-05,
+      "loss": 4.3327,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533064873820972e-05,
+      "loss": 4.3236,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53222627906992e-05,
+      "loss": 4.3108,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531387684318868e-05,
+      "loss": 4.3078,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530549089567816e-05,
+      "loss": 4.3161,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529712132697137e-05,
+      "loss": 4.3346,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528873537946085e-05,
+      "loss": 4.3192,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528034943195033e-05,
+      "loss": 4.3319,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527196348443981e-05,
+      "loss": 4.3216,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263577536929295e-05,
+      "loss": 4.3238,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255191589418775e-05,
+      "loss": 4.3243,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246805641908255e-05,
+      "loss": 4.314,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238419694397735e-05,
+      "loss": 4.315,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230050125690944e-05,
+      "loss": 4.3225,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.3114,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.316,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520492504076684e-05,
+      "loss": 4.3206,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519653909325632e-05,
+      "loss": 4.3182,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.3086,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.3071,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517138125072476e-05,
+      "loss": 4.3291,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516301168201798e-05,
+      "loss": 4.3093,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515462573450746e-05,
+      "loss": 4.3231,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514623978699694e-05,
+      "loss": 4.3156,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513785383948642e-05,
+      "loss": 4.322,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51294678919759e-05,
+      "loss": 4.3115,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512108194446538e-05,
+      "loss": 4.315,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511269599695486e-05,
+      "loss": 4.3152,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510431004944434e-05,
+      "loss": 4.3109,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5095940480737546e-05,
+      "loss": 4.325,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087554533227026e-05,
+      "loss": 4.3144,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079168585716506e-05,
+      "loss": 4.3181,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070799017009715e-05,
+      "loss": 4.3171,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062413069499195e-05,
+      "loss": 4.3042,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.3109,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.3017,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503725522696764e-05,
+      "loss": 4.3078,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028869279457115e-05,
+      "loss": 4.306,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020483331946595e-05,
+      "loss": 4.3056,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012097384436075e-05,
+      "loss": 4.3008,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003711436925555e-05,
+      "loss": 4.3064,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.284051418304443,
+      "eval_runtime": 539.6897,
+      "eval_samples_per_second": 707.056,
+      "eval_steps_per_second": 22.096,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995325489415035e-05,
+      "loss": 4.3165,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.3153,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.3053,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.301,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.3073,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2855,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.3041,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.3006,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2929,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.3061,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4911466014309834e-05,
+      "loss": 4.3124,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4903080066799314e-05,
+      "loss": 4.2929,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4894694119288794e-05,
+      "loss": 4.298,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886308171778274e-05,
+      "loss": 4.2963,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877922224267754e-05,
+      "loss": 4.3024,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4869536276757234e-05,
+      "loss": 4.2911,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4861150329246714e-05,
+      "loss": 4.2916,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4852764381736194e-05,
+      "loss": 4.2914,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4844378434225674e-05,
+      "loss": 4.2923,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483600886551888e-05,
+      "loss": 4.3041,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482762291800836e-05,
+      "loss": 4.296,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481925334930158e-05,
+      "loss": 4.3043,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481086740179106e-05,
+      "loss": 4.3017,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480248145428054e-05,
+      "loss": 4.2943,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479409550677002e-05,
+      "loss": 4.2928,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2946,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.3003,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.2877,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476055171672794e-05,
+      "loss": 4.2922,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2885,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2808,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.298,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.3009,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4718654736782804e-05,
+      "loss": 4.2895,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471028516807601e-05,
+      "loss": 4.2844,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.470189922056549e-05,
+      "loss": 4.2914,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.469351327305497e-05,
+      "loss": 4.2903,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.468512732554445e-05,
+      "loss": 4.292,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.467674137803393e-05,
+      "loss": 4.2871,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.466835543052341e-05,
+      "loss": 4.2774,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465998586181662e-05,
+      "loss": 4.2757,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465161629310983e-05,
+      "loss": 4.2759,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.464323034559931e-05,
+      "loss": 4.288,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.463484439808879e-05,
+      "loss": 4.2819,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.462645845057827e-05,
+      "loss": 4.2831,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.461807250306775e-05,
+      "loss": 4.2794,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460968655555724e-05,
+      "loss": 4.2785,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460130060804672e-05,
+      "loss": 4.2961,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459291466053619e-05,
+      "loss": 4.2766,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4584545091829406e-05,
+      "loss": 4.2744,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4576159144318886e-05,
+      "loss": 4.2821,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4567773196808366e-05,
+      "loss": 4.284,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455938724929784e-05,
+      "loss": 4.2707,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455100130178732e-05,
+      "loss": 4.2814,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45426153542768e-05,
+      "loss": 4.2703,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2762,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452585983805949e-05,
+      "loss": 4.2828,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4517473890548975e-05,
+      "loss": 4.2665,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509087943038455e-05,
+      "loss": 4.2827,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500718374331664e-05,
+      "loss": 4.2699,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492332426821144e-05,
+      "loss": 4.2612,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483946479310624e-05,
+      "loss": 4.2692,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475560531800104e-05,
+      "loss": 4.2739,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467174584289584e-05,
+      "loss": 4.2766,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445880501558279e-05,
+      "loss": 4.2654,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.445041906807227e-05,
+      "loss": 4.2711,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.444203312056175e-05,
+      "loss": 4.2467,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.443366355185496e-05,
+      "loss": 4.2791,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.442527760434444e-05,
+      "loss": 4.2629,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.441689165683393e-05,
+      "loss": 4.2642,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440850570932341e-05,
+      "loss": 4.2751,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.440011976181289e-05,
+      "loss": 4.2663,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.439173381430237e-05,
+      "loss": 4.2714,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438334786679185e-05,
+      "loss": 4.278,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.437496191928133e-05,
+      "loss": 4.2755,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436657597177081e-05,
+      "loss": 4.2768,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435819002426029e-05,
+      "loss": 4.2639,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434980407674977e-05,
+      "loss": 4.2552,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434141812923925e-05,
+      "loss": 4.2662,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433304856053246e-05,
+      "loss": 4.2752,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432466261302194e-05,
+      "loss": 4.266,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431627666551142e-05,
+      "loss": 4.2616,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43078907180009e-05,
+      "loss": 4.2467,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429950477049038e-05,
+      "loss": 4.2606,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429111882297986e-05,
+      "loss": 4.2549,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428273287546934e-05,
+      "loss": 4.2649,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427436330676255e-05,
+      "loss": 4.2638,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4265977359252026e-05,
+      "loss": 4.2736,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257591411741506e-05,
+      "loss": 4.2582,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4249205464230986e-05,
+      "loss": 4.2609,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4240819516720466e-05,
+      "loss": 4.2486,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232433569209946e-05,
+      "loss": 4.2555,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224047621699426e-05,
+      "loss": 4.261,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215661674188906e-05,
+      "loss": 4.2506,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4207275726678386e-05,
+      "loss": 4.2508,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4198889779167866e-05,
+      "loss": 4.2634,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419052021046108e-05,
+      "loss": 4.2535,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418213426295056e-05,
+      "loss": 4.2748,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417374831544004e-05,
+      "loss": 4.2549,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416536236792952e-05,
+      "loss": 4.2638,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4156976420419e-05,
+      "loss": 4.2558,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414862323051594e-05,
+      "loss": 4.2531,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414023728300542e-05,
+      "loss": 4.2686,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.41318513354949e-05,
+      "loss": 4.249,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412346538798438e-05,
+      "loss": 4.27,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411507944047386e-05,
+      "loss": 4.2457,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410669349296334e-05,
+      "loss": 4.2424,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409830754545282e-05,
+      "loss": 4.2439,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.40899215979423e-05,
+      "loss": 4.2589,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408153565043178e-05,
+      "loss": 4.2551,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407314970292126e-05,
+      "loss": 4.2416,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064780134214475e-05,
+      "loss": 4.2341,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056394186703955e-05,
+      "loss": 4.2457,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048008239193435e-05,
+      "loss": 4.2628,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039622291682915e-05,
+      "loss": 4.2456,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.403123634417239e-05,
+      "loss": 4.2616,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402285039666187e-05,
+      "loss": 4.2512,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4014480827955084e-05,
+      "loss": 4.2546,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400609488044456e-05,
+      "loss": 4.2545,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399770893293404e-05,
+      "loss": 4.2427,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398932298542352e-05,
+      "loss": 4.2434,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398095341671673e-05,
+      "loss": 4.2525,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397256746920621e-05,
+      "loss": 4.2419,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396418152169569e-05,
+      "loss": 4.2468,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395579557418517e-05,
+      "loss": 4.2483,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394740962667465e-05,
+      "loss": 4.2489,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393902367916413e-05,
+      "loss": 4.2392,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393065411045734e-05,
+      "loss": 4.2393,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392226816294682e-05,
+      "loss": 4.2549,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39138822154363e-05,
+      "loss": 4.2405,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390549626792578e-05,
+      "loss": 4.2545,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389712669921899e-05,
+      "loss": 4.2501,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388874075170847e-05,
+      "loss": 4.2496,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388035480419795e-05,
+      "loss": 4.2459,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387196885668744e-05,
+      "loss": 4.2438,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863599287980647e-05,
+      "loss": 4.2475,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855213340470127e-05,
+      "loss": 4.2419,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846827392959606e-05,
+      "loss": 4.256,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838441445449086e-05,
+      "loss": 4.2502,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830055497938566e-05,
+      "loss": 4.2509,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821669550428046e-05,
+      "loss": 4.2494,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813299981721255e-05,
+      "loss": 4.2419,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804914034210735e-05,
+      "loss": 4.2427,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796528086700215e-05,
+      "loss": 4.236,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788142139189695e-05,
+      "loss": 4.2389,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779772570482904e-05,
+      "loss": 4.242,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377138662297239e-05,
+      "loss": 4.2393,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376300067546187e-05,
+      "loss": 4.2289,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375461472795135e-05,
+      "loss": 4.2412,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.225332260131836,
+      "eval_runtime": 541.0223,
+      "eval_samples_per_second": 705.315,
+      "eval_steps_per_second": 22.042,
+      "step": 381595
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374622878044083e-05,
+      "loss": 4.2473,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373785921173404e-05,
+      "loss": 4.2438,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372947326422352e-05,
+      "loss": 4.2419,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3721087316713e-05,
+      "loss": 4.238,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371270136920248e-05,
+      "loss": 4.2352,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370433180049569e-05,
+      "loss": 4.2252,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36959622317889e-05,
+      "loss": 4.2364,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3687592663082114e-05,
+      "loss": 4.2306,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3679206715571594e-05,
+      "loss": 4.2351,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670820768061074e-05,
+      "loss": 4.2413,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662434820550554e-05,
+      "loss": 4.2426,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654048873040034e-05,
+      "loss": 4.2271,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645662925529514e-05,
+      "loss": 4.2349,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637276978018994e-05,
+      "loss": 4.2302,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3628891030508474e-05,
+      "loss": 4.2369,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3620505082997954e-05,
+      "loss": 4.2272,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3612119135487434e-05,
+      "loss": 4.2253,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3603733187976914e-05,
+      "loss": 4.2272,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359536361927012e-05,
+      "loss": 4.2314,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35869776717596e-05,
+      "loss": 4.2392,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357859172424908e-05,
+      "loss": 4.2317,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357020577673856e-05,
+      "loss": 4.2421,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356181982922804e-05,
+      "loss": 4.2357,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355343388171752e-05,
+      "loss": 4.2333,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3545047934207e-05,
+      "loss": 4.2279,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353666198669648e-05,
+      "loss": 4.2286,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352827603918596e-05,
+      "loss": 4.2389,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351989009167544e-05,
+      "loss": 4.2229,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351152052296865e-05,
+      "loss": 4.231,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350313457545813e-05,
+      "loss": 4.2256,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349474862794761e-05,
+      "loss": 4.2156,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348636268043709e-05,
+      "loss": 4.2383,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.347797673292657e-05,
+      "loss": 4.2327,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346959078541605e-05,
+      "loss": 4.2314,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.346120483790553e-05,
+      "loss": 4.2187,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.345283526919875e-05,
+      "loss": 4.2334,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.344444932168823e-05,
+      "loss": 4.2239,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.343606337417771e-05,
+      "loss": 4.2337,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.342767742666719e-05,
+      "loss": 4.2276,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.341929147915667e-05,
+      "loss": 4.2149,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3410921910449876e-05,
+      "loss": 4.215,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402535962939356e-05,
+      "loss": 4.2117,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394150015428836e-05,
+      "loss": 4.2317,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385764067918316e-05,
+      "loss": 4.2212,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377378120407796e-05,
+      "loss": 4.2167,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3368992172897276e-05,
+      "loss": 4.222,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360606225386756e-05,
+      "loss": 4.218,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3352236656679965e-05,
+      "loss": 4.2329,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3343850709169445e-05,
+      "loss": 4.2137,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333546476165893e-05,
+      "loss": 4.2181,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332707881414841e-05,
+      "loss": 4.222,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3318692866637885e-05,
+      "loss": 4.2231,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3310306919127365e-05,
+      "loss": 4.2095,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330193735042058e-05,
+      "loss": 4.2196,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3293551402910054e-05,
+      "loss": 4.2107,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.328518183420327e-05,
+      "loss": 4.2181,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327679588669275e-05,
+      "loss": 4.2187,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326840993918223e-05,
+      "loss": 4.2143,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32600239916717e-05,
+      "loss": 4.2235,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325163804416118e-05,
+      "loss": 4.2085,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32432684754544e-05,
+      "loss": 4.2055,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3234882527943885e-05,
+      "loss": 4.2087,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.322649658043336e-05,
+      "loss": 4.2089,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.321811063292284e-05,
+      "loss": 4.2254,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320972468541232e-05,
+      "loss": 4.2047,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320135511670553e-05,
+      "loss": 4.214,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319296916919501e-05,
+      "loss": 4.1926,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318458322168449e-05,
+      "loss": 4.2149,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.317619727417397e-05,
+      "loss": 4.2047,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.316781132666345e-05,
+      "loss": 4.205,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315942537915293e-05,
+      "loss": 4.2168,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.315103943164241e-05,
+      "loss": 4.2103,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314265348413189e-05,
+      "loss": 4.2137,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313426753662137e-05,
+      "loss": 4.2202,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3125881589110854e-05,
+      "loss": 4.2164,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3117495641600334e-05,
+      "loss": 4.2234,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310912607289354e-05,
+      "loss": 4.2073,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310074012538302e-05,
+      "loss": 4.198,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30923541778725e-05,
+      "loss": 4.2073,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308396823036198e-05,
+      "loss": 4.2189,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307559866165519e-05,
+      "loss": 4.2061,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306721271414467e-05,
+      "loss": 4.2059,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305884314543788e-05,
+      "loss": 4.1895,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.305045719792736e-05,
+      "loss": 4.2048,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.304207125041684e-05,
+      "loss": 4.1966,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.303368530290632e-05,
+      "loss": 4.2087,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30252993553958e-05,
+      "loss": 4.21,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.301691340788529e-05,
+      "loss": 4.2153,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.300852746037477e-05,
+      "loss": 4.2044,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.300014151286425e-05,
+      "loss": 4.2055,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.299175556535372e-05,
+      "loss": 4.1921,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.29833696178432e-05,
+      "loss": 4.1987,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.297498367033268e-05,
+      "loss": 4.2033,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.296659772282216e-05,
+      "loss": 4.1988,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.295822815411537e-05,
+      "loss": 4.1955,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949858585408585e-05,
+      "loss": 4.2097,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2941472637898065e-05,
+      "loss": 4.1972,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293308669038754e-05,
+      "loss": 4.2194,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2924700742877025e-05,
+      "loss": 4.1945,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2916314795366505e-05,
+      "loss": 4.2136,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2907928847855985e-05,
+      "loss": 4.2011,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2899559279149194e-05,
+      "loss": 4.1986,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2891173331638674e-05,
+      "loss": 4.2156,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2882787384128154e-05,
+      "loss": 4.1928,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2874401436617634e-05,
+      "loss": 4.2153,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2866015489107114e-05,
+      "loss": 4.1897,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2857629541596594e-05,
+      "loss": 4.189,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2849243594086074e-05,
+      "loss": 4.1895,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284087402537928e-05,
+      "loss": 4.2023,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283248807786876e-05,
+      "loss": 4.2012,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282410213035824e-05,
+      "loss": 4.1923,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.281571618284772e-05,
+      "loss": 4.1776,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.280733023533721e-05,
+      "loss": 4.1917,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279894428782669e-05,
+      "loss": 4.206,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.279055834031617e-05,
+      "loss": 4.1983,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278217239280565e-05,
+      "loss": 4.2032,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277378644529513e-05,
+      "loss": 4.1993,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27654004977846e-05,
+      "loss": 4.1994,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275701455027408e-05,
+      "loss": 4.2033,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274862860276356e-05,
+      "loss": 4.1937,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274025903405678e-05,
+      "loss": 4.1879,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273187308654625e-05,
+      "loss": 4.1967,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272348713903573e-05,
+      "loss": 4.1904,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271510119152521e-05,
+      "loss": 4.1934,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270673162281843e-05,
+      "loss": 4.1957,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269836205411164e-05,
+      "loss": 4.1948,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268997610660112e-05,
+      "loss": 4.192,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268160653789433e-05,
+      "loss": 4.1856,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267322059038381e-05,
+      "loss": 4.2017,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266483464287329e-05,
+      "loss": 4.1879,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.265644869536277e-05,
+      "loss": 4.2,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264806274785225e-05,
+      "loss": 4.2024,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2639676800341725e-05,
+      "loss": 4.1977,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2631290852831205e-05,
+      "loss": 4.1964,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2622904905320685e-05,
+      "loss": 4.1918,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26145353366139e-05,
+      "loss": 4.1935,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.260614938910338e-05,
+      "loss": 4.1868,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.259776344159286e-05,
+      "loss": 4.2055,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258937749408234e-05,
+      "loss": 4.2018,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.258099154657182e-05,
+      "loss": 4.197,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.25726055990613e-05,
+      "loss": 4.1943,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.256421965155078e-05,
+      "loss": 4.1918,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.255585008284399e-05,
+      "loss": 4.1938,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.254746413533347e-05,
+      "loss": 4.1857,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253907818782295e-05,
+      "loss": 4.1815,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.253069224031243e-05,
+      "loss": 4.1947,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.252230629280191e-05,
+      "loss": 4.1879,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.251392034529139e-05,
+      "loss": 4.1773,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250553439778087e-05,
+      "loss": 4.1888,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.182226657867432,
+      "eval_runtime": 540.6268,
+      "eval_samples_per_second": 705.831,
+      "eval_steps_per_second": 22.058,
+      "step": 457914
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249714845027035e-05,
+      "loss": 4.203,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248876250275983e-05,
+      "loss": 4.1941,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.248037655524931e-05,
+      "loss": 4.1904,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2471990607738796e-05,
+      "loss": 4.187,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.246360466022827e-05,
+      "loss": 4.1876,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.245521871271775e-05,
+      "loss": 4.1778,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.244683276520723e-05,
+      "loss": 4.1862,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243844681769671e-05,
+      "loss": 4.1799,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243006087018619e-05,
+      "loss": 4.1846,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242167492267567e-05,
+      "loss": 4.1901,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241330535396888e-05,
+      "loss": 4.1953,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240491940645836e-05,
+      "loss": 4.176,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239653345894784e-05,
+      "loss": 4.1889,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238814751143732e-05,
+      "loss": 4.1787,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.23797615639268e-05,
+      "loss": 4.1907,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237137561641628e-05,
+      "loss": 4.1757,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2362989668905765e-05,
+      "loss": 4.1753,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2354603721395245e-05,
+      "loss": 4.1828,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2346217773884725e-05,
+      "loss": 4.1796,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2337848205177934e-05,
+      "loss": 4.1891,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2329462257667414e-05,
+      "loss": 4.1841,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2321076310156894e-05,
+      "loss": 4.1926,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2312690362646374e-05,
+      "loss": 4.1884,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230432079393958e-05,
+      "loss": 4.1835,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229593484642906e-05,
+      "loss": 4.1795,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228754889891854e-05,
+      "loss": 4.1802,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227916295140802e-05,
+      "loss": 4.1911,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22707770038975e-05,
+      "loss": 4.1794,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226239105638698e-05,
+      "loss": 4.1812,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.225400510887646e-05,
+      "loss": 4.1808,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.224561916136594e-05,
+      "loss": 4.1701,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.223723321385542e-05,
+      "loss": 4.187,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.22288472663449e-05,
+      "loss": 4.1862,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.222047769763811e-05,
+      "loss": 4.1839,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.221209175012759e-05,
+      "loss": 4.1703,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.220370580261707e-05,
+      "loss": 4.184,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.219531985510655e-05,
+      "loss": 4.1774,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.218693390759603e-05,
+      "loss": 4.1849,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217854796008551e-05,
+      "loss": 4.1812,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.217016201257499e-05,
+      "loss": 4.1695,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216177606506447e-05,
+      "loss": 4.1634,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215340649635769e-05,
+      "loss": 4.1636,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214502054884717e-05,
+      "loss": 4.1873,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213663460133665e-05,
+      "loss": 4.1723,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212824865382613e-05,
+      "loss": 4.1681,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211986270631561e-05,
+      "loss": 4.1791,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2111493137608816e-05,
+      "loss": 4.1706,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2103107190098296e-05,
+      "loss": 4.1828,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2094721242587776e-05,
+      "loss": 4.1692,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2086335295077256e-05,
+      "loss": 4.1749,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2077982105174194e-05,
+      "loss": 4.1726,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2069596157663674e-05,
+      "loss": 4.1776,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061210210153154e-05,
+      "loss": 4.167,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052824262642634e-05,
+      "loss": 4.1679,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.204443831513212e-05,
+      "loss": 4.1684,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.20360523676216e-05,
+      "loss": 4.1743,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.202768279891481e-05,
+      "loss": 4.1695,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201929685140429e-05,
+      "loss": 4.1677,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.201091090389377e-05,
+      "loss": 4.1785,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.200252495638325e-05,
+      "loss": 4.1627,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.199413900887273e-05,
+      "loss": 4.1589,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.198575306136221e-05,
+      "loss": 4.1619,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197736711385169e-05,
+      "loss": 4.1636,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196898116634116e-05,
+      "loss": 4.1807,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.196059521883064e-05,
+      "loss": 4.1595,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195222565012386e-05,
+      "loss": 4.1669,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194383970261334e-05,
+      "loss": 4.1484,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193545375510282e-05,
+      "loss": 4.1652,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19270678075923e-05,
+      "loss": 4.1599,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191868186008178e-05,
+      "loss": 4.1623,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191032867017872e-05,
+      "loss": 4.1693,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19019427226682e-05,
+      "loss": 4.1642,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189355677515768e-05,
+      "loss": 4.17,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188517082764716e-05,
+      "loss": 4.1737,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1876784880136636e-05,
+      "loss": 4.1709,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186841531142985e-05,
+      "loss": 4.1756,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186002936391933e-05,
+      "loss": 4.1701,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185164341640881e-05,
+      "loss": 4.1514,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184325746889829e-05,
+      "loss": 4.1619,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183488790019151e-05,
+      "loss": 4.1694,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182651833148472e-05,
+      "loss": 4.1658,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18181323839742e-05,
+      "loss": 4.1643,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180974643646368e-05,
+      "loss": 4.1436,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180136048895316e-05,
+      "loss": 4.161,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.179297454144264e-05,
+      "loss": 4.1526,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.178458859393211e-05,
+      "loss": 4.168,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.177620264642159e-05,
+      "loss": 4.1642,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.176781669891107e-05,
+      "loss": 4.1711,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.175943075140055e-05,
+      "loss": 4.1585,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.175104480389003e-05,
+      "loss": 4.1613,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174265885637951e-05,
+      "loss": 4.1507,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1734272908868996e-05,
+      "loss": 4.1614,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1725886961358476e-05,
+      "loss": 4.1572,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1717501013847956e-05,
+      "loss": 4.1555,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1709115066337436e-05,
+      "loss": 4.1497,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1700729118826916e-05,
+      "loss": 4.1648,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1692375928923855e-05,
+      "loss": 4.155,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1683989981413334e-05,
+      "loss": 4.1791,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1675604033902814e-05,
+      "loss": 4.153,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1667218086392294e-05,
+      "loss": 4.1691,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1658832138881774e-05,
+      "loss": 4.1593,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1650462570174983e-05,
+      "loss": 4.156,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.164207662266446e-05,
+      "loss": 4.1696,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163369067515395e-05,
+      "loss": 4.1567,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.162530472764343e-05,
+      "loss": 4.1682,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.161691878013291e-05,
+      "loss": 4.1532,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160853283262239e-05,
+      "loss": 4.1434,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160014688511187e-05,
+      "loss": 4.1464,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159176093760135e-05,
+      "loss": 4.1601,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158337499009082e-05,
+      "loss": 4.1589,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.157500542138404e-05,
+      "loss": 4.1542,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156661947387352e-05,
+      "loss": 4.1324,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558233526363e-05,
+      "loss": 4.1518,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154984757885247e-05,
+      "loss": 4.1657,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154146163134195e-05,
+      "loss": 4.1566,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153309206263517e-05,
+      "loss": 4.1637,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152470611512465e-05,
+      "loss": 4.157,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151632016761413e-05,
+      "loss": 4.1576,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150793422010361e-05,
+      "loss": 4.1597,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149954827259309e-05,
+      "loss": 4.1548,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.14911787038863e-05,
+      "loss": 4.1455,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148279275637578e-05,
+      "loss": 4.1552,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147440680886526e-05,
+      "loss": 4.1487,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146602086135474e-05,
+      "loss": 4.153,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1457651292647946e-05,
+      "loss": 4.1551,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1449265345137426e-05,
+      "loss": 4.1539,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1440879397626906e-05,
+      "loss": 4.1488,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432493450116386e-05,
+      "loss": 4.1463,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424107502605866e-05,
+      "loss": 4.1588,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.141573793389908e-05,
+      "loss": 4.1465,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.140735198638856e-05,
+      "loss": 4.1608,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139896603887804e-05,
+      "loss": 4.1619,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.139058009136752e-05,
+      "loss": 4.1552,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.138221052266073e-05,
+      "loss": 4.1578,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.137382457515021e-05,
+      "loss": 4.1535,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.136543862763969e-05,
+      "loss": 4.1515,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.135705268012917e-05,
+      "loss": 4.1464,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134866673261865e-05,
+      "loss": 4.1664,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134028078510813e-05,
+      "loss": 4.1626,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133189483759761e-05,
+      "loss": 4.1584,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132352526889082e-05,
+      "loss": 4.1562,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1315155700184035e-05,
+      "loss": 4.1484,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1306769752673515e-05,
+      "loss": 4.1522,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1298383805162995e-05,
+      "loss": 4.1484,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1289997857652475e-05,
+      "loss": 4.1423,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128164466774942e-05,
+      "loss": 4.1584,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127325872023889e-05,
+      "loss": 4.1487,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126487277272837e-05,
+      "loss": 4.1389,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125648682521785e-05,
+      "loss": 4.1518,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.150303840637207,
+      "eval_runtime": 543.7219,
+      "eval_samples_per_second": 701.813,
+      "eval_steps_per_second": 21.932,
+      "step": 534233
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.124810087770733e-05,
+      "loss": 4.1551,
+      "step": 534528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123971493019681e-05,
+      "loss": 4.152,
+      "step": 535040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.123132898268629e-05,
+      "loss": 4.1567,
+      "step": 535552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.122294303517577e-05,
+      "loss": 4.1484,
+      "step": 536064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.121455708766526e-05,
+      "loss": 4.1449,
+      "step": 536576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.120617114015474e-05,
+      "loss": 4.1427,
+      "step": 537088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.119780157144795e-05,
+      "loss": 4.1452,
+      "step": 537600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118941562393743e-05,
+      "loss": 4.1435,
+      "step": 538112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.118102967642691e-05,
+      "loss": 4.1424,
+      "step": 538624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.117264372891639e-05,
+      "loss": 4.1466,
+      "step": 539136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.11642741602096e-05,
+      "loss": 4.1586,
+      "step": 539648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.115588821269908e-05,
+      "loss": 4.1401,
+      "step": 540160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.114750226518856e-05,
+      "loss": 4.1508,
+      "step": 540672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113911631767804e-05,
+      "loss": 4.1384,
+      "step": 541184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.113073037016752e-05,
+      "loss": 4.1542,
+      "step": 541696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1122344422657e-05,
+      "loss": 4.1356,
+      "step": 542208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.111395847514648e-05,
+      "loss": 4.1396,
+      "step": 542720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.110557252763596e-05,
+      "loss": 4.1423,
+      "step": 543232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.109718658012544e-05,
+      "loss": 4.1413,
+      "step": 543744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.108880063261492e-05,
+      "loss": 4.1514,
+      "step": 544256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.10804146851044e-05,
+      "loss": 4.1444,
+      "step": 544768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.107202873759388e-05,
+      "loss": 4.1573,
+      "step": 545280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.106364279008336e-05,
+      "loss": 4.145,
+      "step": 545792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.105525684257284e-05,
+      "loss": 4.1475,
+      "step": 546304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.104687089506232e-05,
+      "loss": 4.1437,
+      "step": 546816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1038501326355526e-05,
+      "loss": 4.1433,
+      "step": 547328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1030115378845006e-05,
+      "loss": 4.1468,
+      "step": 547840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1021745810138215e-05,
+      "loss": 4.1413,
+      "step": 548352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.1013359862627695e-05,
+      "loss": 4.1467,
+      "step": 548864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.100497391511718e-05,
+      "loss": 4.143,
+      "step": 549376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.099658796760666e-05,
+      "loss": 4.1332,
+      "step": 549888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.098820202009614e-05,
+      "loss": 4.146,
+      "step": 550400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.097981607258562e-05,
+      "loss": 4.1514,
+      "step": 550912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09714301250751e-05,
+      "loss": 4.1439,
+      "step": 551424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.096304417756458e-05,
+      "loss": 4.1357,
+      "step": 551936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.095467460885779e-05,
+      "loss": 4.1477,
+      "step": 552448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.094628866134727e-05,
+      "loss": 4.1422,
+      "step": 552960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.093790271383675e-05,
+      "loss": 4.1461,
+      "step": 553472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092951676632623e-05,
+      "loss": 4.1423,
+      "step": 553984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.092114719761944e-05,
+      "loss": 4.1355,
+      "step": 554496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.091276125010892e-05,
+      "loss": 4.1267,
+      "step": 555008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.09043753025984e-05,
+      "loss": 4.1282,
+      "step": 555520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.089598935508788e-05,
+      "loss": 4.1464,
+      "step": 556032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.088760340757736e-05,
+      "loss": 4.1367,
+      "step": 556544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0879233838870575e-05,
+      "loss": 4.1304,
+      "step": 557056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0870847891360055e-05,
+      "loss": 4.1436,
+      "step": 557568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0862461943849535e-05,
+      "loss": 4.1331,
+      "step": 558080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0854075996339015e-05,
+      "loss": 4.1478,
+      "step": 558592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0845690048828495e-05,
+      "loss": 4.1336,
+      "step": 559104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.083730410131797e-05,
+      "loss": 4.1402,
+      "step": 559616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0828934532611184e-05,
+      "loss": 4.133,
+      "step": 560128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0820548585100664e-05,
+      "loss": 4.1396,
+      "step": 560640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.081216263759014e-05,
+      "loss": 4.1322,
+      "step": 561152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.080379306888335e-05,
+      "loss": 4.1292,
+      "step": 561664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.079540712137283e-05,
+      "loss": 4.1369,
+      "step": 562176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.078702117386231e-05,
+      "loss": 4.1349,
+      "step": 562688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077863522635179e-05,
+      "loss": 4.1334,
+      "step": 563200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.077026565764501e-05,
+      "loss": 4.1321,
+      "step": 563712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.076187971013449e-05,
+      "loss": 4.144,
+      "step": 564224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.075349376262397e-05,
+      "loss": 4.1226,
+      "step": 564736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.074510781511344e-05,
+      "loss": 4.127,
+      "step": 565248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.073672186760292e-05,
+      "loss": 4.1259,
+      "step": 565760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.07283359200924e-05,
+      "loss": 4.1316,
+      "step": 566272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071994997258188e-05,
+      "loss": 4.1422,
+      "step": 566784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.071156402507136e-05,
+      "loss": 4.1244,
+      "step": 567296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.070317807756084e-05,
+      "loss": 4.1316,
+      "step": 567808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.069479213005032e-05,
+      "loss": 4.1155,
+      "step": 568320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06864061825398e-05,
+      "loss": 4.1318,
+      "step": 568832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.067802023502928e-05,
+      "loss": 4.1219,
+      "step": 569344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.06696506663225e-05,
+      "loss": 4.129,
+      "step": 569856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.066126471881198e-05,
+      "loss": 4.1332,
+      "step": 570368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.065287877130146e-05,
+      "loss": 4.1286,
+      "step": 570880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.064450920259467e-05,
+      "loss": 4.1333,
+      "step": 571392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0636123255084147e-05,
+      "loss": 4.1424,
+      "step": 571904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0627737307573627e-05,
+      "loss": 4.1339,
+      "step": 572416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0619351360063106e-05,
+      "loss": 4.1385,
+      "step": 572928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0610965412552586e-05,
+      "loss": 4.1332,
+      "step": 573440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0602579465042066e-05,
+      "loss": 4.1194,
+      "step": 573952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0594193517531546e-05,
+      "loss": 4.1276,
+      "step": 574464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0585823948824755e-05,
+      "loss": 4.1335,
+      "step": 574976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0577438001314235e-05,
+      "loss": 4.1337,
+      "step": 575488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.056905205380372e-05,
+      "loss": 4.127,
+      "step": 576000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.05606661062932e-05,
+      "loss": 4.1132,
+      "step": 576512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.055228015878268e-05,
+      "loss": 4.1223,
+      "step": 577024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0543894211272155e-05,
+      "loss": 4.1194,
+      "step": 577536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0535508263761635e-05,
+      "loss": 4.1321,
+      "step": 578048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0527122316251115e-05,
+      "loss": 4.1275,
+      "step": 578560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.0518752747544324e-05,
+      "loss": 4.1355,
+      "step": 579072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.051038317883754e-05,
+      "loss": 4.1293,
+      "step": 579584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.050199723132702e-05,
+      "loss": 4.1242,
+      "step": 580096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04936112838165e-05,
+      "loss": 4.1196,
+      "step": 580608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.048522533630597e-05,
+      "loss": 4.1242,
+      "step": 581120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.047683938879546e-05,
+      "loss": 4.1193,
+      "step": 581632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046845344128494e-05,
+      "loss": 4.1221,
+      "step": 582144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.046006749377442e-05,
+      "loss": 4.117,
+      "step": 582656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.04516815462639e-05,
+      "loss": 4.1291,
+      "step": 583168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.044331197755711e-05,
+      "loss": 4.1224,
+      "step": 583680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.043492603004659e-05,
+      "loss": 4.1458,
+      "step": 584192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.042654008253607e-05,
+      "loss": 4.1159,
+      "step": 584704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.041815413502555e-05,
+      "loss": 4.1348,
+      "step": 585216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040978456631876e-05,
+      "loss": 4.1285,
+      "step": 585728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.040139861880824e-05,
+      "loss": 4.1222,
+      "step": 586240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.039301267129772e-05,
+      "loss": 4.1279,
+      "step": 586752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03846267237872e-05,
+      "loss": 4.1253,
+      "step": 587264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0376257155080414e-05,
+      "loss": 4.1335,
+      "step": 587776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0367871207569894e-05,
+      "loss": 4.1208,
+      "step": 588288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0359485260059374e-05,
+      "loss": 4.1109,
+      "step": 588800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0351099312548853e-05,
+      "loss": 4.1144,
+      "step": 589312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0342713365038333e-05,
+      "loss": 4.1257,
+      "step": 589824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.033434379633154e-05,
+      "loss": 4.1264,
+      "step": 590336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.032595784882102e-05,
+      "loss": 4.1174,
+      "step": 590848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.03175719013105e-05,
+      "loss": 4.0983,
+      "step": 591360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030918595379998e-05,
+      "loss": 4.1186,
+      "step": 591872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.030080000628946e-05,
+      "loss": 4.1318,
+      "step": 592384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.029241405877894e-05,
+      "loss": 4.122,
+      "step": 592896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.028402811126842e-05,
+      "loss": 4.1326,
+      "step": 593408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.027565854256163e-05,
+      "loss": 4.124,
+      "step": 593920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.026727259505111e-05,
+      "loss": 4.1193,
+      "step": 594432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025888664754059e-05,
+      "loss": 4.1279,
+      "step": 594944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.025050070003008e-05,
+      "loss": 4.1216,
+      "step": 595456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.024213113132329e-05,
+      "loss": 4.1147,
+      "step": 595968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.023374518381277e-05,
+      "loss": 4.1214,
+      "step": 596480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.022535923630225e-05,
+      "loss": 4.1169,
+      "step": 596992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.021697328879173e-05,
+      "loss": 4.1173,
+      "step": 597504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.020858734128121e-05,
+      "loss": 4.1177,
+      "step": 598016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0200217772574416e-05,
+      "loss": 4.126,
+      "step": 598528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0191831825063896e-05,
+      "loss": 4.1168,
+      "step": 599040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0183445877553376e-05,
+      "loss": 4.1096,
+      "step": 599552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0175059930042856e-05,
+      "loss": 4.1294,
+      "step": 600064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0166673982532336e-05,
+      "loss": 4.1112,
+      "step": 600576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0158288035021816e-05,
+      "loss": 4.1256,
+      "step": 601088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0149902087511296e-05,
+      "loss": 4.1314,
+      "step": 601600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.014153251880451e-05,
+      "loss": 4.1223,
+      "step": 602112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.013314657129399e-05,
+      "loss": 4.1218,
+      "step": 602624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0124760623783465e-05,
+      "loss": 4.1173,
+      "step": 603136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0116374676272945e-05,
+      "loss": 4.1193,
+      "step": 603648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0107988728762425e-05,
+      "loss": 4.1165,
+      "step": 604160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0099619160055634e-05,
+      "loss": 4.1319,
+      "step": 604672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0091233212545114e-05,
+      "loss": 4.1284,
+      "step": 605184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0082847265034594e-05,
+      "loss": 4.1277,
+      "step": 605696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.0074461317524074e-05,
+      "loss": 4.1206,
+      "step": 606208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.006609174881728e-05,
+      "loss": 4.116,
+      "step": 606720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.005770580130677e-05,
+      "loss": 4.1219,
+      "step": 607232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004931985379625e-05,
+      "loss": 4.1125,
+      "step": 607744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.004093390628573e-05,
+      "loss": 4.1122,
+      "step": 608256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.003254795877521e-05,
+      "loss": 4.1236,
+      "step": 608768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.002416201126469e-05,
+      "loss": 4.1157,
+      "step": 609280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.001577606375417e-05,
+      "loss": 4.1036,
+      "step": 609792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.000740649504738e-05,
+      "loss": 4.1231,
+      "step": 610304
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.125118732452393,
+      "eval_runtime": 541.3452,
+      "eval_samples_per_second": 704.894,
+      "eval_steps_per_second": 22.028,
+      "step": 610552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999903692634059e-05,
+      "loss": 4.129,
+      "step": 610816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.999065097883007e-05,
+      "loss": 4.1251,
+      "step": 611328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.998226503131955e-05,
+      "loss": 4.1223,
+      "step": 611840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.997387908380903e-05,
+      "loss": 4.1142,
+      "step": 612352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.996549313629851e-05,
+      "loss": 4.1123,
+      "step": 612864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.995710718878799e-05,
+      "loss": 4.1112,
+      "step": 613376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.994872124127747e-05,
+      "loss": 4.1145,
+      "step": 613888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9940335293766954e-05,
+      "loss": 4.1084,
+      "step": 614400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9931949346256434e-05,
+      "loss": 4.1102,
+      "step": 614912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9923563398745914e-05,
+      "loss": 4.1177,
+      "step": 615424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9915177451235394e-05,
+      "loss": 4.1257,
+      "step": 615936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9906791503724874e-05,
+      "loss": 4.1087,
+      "step": 616448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989840555621435e-05,
+      "loss": 4.117,
+      "step": 616960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.989001960870383e-05,
+      "loss": 4.1041,
+      "step": 617472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.988163366119331e-05,
+      "loss": 4.1255,
+      "step": 617984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.987324771368279e-05,
+      "loss": 4.1018,
+      "step": 618496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.986486176617227e-05,
+      "loss": 4.1087,
+      "step": 619008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.985650857626921e-05,
+      "loss": 4.1124,
+      "step": 619520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.984812262875869e-05,
+      "loss": 4.1128,
+      "step": 620032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983973668124817e-05,
+      "loss": 4.1183,
+      "step": 620544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.983135073373765e-05,
+      "loss": 4.1124,
+      "step": 621056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.982296478622713e-05,
+      "loss": 4.127,
+      "step": 621568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.981457883871661e-05,
+      "loss": 4.115,
+      "step": 622080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.980619289120609e-05,
+      "loss": 4.1155,
+      "step": 622592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.979780694369557e-05,
+      "loss": 4.1085,
+      "step": 623104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978942099618505e-05,
+      "loss": 4.1171,
+      "step": 623616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.978105142747826e-05,
+      "loss": 4.1148,
+      "step": 624128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.977266547996774e-05,
+      "loss": 4.1094,
+      "step": 624640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.976429591126095e-05,
+      "loss": 4.1139,
+      "step": 625152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.975590996375043e-05,
+      "loss": 4.1119,
+      "step": 625664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.974752401623991e-05,
+      "loss": 4.0996,
+      "step": 626176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973913806872939e-05,
+      "loss": 4.1174,
+      "step": 626688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.973075212121887e-05,
+      "loss": 4.1192,
+      "step": 627200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9722366173708356e-05,
+      "loss": 4.1129,
+      "step": 627712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9713980226197836e-05,
+      "loss": 4.1042,
+      "step": 628224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9705594278687316e-05,
+      "loss": 4.1172,
+      "step": 628736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9697208331176796e-05,
+      "loss": 4.1091,
+      "step": 629248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9688822383666276e-05,
+      "loss": 4.1134,
+      "step": 629760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9680436436155756e-05,
+      "loss": 4.1156,
+      "step": 630272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9672066867448965e-05,
+      "loss": 4.1033,
+      "step": 630784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9663680919938445e-05,
+      "loss": 4.1003,
+      "step": 631296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9655294972427925e-05,
+      "loss": 4.0952,
+      "step": 631808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9646909024917405e-05,
+      "loss": 4.117,
+      "step": 632320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9638523077406885e-05,
+      "loss": 4.1062,
+      "step": 632832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.963013712989636e-05,
+      "loss": 4.1007,
+      "step": 633344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.962175118238584e-05,
+      "loss": 4.1148,
+      "step": 633856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9613365234875325e-05,
+      "loss": 4.097,
+      "step": 634368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.960499566616854e-05,
+      "loss": 4.1172,
+      "step": 634880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.959662609746175e-05,
+      "loss": 4.1042,
+      "step": 635392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.958824014995123e-05,
+      "loss": 4.1112,
+      "step": 635904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957985420244071e-05,
+      "loss": 4.1048,
+      "step": 636416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.957146825493018e-05,
+      "loss": 4.1043,
+      "step": 636928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.956308230741966e-05,
+      "loss": 4.1082,
+      "step": 637440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.955471273871288e-05,
+      "loss": 4.0952,
+      "step": 637952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.954634317000609e-05,
+      "loss": 4.1078,
+      "step": 638464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.953795722249557e-05,
+      "loss": 4.103,
+      "step": 638976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952957127498505e-05,
+      "loss": 4.1045,
+      "step": 639488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.952118532747453e-05,
+      "loss": 4.1016,
+      "step": 640000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.951279937996401e-05,
+      "loss": 4.1156,
+      "step": 640512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.950441343245349e-05,
+      "loss": 4.0885,
+      "step": 641024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.949602748494297e-05,
+      "loss": 4.101,
+      "step": 641536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.948764153743245e-05,
+      "loss": 4.0976,
+      "step": 642048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.947925558992193e-05,
+      "loss": 4.0997,
+      "step": 642560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9470886021215136e-05,
+      "loss": 4.1092,
+      "step": 643072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9462500073704616e-05,
+      "loss": 4.0961,
+      "step": 643584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9454114126194096e-05,
+      "loss": 4.107,
+      "step": 644096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9445744557487305e-05,
+      "loss": 4.0794,
+      "step": 644608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9437358609976785e-05,
+      "loss": 4.105,
+      "step": 645120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9428972662466265e-05,
+      "loss": 4.0931,
+      "step": 645632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9420586714955745e-05,
+      "loss": 4.0962,
+      "step": 646144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.941220076744523e-05,
+      "loss": 4.0999,
+      "step": 646656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.940381481993471e-05,
+      "loss": 4.1025,
+      "step": 647168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.939542887242419e-05,
+      "loss": 4.1007,
+      "step": 647680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.938704292491367e-05,
+      "loss": 4.1147,
+      "step": 648192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937865697740315e-05,
+      "loss": 4.1005,
+      "step": 648704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.937027102989263e-05,
+      "loss": 4.1091,
+      "step": 649216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.936190146118584e-05,
+      "loss": 4.1047,
+      "step": 649728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.935353189247905e-05,
+      "loss": 4.0906,
+      "step": 650240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.934514594496853e-05,
+      "loss": 4.0911,
+      "step": 650752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.933675999745801e-05,
+      "loss": 4.1028,
+      "step": 651264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.932837404994749e-05,
+      "loss": 4.1077,
+      "step": 651776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.931998810243697e-05,
+      "loss": 4.1024,
+      "step": 652288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9311618533730186e-05,
+      "loss": 4.0802,
+      "step": 652800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9303232586219666e-05,
+      "loss": 4.0896,
+      "step": 653312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9294846638709146e-05,
+      "loss": 4.0926,
+      "step": 653824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9286460691198626e-05,
+      "loss": 4.1008,
+      "step": 654336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9278074743688105e-05,
+      "loss": 4.1012,
+      "step": 654848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9269688796177585e-05,
+      "loss": 4.1053,
+      "step": 655360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9261302848667065e-05,
+      "loss": 4.0973,
+      "step": 655872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9252916901156545e-05,
+      "loss": 4.0952,
+      "step": 656384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924453095364602e-05,
+      "loss": 4.0929,
+      "step": 656896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9236161384939234e-05,
+      "loss": 4.0923,
+      "step": 657408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9227775437428714e-05,
+      "loss": 4.0856,
+      "step": 657920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9219405868721923e-05,
+      "loss": 4.0984,
+      "step": 658432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9211019921211403e-05,
+      "loss": 4.0903,
+      "step": 658944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.920263397370088e-05,
+      "loss": 4.0957,
+      "step": 659456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.919424802619036e-05,
+      "loss": 4.0946,
+      "step": 659968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.918586207867984e-05,
+      "loss": 4.1118,
+      "step": 660480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.917749250997306e-05,
+      "loss": 4.0888,
+      "step": 660992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916910656246254e-05,
+      "loss": 4.1039,
+      "step": 661504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.916072061495202e-05,
+      "loss": 4.102,
+      "step": 662016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.915233466744149e-05,
+      "loss": 4.0938,
+      "step": 662528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.914394871993097e-05,
+      "loss": 4.0982,
+      "step": 663040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.913557915122419e-05,
+      "loss": 4.0989,
+      "step": 663552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.912719320371367e-05,
+      "loss": 4.103,
+      "step": 664064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911880725620314e-05,
+      "loss": 4.0911,
+      "step": 664576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.911042130869262e-05,
+      "loss": 4.0835,
+      "step": 665088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.910205173998584e-05,
+      "loss": 4.087,
+      "step": 665600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.909366579247532e-05,
+      "loss": 4.0953,
+      "step": 666112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90852798449648e-05,
+      "loss": 4.0938,
+      "step": 666624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.907689389745428e-05,
+      "loss": 4.0928,
+      "step": 667136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906850794994376e-05,
+      "loss": 4.0692,
+      "step": 667648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.906012200243324e-05,
+      "loss": 4.093,
+      "step": 668160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.905173605492272e-05,
+      "loss": 4.098,
+      "step": 668672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.90433501074122e-05,
+      "loss": 4.0942,
+      "step": 669184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903496415990168e-05,
+      "loss": 4.1042,
+      "step": 669696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9026594591194886e-05,
+      "loss": 4.0973,
+      "step": 670208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9018208643684366e-05,
+      "loss": 4.0878,
+      "step": 670720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9009822696173846e-05,
+      "loss": 4.0992,
+      "step": 671232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9001436748663326e-05,
+      "loss": 4.0965,
+      "step": 671744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.899306717995654e-05,
+      "loss": 4.0838,
+      "step": 672256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.898468123244602e-05,
+      "loss": 4.0931,
+      "step": 672768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89762952849355e-05,
+      "loss": 4.09,
+      "step": 673280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.896790933742498e-05,
+      "loss": 4.0893,
+      "step": 673792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895953976871819e-05,
+      "loss": 4.0896,
+      "step": 674304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.895115382120767e-05,
+      "loss": 4.0947,
+      "step": 674816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.894276787369715e-05,
+      "loss": 4.0932,
+      "step": 675328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.893438192618663e-05,
+      "loss": 4.0815,
+      "step": 675840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.892599597867611e-05,
+      "loss": 4.0974,
+      "step": 676352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.891761003116559e-05,
+      "loss": 4.0836,
+      "step": 676864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.89092404624588e-05,
+      "loss": 4.0979,
+      "step": 677376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.890085451494828e-05,
+      "loss": 4.1049,
+      "step": 677888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.889246856743776e-05,
+      "loss": 4.0923,
+      "step": 678400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.888408261992724e-05,
+      "loss": 4.0939,
+      "step": 678912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8875696672416726e-05,
+      "loss": 4.0884,
+      "step": 679424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8867310724906206e-05,
+      "loss": 4.0915,
+      "step": 679936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885892477739568e-05,
+      "loss": 4.0902,
+      "step": 680448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.885053882988516e-05,
+      "loss": 4.1029,
+      "step": 680960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8842169261178375e-05,
+      "loss": 4.104,
+      "step": 681472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8833783313667855e-05,
+      "loss": 4.0946,
+      "step": 681984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8825413744961064e-05,
+      "loss": 4.0929,
+      "step": 682496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8817027797450544e-05,
+      "loss": 4.0952,
+      "step": 683008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880865822874375e-05,
+      "loss": 4.0902,
+      "step": 683520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.880027228123323e-05,
+      "loss": 4.0884,
+      "step": 684032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.879188633372271e-05,
+      "loss": 4.084,
+      "step": 684544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.878350038621219e-05,
+      "loss": 4.0945,
+      "step": 685056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.877511443870168e-05,
+      "loss": 4.0889,
+      "step": 685568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.876672849119115e-05,
+      "loss": 4.0765,
+      "step": 686080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.875834254368063e-05,
+      "loss": 4.0901,
+      "step": 686592
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.104300022125244,
+      "eval_runtime": 541.7623,
+      "eval_samples_per_second": 704.351,
+      "eval_steps_per_second": 22.011,
+      "step": 686871
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874995659617011e-05,
+      "loss": 4.0894,
+      "step": 687104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.874157064865959e-05,
+      "loss": 4.0975,
+      "step": 687616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.873318470114907e-05,
+      "loss": 4.096,
+      "step": 688128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.872479875363855e-05,
+      "loss": 4.0825,
+      "step": 688640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.871641280612803e-05,
+      "loss": 4.0913,
+      "step": 689152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.870805961622497e-05,
+      "loss": 4.0852,
+      "step": 689664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869967366871445e-05,
+      "loss": 4.082,
+      "step": 690176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.869128772120393e-05,
+      "loss": 4.0848,
+      "step": 690688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.868290177369342e-05,
+      "loss": 4.081,
+      "step": 691200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.86745158261829e-05,
+      "loss": 4.0938,
+      "step": 691712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8666146257476106e-05,
+      "loss": 4.094,
+      "step": 692224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8657760309965586e-05,
+      "loss": 4.0826,
+      "step": 692736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8649374362455066e-05,
+      "loss": 4.091,
+      "step": 693248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8640988414944546e-05,
+      "loss": 4.0766,
+      "step": 693760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8632602467434026e-05,
+      "loss": 4.0938,
+      "step": 694272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8624232898727235e-05,
+      "loss": 4.0744,
+      "step": 694784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8615846951216715e-05,
+      "loss": 4.084,
+      "step": 695296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8607461003706195e-05,
+      "loss": 4.0868,
+      "step": 695808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8599075056195675e-05,
+      "loss": 4.0842,
+      "step": 696320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.8590705487488884e-05,
+      "loss": 4.0954,
+      "step": 696832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.858231953997837e-05,
+      "loss": 4.0799,
+      "step": 697344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.857393359246785e-05,
+      "loss": 4.1051,
+      "step": 697856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.856554764495733e-05,
+      "loss": 4.0845,
+      "step": 698368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.855716169744681e-05,
+      "loss": 4.087,
+      "step": 698880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854877574993629e-05,
+      "loss": 4.0847,
+      "step": 699392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.854038980242577e-05,
+      "loss": 4.0924,
+      "step": 699904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.853202023371898e-05,
+      "loss": 4.0901,
+      "step": 700416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.852363428620846e-05,
+      "loss": 4.0792,
+      "step": 700928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.851524833869794e-05,
+      "loss": 4.0877,
+      "step": 701440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.850686239118742e-05,
+      "loss": 4.088,
+      "step": 701952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.84984764436769e-05,
+      "loss": 4.0671,
+      "step": 702464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.849009049616638e-05,
+      "loss": 4.0913,
+      "step": 702976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.848170454865586e-05,
+      "loss": 4.093,
+      "step": 703488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.847331860114533e-05,
+      "loss": 4.0877,
+      "step": 704000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.846494903243855e-05,
+      "loss": 4.0768,
+      "step": 704512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8456563084928035e-05,
+      "loss": 4.0901,
+      "step": 705024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8448177137417515e-05,
+      "loss": 4.0841,
+      "step": 705536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843979118990699e-05,
+      "loss": 4.0855,
+      "step": 706048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.843140524239647e-05,
+      "loss": 4.0884,
+      "step": 706560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.842301929488595e-05,
+      "loss": 4.0761,
+      "step": 707072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8414649726179164e-05,
+      "loss": 4.0762,
+      "step": 707584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.840626377866864e-05,
+      "loss": 4.0682,
+      "step": 708096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.839787783115812e-05,
+      "loss": 4.0902,
+      "step": 708608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.83894918836476e-05,
+      "loss": 4.0804,
+      "step": 709120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.838110593613708e-05,
+      "loss": 4.0742,
+      "step": 709632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.837271998862656e-05,
+      "loss": 4.0862,
+      "step": 710144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.836433404111604e-05,
+      "loss": 4.0709,
+      "step": 710656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.835594809360552e-05,
+      "loss": 4.09,
+      "step": 711168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.834757852489873e-05,
+      "loss": 4.085,
+      "step": 711680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833919257738821e-05,
+      "loss": 4.0827,
+      "step": 712192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.833080662987769e-05,
+      "loss": 4.0796,
+      "step": 712704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.832242068236717e-05,
+      "loss": 4.0763,
+      "step": 713216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.831405111366038e-05,
+      "loss": 4.0843,
+      "step": 713728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.830566516614986e-05,
+      "loss": 4.067,
+      "step": 714240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.829727921863934e-05,
+      "loss": 4.0831,
+      "step": 714752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.828889327112882e-05,
+      "loss": 4.0737,
+      "step": 715264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.82805073236183e-05,
+      "loss": 4.0782,
+      "step": 715776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.827213775491151e-05,
+      "loss": 4.0746,
+      "step": 716288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.826376818620473e-05,
+      "loss": 4.0872,
+      "step": 716800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.825538223869421e-05,
+      "loss": 4.07,
+      "step": 717312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.824699629118369e-05,
+      "loss": 4.0673,
+      "step": 717824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823861034367317e-05,
+      "loss": 4.0766,
+      "step": 718336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.823022439616265e-05,
+      "loss": 4.069,
+      "step": 718848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.822183844865213e-05,
+      "loss": 4.0812,
+      "step": 719360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.821345250114161e-05,
+      "loss": 4.0774,
+      "step": 719872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8205066553631087e-05,
+      "loss": 4.0794,
+      "step": 720384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8196680606120567e-05,
+      "loss": 4.054,
+      "step": 720896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8188294658610047e-05,
+      "loss": 4.0766,
+      "step": 721408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.817990871109952e-05,
+      "loss": 4.0703,
+      "step": 721920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8171522763589e-05,
+      "loss": 4.0691,
+      "step": 722432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8163153194882216e-05,
+      "loss": 4.0734,
+      "step": 722944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8154783626175425e-05,
+      "loss": 4.0754,
+      "step": 723456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.814641405746864e-05,
+      "loss": 4.0756,
+      "step": 723968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.813802810995812e-05,
+      "loss": 4.0893,
+      "step": 724480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.81296421624476e-05,
+      "loss": 4.0757,
+      "step": 724992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.812125621493708e-05,
+      "loss": 4.0836,
+      "step": 725504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.811287026742656e-05,
+      "loss": 4.079,
+      "step": 726016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.810448431991604e-05,
+      "loss": 4.0702,
+      "step": 726528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.809609837240552e-05,
+      "loss": 4.0661,
+      "step": 727040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8087712424894993e-05,
+      "loss": 4.0761,
+      "step": 727552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807934285618821e-05,
+      "loss": 4.0788,
+      "step": 728064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.807095690867769e-05,
+      "loss": 4.0833,
+      "step": 728576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.806257096116717e-05,
+      "loss": 4.0531,
+      "step": 729088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.805418501365665e-05,
+      "loss": 4.0622,
+      "step": 729600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.804579906614613e-05,
+      "loss": 4.0669,
+      "step": 730112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8037429497439345e-05,
+      "loss": 4.0757,
+      "step": 730624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.8029043549928825e-05,
+      "loss": 4.0738,
+      "step": 731136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.80206576024183e-05,
+      "loss": 4.0824,
+      "step": 731648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.801227165490778e-05,
+      "loss": 4.0733,
+      "step": 732160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.800388570739726e-05,
+      "loss": 4.0715,
+      "step": 732672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.799549975988674e-05,
+      "loss": 4.0675,
+      "step": 733184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.798711381237622e-05,
+      "loss": 4.068,
+      "step": 733696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79787278648657e-05,
+      "loss": 4.0592,
+      "step": 734208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.797037467496264e-05,
+      "loss": 4.0723,
+      "step": 734720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7961988727452116e-05,
+      "loss": 4.0679,
+      "step": 735232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7953602779941596e-05,
+      "loss": 4.0686,
+      "step": 735744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.794523321123481e-05,
+      "loss": 4.0714,
+      "step": 736256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.79368472637243e-05,
+      "loss": 4.0844,
+      "step": 736768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792846131621377e-05,
+      "loss": 4.0661,
+      "step": 737280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.792007536870325e-05,
+      "loss": 4.0788,
+      "step": 737792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.791168942119273e-05,
+      "loss": 4.0774,
+      "step": 738304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.790330347368221e-05,
+      "loss": 4.0732,
+      "step": 738816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.789491752617169e-05,
+      "loss": 4.0676,
+      "step": 739328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.788653157866117e-05,
+      "loss": 4.0752,
+      "step": 739840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.787816200995438e-05,
+      "loss": 4.0786,
+      "step": 740352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786977606244386e-05,
+      "loss": 4.0641,
+      "step": 740864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.786140649373707e-05,
+      "loss": 4.0632,
+      "step": 741376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.785302054622655e-05,
+      "loss": 4.0601,
+      "step": 741888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7844634598716036e-05,
+      "loss": 4.0669,
+      "step": 742400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7836248651205516e-05,
+      "loss": 4.0729,
+      "step": 742912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7827862703694996e-05,
+      "loss": 4.0697,
+      "step": 743424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7819493134988205e-05,
+      "loss": 4.0503,
+      "step": 743936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7811107187477685e-05,
+      "loss": 4.0621,
+      "step": 744448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7802721239967165e-05,
+      "loss": 4.0726,
+      "step": 744960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7794335292456645e-05,
+      "loss": 4.074,
+      "step": 745472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7785949344946125e-05,
+      "loss": 4.0822,
+      "step": 745984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7777563397435605e-05,
+      "loss": 4.0715,
+      "step": 746496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7769193828728814e-05,
+      "loss": 4.0654,
+      "step": 747008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7760807881218294e-05,
+      "loss": 4.0719,
+      "step": 747520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7752421933707774e-05,
+      "loss": 4.074,
+      "step": 748032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7744035986197254e-05,
+      "loss": 4.0646,
+      "step": 748544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7735650038686734e-05,
+      "loss": 4.0641,
+      "step": 749056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.772726409117622e-05,
+      "loss": 4.068,
+      "step": 749568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.77188781436657e-05,
+      "loss": 4.0604,
+      "step": 750080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.771049219615518e-05,
+      "loss": 4.0697,
+      "step": 750592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.770212262744839e-05,
+      "loss": 4.0681,
+      "step": 751104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.769373667993787e-05,
+      "loss": 4.072,
+      "step": 751616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.768535073242735e-05,
+      "loss": 4.0574,
+      "step": 752128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.767698116372056e-05,
+      "loss": 4.0731,
+      "step": 752640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766859521621004e-05,
+      "loss": 4.0615,
+      "step": 753152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.766020926869952e-05,
+      "loss": 4.0757,
+      "step": 753664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7651823321189e-05,
+      "loss": 4.08,
+      "step": 754176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.764345375248221e-05,
+      "loss": 4.0675,
+      "step": 754688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.763506780497169e-05,
+      "loss": 4.0679,
+      "step": 755200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7626681857461175e-05,
+      "loss": 4.0704,
+      "step": 755712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7618295909950654e-05,
+      "loss": 4.0661,
+      "step": 756224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760990996244013e-05,
+      "loss": 4.0655,
+      "step": 756736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.760152401492961e-05,
+      "loss": 4.0787,
+      "step": 757248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7593154446222823e-05,
+      "loss": 4.0766,
+      "step": 757760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7584768498712303e-05,
+      "loss": 4.0746,
+      "step": 758272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.757639893000551e-05,
+      "loss": 4.0686,
+      "step": 758784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.756801298249499e-05,
+      "loss": 4.0749,
+      "step": 759296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755962703498447e-05,
+      "loss": 4.064,
+      "step": 759808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.755124108747395e-05,
+      "loss": 4.0625,
+      "step": 760320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.7542855139963426e-05,
+      "loss": 4.0624,
+      "step": 760832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.753446919245291e-05,
+      "loss": 4.0716,
+      "step": 761344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.752608324494239e-05,
+      "loss": 4.0658,
+      "step": 761856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.751769729743187e-05,
+      "loss": 4.051,
+      "step": 762368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.750931134992135e-05,
+      "loss": 4.0684,
+      "step": 762880
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0882673263549805,
+      "eval_runtime": 539.6559,
+      "eval_samples_per_second": 707.101,
+      "eval_steps_per_second": 22.097,
+      "step": 763190
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.750092540241083e-05,
+      "loss": 4.0868,
+      "step": 763392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.749253945490031e-05,
+      "loss": 4.0715,
+      "step": 763904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.748415350738979e-05,
+      "loss": 4.0768,
+      "step": 764416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.747576755987927e-05,
+      "loss": 4.057,
+      "step": 764928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.746738161236875e-05,
+      "loss": 4.0719,
+      "step": 765440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745899566485823e-05,
+      "loss": 4.0634,
+      "step": 765952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.745060971734771e-05,
+      "loss": 4.0595,
+      "step": 766464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.744222376983719e-05,
+      "loss": 4.0586,
+      "step": 766976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7433837822326665e-05,
+      "loss": 4.0586,
+      "step": 767488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7425451874816145e-05,
+      "loss": 4.0703,
+      "step": 768000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7417065927305625e-05,
+      "loss": 4.068,
+      "step": 768512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7408679979795105e-05,
+      "loss": 4.0602,
+      "step": 769024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.740029403228459e-05,
+      "loss": 4.067,
+      "step": 769536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.739190808477407e-05,
+      "loss": 4.053,
+      "step": 770048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.738352213726355e-05,
+      "loss": 4.073,
+      "step": 770560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.737513618975303e-05,
+      "loss": 4.0525,
+      "step": 771072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.736676662104624e-05,
+      "loss": 4.0578,
+      "step": 771584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.735838067353572e-05,
+      "loss": 4.0638,
+      "step": 772096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.73499947260252e-05,
+      "loss": 4.0625,
+      "step": 772608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.734160877851468e-05,
+      "loss": 4.0688,
+      "step": 773120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.733322283100416e-05,
+      "loss": 4.054,
+      "step": 773632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.732483688349364e-05,
+      "loss": 4.0817,
+      "step": 774144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.731646731478685e-05,
+      "loss": 4.0612,
+      "step": 774656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.730808136727633e-05,
+      "loss": 4.0685,
+      "step": 775168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729969541976581e-05,
+      "loss": 4.0608,
+      "step": 775680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.729130947225529e-05,
+      "loss": 4.0723,
+      "step": 776192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7282923524744776e-05,
+      "loss": 4.0651,
+      "step": 776704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7274537577234256e-05,
+      "loss": 4.0581,
+      "step": 777216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7266151629723736e-05,
+      "loss": 4.0674,
+      "step": 777728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.7257782061016945e-05,
+      "loss": 4.0648,
+      "step": 778240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7249396113506425e-05,
+      "loss": 4.0473,
+      "step": 778752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7241010165995905e-05,
+      "loss": 4.0662,
+      "step": 779264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.723262421848538e-05,
+      "loss": 4.07,
+      "step": 779776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.722423827097486e-05,
+      "loss": 4.0635,
+      "step": 780288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.721585232346434e-05,
+      "loss": 4.0536,
+      "step": 780800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.720746637595382e-05,
+      "loss": 4.0723,
+      "step": 781312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.719909680724703e-05,
+      "loss": 4.0571,
+      "step": 781824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7190710859736514e-05,
+      "loss": 4.069,
+      "step": 782336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7182324912225994e-05,
+      "loss": 4.0632,
+      "step": 782848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.717395534351921e-05,
+      "loss": 4.0525,
+      "step": 783360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.716556939600868e-05,
+      "loss": 4.0561,
+      "step": 783872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.715718344849816e-05,
+      "loss": 4.0443,
+      "step": 784384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714879750098764e-05,
+      "loss": 4.0642,
+      "step": 784896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.714041155347712e-05,
+      "loss": 4.059,
+      "step": 785408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.71320256059666e-05,
+      "loss": 4.052,
+      "step": 785920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.712363965845608e-05,
+      "loss": 4.0649,
+      "step": 786432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.711525371094556e-05,
+      "loss": 4.051,
+      "step": 786944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.710686776343504e-05,
+      "loss": 4.0655,
+      "step": 787456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709849819472825e-05,
+      "loss": 4.0609,
+      "step": 787968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.709011224721773e-05,
+      "loss": 4.0569,
+      "step": 788480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.708172629970721e-05,
+      "loss": 4.059,
+      "step": 788992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.707334035219669e-05,
+      "loss": 4.0551,
+      "step": 789504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.706497078348991e-05,
+      "loss": 4.0598,
+      "step": 790016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.705658483597939e-05,
+      "loss": 4.0478,
+      "step": 790528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.704819888846887e-05,
+      "loss": 4.059,
+      "step": 791040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703981294095835e-05,
+      "loss": 4.0535,
+      "step": 791552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.703142699344783e-05,
+      "loss": 4.0571,
+      "step": 792064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7023057424741036e-05,
+      "loss": 4.0528,
+      "step": 792576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7014671477230516e-05,
+      "loss": 4.0624,
+      "step": 793088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7006285529719996e-05,
+      "loss": 4.048,
+      "step": 793600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6997899582209476e-05,
+      "loss": 4.0485,
+      "step": 794112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6989513634698956e-05,
+      "loss": 4.0548,
+      "step": 794624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6981127687188436e-05,
+      "loss": 4.0467,
+      "step": 795136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6972758118481645e-05,
+      "loss": 4.0594,
+      "step": 795648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.696437217097113e-05,
+      "loss": 4.0565,
+      "step": 796160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.695600260226434e-05,
+      "loss": 4.0576,
+      "step": 796672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.694761665475382e-05,
+      "loss": 4.0357,
+      "step": 797184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.69392307072433e-05,
+      "loss": 4.058,
+      "step": 797696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.693084475973278e-05,
+      "loss": 4.0475,
+      "step": 798208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.692245881222226e-05,
+      "loss": 4.0478,
+      "step": 798720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.691407286471174e-05,
+      "loss": 4.0493,
+      "step": 799232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6905686917201214e-05,
+      "loss": 4.0554,
+      "step": 799744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6897300969690694e-05,
+      "loss": 4.0493,
+      "step": 800256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6888915022180174e-05,
+      "loss": 4.0689,
+      "step": 800768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6880529074669654e-05,
+      "loss": 4.0575,
+      "step": 801280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.6872143127159134e-05,
+      "loss": 4.0593,
+      "step": 801792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.686377355845235e-05,
+      "loss": 4.0577,
+      "step": 802304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.685538761094183e-05,
+      "loss": 4.0477,
+      "step": 802816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.684700166343131e-05,
+      "loss": 4.0466,
+      "step": 803328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683861571592079e-05,
+      "loss": 4.0551,
+      "step": 803840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.683022976841027e-05,
+      "loss": 4.0562,
+      "step": 804352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.682184382089975e-05,
+      "loss": 4.0621,
+      "step": 804864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.681345787338923e-05,
+      "loss": 4.0345,
+      "step": 805376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.680507192587871e-05,
+      "loss": 4.0396,
+      "step": 805888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.679670235717192e-05,
+      "loss": 4.0424,
+      "step": 806400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.678833278846513e-05,
+      "loss": 4.059,
+      "step": 806912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677994684095461e-05,
+      "loss": 4.0503,
+      "step": 807424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.677156089344409e-05,
+      "loss": 4.0617,
+      "step": 807936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.676317494593357e-05,
+      "loss": 4.0533,
+      "step": 808448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.675480537722678e-05,
+      "loss": 4.0532,
+      "step": 808960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.674643580851999e-05,
+      "loss": 4.0427,
+      "step": 809472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.673804986100947e-05,
+      "loss": 4.0507,
+      "step": 809984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672966391349895e-05,
+      "loss": 4.0404,
+      "step": 810496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672127796598843e-05,
+      "loss": 4.0482,
+      "step": 811008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.671289201847791e-05,
+      "loss": 4.0465,
+      "step": 811520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.670450607096739e-05,
+      "loss": 4.0489,
+      "step": 812032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.669612012345687e-05,
+      "loss": 4.0517,
+      "step": 812544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.668775055475008e-05,
+      "loss": 4.0595,
+      "step": 813056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667936460723956e-05,
+      "loss": 4.051,
+      "step": 813568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.667097865972904e-05,
+      "loss": 4.0573,
+      "step": 814080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.666259271221852e-05,
+      "loss": 4.0531,
+      "step": 814592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.665420676470801e-05,
+      "loss": 4.054,
+      "step": 815104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.664582081719749e-05,
+      "loss": 4.0492,
+      "step": 815616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.663743486968697e-05,
+      "loss": 4.0562,
+      "step": 816128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662904892217645e-05,
+      "loss": 4.0549,
+      "step": 816640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.662066297466593e-05,
+      "loss": 4.0445,
+      "step": 817152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.66122770271554e-05,
+      "loss": 4.0421,
+      "step": 817664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.660390745844862e-05,
+      "loss": 4.0381,
+      "step": 818176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65955215109381e-05,
+      "loss": 4.0496,
+      "step": 818688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6587135563427577e-05,
+      "loss": 4.0477,
+      "step": 819200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6578765994720786e-05,
+      "loss": 4.0534,
+      "step": 819712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6570380047210266e-05,
+      "loss": 4.0267,
+      "step": 820224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6561994099699746e-05,
+      "loss": 4.0361,
+      "step": 820736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6553608152189226e-05,
+      "loss": 4.0537,
+      "step": 821248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6545222204678705e-05,
+      "loss": 4.054,
+      "step": 821760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6536836257168185e-05,
+      "loss": 4.0617,
+      "step": 822272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.65284666884614e-05,
+      "loss": 4.0538,
+      "step": 822784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6520080740950875e-05,
+      "loss": 4.0391,
+      "step": 823296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6511694793440354e-05,
+      "loss": 4.0522,
+      "step": 823808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6503308845929834e-05,
+      "loss": 4.0516,
+      "step": 824320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6494922898419314e-05,
+      "loss": 4.0427,
+      "step": 824832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6486536950908794e-05,
+      "loss": 4.0436,
+      "step": 825344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6478151003398274e-05,
+      "loss": 4.0473,
+      "step": 825856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6469765055887754e-05,
+      "loss": 4.0437,
+      "step": 826368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6461379108377234e-05,
+      "loss": 4.0466,
+      "step": 826880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6452993160866714e-05,
+      "loss": 4.0447,
+      "step": 827392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6444607213356194e-05,
+      "loss": 4.0547,
+      "step": 827904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6436221265845674e-05,
+      "loss": 4.0373,
+      "step": 828416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.642785169713889e-05,
+      "loss": 4.0477,
+      "step": 828928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641946574962837e-05,
+      "loss": 4.0463,
+      "step": 829440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.641107980211785e-05,
+      "loss": 4.0568,
+      "step": 829952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.640271023341106e-05,
+      "loss": 4.0577,
+      "step": 830464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.639432428590054e-05,
+      "loss": 4.0462,
+      "step": 830976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.638593833839002e-05,
+      "loss": 4.0484,
+      "step": 831488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.63775523908795e-05,
+      "loss": 4.0505,
+      "step": 832000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636916644336898e-05,
+      "loss": 4.046,
+      "step": 832512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.636079687466219e-05,
+      "loss": 4.0447,
+      "step": 833024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.635241092715167e-05,
+      "loss": 4.0587,
+      "step": 833536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.634402497964115e-05,
+      "loss": 4.0572,
+      "step": 834048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.633563903213063e-05,
+      "loss": 4.0554,
+      "step": 834560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.632725308462011e-05,
+      "loss": 4.0494,
+      "step": 835072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6318883515913324e-05,
+      "loss": 4.0553,
+      "step": 835584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6310497568402804e-05,
+      "loss": 4.0401,
+      "step": 836096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6302111620892283e-05,
+      "loss": 4.0433,
+      "step": 836608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6293725673381763e-05,
+      "loss": 4.0454,
+      "step": 837120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.628535610467497e-05,
+      "loss": 4.0489,
+      "step": 837632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.627697015716445e-05,
+      "loss": 4.0451,
+      "step": 838144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626858420965393e-05,
+      "loss": 4.0341,
+      "step": 838656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.626019826214341e-05,
+      "loss": 4.0494,
+      "step": 839168
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.074805736541748,
+      "eval_runtime": 553.552,
+      "eval_samples_per_second": 689.35,
+      "eval_steps_per_second": 21.543,
+      "step": 839509
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6251812314632886e-05,
+      "loss": 4.0461,
+      "step": 839680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6243426367122366e-05,
+      "loss": 4.0529,
+      "step": 840192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6235040419611845e-05,
+      "loss": 4.0513,
+      "step": 840704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.622665447210133e-05,
+      "loss": 4.0395,
+      "step": 841216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.621826852459081e-05,
+      "loss": 4.0496,
+      "step": 841728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.620989895588402e-05,
+      "loss": 4.0442,
+      "step": 842240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.62015130083735e-05,
+      "loss": 4.0375,
+      "step": 842752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.619312706086298e-05,
+      "loss": 4.0413,
+      "step": 843264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.618474111335246e-05,
+      "loss": 4.0424,
+      "step": 843776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.617635516584194e-05,
+      "loss": 4.0474,
+      "step": 844288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.616798559713515e-05,
+      "loss": 4.0457,
+      "step": 844800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615959964962463e-05,
+      "loss": 4.0413,
+      "step": 845312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.615121370211411e-05,
+      "loss": 4.0473,
+      "step": 845824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.614282775460359e-05,
+      "loss": 4.0388,
+      "step": 846336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.613444180709307e-05,
+      "loss": 4.0465,
+      "step": 846848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.612605585958255e-05,
+      "loss": 4.0345,
+      "step": 847360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6117686290875766e-05,
+      "loss": 4.0369,
+      "step": 847872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6109300343365246e-05,
+      "loss": 4.0446,
+      "step": 848384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6100914395854726e-05,
+      "loss": 4.0398,
+      "step": 848896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6092528448344206e-05,
+      "loss": 4.0491,
+      "step": 849408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6084142500833686e-05,
+      "loss": 4.0392,
+      "step": 849920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6075756553323166e-05,
+      "loss": 4.0621,
+      "step": 850432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6067370605812646e-05,
+      "loss": 4.0419,
+      "step": 850944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6059001037105855e-05,
+      "loss": 4.045,
+      "step": 851456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6050615089595335e-05,
+      "loss": 4.0421,
+      "step": 851968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6042229142084815e-05,
+      "loss": 4.0533,
+      "step": 852480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6033843194574295e-05,
+      "loss": 4.0444,
+      "step": 852992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6025457247063774e-05,
+      "loss": 4.0371,
+      "step": 853504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6017071299553254e-05,
+      "loss": 4.0472,
+      "step": 854016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.6008685352042734e-05,
+      "loss": 4.0479,
+      "step": 854528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.600031578333595e-05,
+      "loss": 4.0295,
+      "step": 855040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5991929835825423e-05,
+      "loss": 4.042,
+      "step": 855552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5983543888314903e-05,
+      "loss": 4.0517,
+      "step": 856064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.597515794080438e-05,
+      "loss": 4.048,
+      "step": 856576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.596677199329386e-05,
+      "loss": 4.0334,
+      "step": 857088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595838604578334e-05,
+      "loss": 4.047,
+      "step": 857600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.595001647707655e-05,
+      "loss": 4.0428,
+      "step": 858112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.594163052956603e-05,
+      "loss": 4.0461,
+      "step": 858624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.593324458205551e-05,
+      "loss": 4.0419,
+      "step": 859136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.592485863454499e-05,
+      "loss": 4.0376,
+      "step": 859648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.591647268703447e-05,
+      "loss": 4.0357,
+      "step": 860160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.590808673952395e-05,
+      "loss": 4.0252,
+      "step": 860672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589971717081717e-05,
+      "loss": 4.0431,
+      "step": 861184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.589133122330665e-05,
+      "loss": 4.0429,
+      "step": 861696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.588294527579613e-05,
+      "loss": 4.0362,
+      "step": 862208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.587455932828561e-05,
+      "loss": 4.0419,
+      "step": 862720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.586617338077509e-05,
+      "loss": 4.0333,
+      "step": 863232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.585778743326457e-05,
+      "loss": 4.0517,
+      "step": 863744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584940148575405e-05,
+      "loss": 4.0412,
+      "step": 864256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.584103191704726e-05,
+      "loss": 4.0317,
+      "step": 864768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.583264596953674e-05,
+      "loss": 4.0384,
+      "step": 865280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.582426002202622e-05,
+      "loss": 4.0344,
+      "step": 865792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.58158740745157e-05,
+      "loss": 4.0464,
+      "step": 866304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.580748812700518e-05,
+      "loss": 4.0223,
+      "step": 866816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.579910217949466e-05,
+      "loss": 4.0439,
+      "step": 867328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5790716231984137e-05,
+      "loss": 4.0334,
+      "step": 867840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5782330284473617e-05,
+      "loss": 4.039,
+      "step": 868352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5773944336963096e-05,
+      "loss": 4.0355,
+      "step": 868864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.576557476825631e-05,
+      "loss": 4.0453,
+      "step": 869376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.575720519954952e-05,
+      "loss": 4.0261,
+      "step": 869888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574883563084273e-05,
+      "loss": 4.0295,
+      "step": 870400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.574044968333221e-05,
+      "loss": 4.0358,
+      "step": 870912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.573206373582169e-05,
+      "loss": 4.0279,
+      "step": 871424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.572367778831117e-05,
+      "loss": 4.0385,
+      "step": 871936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.571529184080065e-05,
+      "loss": 4.0363,
+      "step": 872448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.570690589329013e-05,
+      "loss": 4.042,
+      "step": 872960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.569853632458334e-05,
+      "loss": 4.0131,
+      "step": 873472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5690150377072826e-05,
+      "loss": 4.0357,
+      "step": 873984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5681764429562306e-05,
+      "loss": 4.0314,
+      "step": 874496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5673378482051786e-05,
+      "loss": 4.0265,
+      "step": 875008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5665008913344995e-05,
+      "loss": 4.0317,
+      "step": 875520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5656622965834475e-05,
+      "loss": 4.0399,
+      "step": 876032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5648237018323955e-05,
+      "loss": 4.028,
+      "step": 876544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5639851070813435e-05,
+      "loss": 4.0498,
+      "step": 877056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.563146512330291e-05,
+      "loss": 4.0378,
+      "step": 877568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5623095554596124e-05,
+      "loss": 4.0413,
+      "step": 878080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5614709607085604e-05,
+      "loss": 4.039,
+      "step": 878592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.560632365957508e-05,
+      "loss": 4.0271,
+      "step": 879104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5597937712064564e-05,
+      "loss": 4.029,
+      "step": 879616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5589551764554044e-05,
+      "loss": 4.0349,
+      "step": 880128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.5581165817043524e-05,
+      "loss": 4.0391,
+      "step": 880640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.557279624833673e-05,
+      "loss": 4.0381,
+      "step": 881152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.556441030082621e-05,
+      "loss": 4.0195,
+      "step": 881664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.555602435331569e-05,
+      "loss": 4.0233,
+      "step": 882176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.554763840580517e-05,
+      "loss": 4.021,
+      "step": 882688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553925245829465e-05,
+      "loss": 4.0416,
+      "step": 883200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.553086651078413e-05,
+      "loss": 4.0301,
+      "step": 883712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.552248056327361e-05,
+      "loss": 4.0409,
+      "step": 884224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.551409461576309e-05,
+      "loss": 4.0347,
+      "step": 884736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.550570866825257e-05,
+      "loss": 4.0318,
+      "step": 885248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.549732272074205e-05,
+      "loss": 4.026,
+      "step": 885760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548893677323153e-05,
+      "loss": 4.0296,
+      "step": 886272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.548056720452475e-05,
+      "loss": 4.0263,
+      "step": 886784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.547218125701423e-05,
+      "loss": 4.0258,
+      "step": 887296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.546379530950371e-05,
+      "loss": 4.025,
+      "step": 887808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.545540936199319e-05,
+      "loss": 4.0322,
+      "step": 888320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.544702341448267e-05,
+      "loss": 4.033,
+      "step": 888832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543863746697215e-05,
+      "loss": 4.0396,
+      "step": 889344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.543026789826536e-05,
+      "loss": 4.0334,
+      "step": 889856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.542188195075484e-05,
+      "loss": 4.0363,
+      "step": 890368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.541349600324432e-05,
+      "loss": 4.0348,
+      "step": 890880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.54051100557338e-05,
+      "loss": 4.0383,
+      "step": 891392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.539672410822327e-05,
+      "loss": 4.0325,
+      "step": 891904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.538833816071275e-05,
+      "loss": 4.0395,
+      "step": 892416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.537995221320223e-05,
+      "loss": 4.0305,
+      "step": 892928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5371582644495446e-05,
+      "loss": 4.0308,
+      "step": 893440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.536321307578866e-05,
+      "loss": 4.0261,
+      "step": 893952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.535482712827814e-05,
+      "loss": 4.014,
+      "step": 894464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.534644118076762e-05,
+      "loss": 4.0328,
+      "step": 894976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5338055233257095e-05,
+      "loss": 4.0344,
+      "step": 895488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5329669285746575e-05,
+      "loss": 4.0298,
+      "step": 896000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.532129971703979e-05,
+      "loss": 4.0114,
+      "step": 896512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5312913769529264e-05,
+      "loss": 4.0159,
+      "step": 897024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5304527822018744e-05,
+      "loss": 4.0353,
+      "step": 897536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5296141874508224e-05,
+      "loss": 4.0364,
+      "step": 898048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5287755926997704e-05,
+      "loss": 4.0429,
+      "step": 898560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.527938635829092e-05,
+      "loss": 4.0361,
+      "step": 899072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.52710004107804e-05,
+      "loss": 4.0276,
+      "step": 899584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.526261446326988e-05,
+      "loss": 4.0293,
+      "step": 900096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.525422851575936e-05,
+      "loss": 4.0337,
+      "step": 900608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.524585894705257e-05,
+      "loss": 4.0233,
+      "step": 901120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.523747299954205e-05,
+      "loss": 4.0243,
+      "step": 901632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522908705203153e-05,
+      "loss": 4.0326,
+      "step": 902144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.522070110452101e-05,
+      "loss": 4.0209,
+      "step": 902656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.521231515701049e-05,
+      "loss": 4.03,
+      "step": 903168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.520392920949997e-05,
+      "loss": 4.0255,
+      "step": 903680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.519554326198945e-05,
+      "loss": 4.036,
+      "step": 904192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.518715731447893e-05,
+      "loss": 4.0192,
+      "step": 904704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517878774577214e-05,
+      "loss": 4.0324,
+      "step": 905216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.517040179826162e-05,
+      "loss": 4.025,
+      "step": 905728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5162015850751104e-05,
+      "loss": 4.0388,
+      "step": 906240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.5153629903240584e-05,
+      "loss": 4.0376,
+      "step": 906752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.514526033453379e-05,
+      "loss": 4.0333,
+      "step": 907264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.513687438702327e-05,
+      "loss": 4.0269,
+      "step": 907776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512848843951275e-05,
+      "loss": 4.0314,
+      "step": 908288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.512010249200223e-05,
+      "loss": 4.0287,
+      "step": 908800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.511173292329544e-05,
+      "loss": 4.0285,
+      "step": 909312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.510334697578492e-05,
+      "loss": 4.0399,
+      "step": 909824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.509497740707813e-05,
+      "loss": 4.0357,
+      "step": 910336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.508659145956761e-05,
+      "loss": 4.0384,
+      "step": 910848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.507820551205709e-05,
+      "loss": 4.0353,
+      "step": 911360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506981956454657e-05,
+      "loss": 4.0352,
+      "step": 911872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.506143361703606e-05,
+      "loss": 4.0235,
+      "step": 912384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.505304766952554e-05,
+      "loss": 4.0256,
+      "step": 912896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.504466172201502e-05,
+      "loss": 4.0259,
+      "step": 913408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.50362757745045e-05,
+      "loss": 4.0337,
+      "step": 913920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.502788982699398e-05,
+      "loss": 4.0288,
+      "step": 914432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501950387948346e-05,
+      "loss": 4.0159,
+      "step": 914944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.501113431077667e-05,
+      "loss": 4.0311,
+      "step": 915456
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.06290864944458,
+      "eval_runtime": 545.3001,
+      "eval_samples_per_second": 699.782,
+      "eval_steps_per_second": 21.869,
+      "step": 915828
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500274836326615e-05,
+      "loss": 4.0439,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499436241575563e-05,
+      "loss": 4.037,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.49859764682451e-05,
+      "loss": 4.0324,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497759052073458e-05,
+      "loss": 4.0231,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496920457322406e-05,
+      "loss": 4.0319,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496081862571354e-05,
+      "loss": 4.0291,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4952432678203026e-05,
+      "loss": 4.0201,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4944046730692506e-05,
+      "loss": 4.0236,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4935660783181986e-05,
+      "loss": 4.0238,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4927274835671466e-05,
+      "loss": 4.0279,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4918888888160946e-05,
+      "loss": 4.0318,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4910502940650426e-05,
+      "loss": 4.0243,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4902116993139906e-05,
+      "loss": 4.0278,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4893731045629386e-05,
+      "loss": 4.0204,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4885345098118866e-05,
+      "loss": 4.0278,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4876959150608346e-05,
+      "loss": 4.0196,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486857320309782e-05,
+      "loss": 4.0239,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4860203634391035e-05,
+      "loss": 4.024,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4851817686880515e-05,
+      "loss": 4.0225,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.484343173936999e-05,
+      "loss": 4.0321,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4835045791859475e-05,
+      "loss": 4.0184,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4826659844348955e-05,
+      "loss": 4.0402,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4818273896838435e-05,
+      "loss": 4.0272,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4809887949327915e-05,
+      "loss": 4.0307,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4801518380621124e-05,
+      "loss": 4.0208,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479314881191434e-05,
+      "loss": 4.0344,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478476286440382e-05,
+      "loss": 4.0323,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477637691689329e-05,
+      "loss": 4.0205,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476799096938277e-05,
+      "loss": 4.0252,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475962140067599e-05,
+      "loss": 4.0319,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475123545316546e-05,
+      "loss": 4.0143,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474286588445868e-05,
+      "loss": 4.023,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473447993694816e-05,
+      "loss": 4.0343,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4726093989437645e-05,
+      "loss": 4.0299,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471770804192712e-05,
+      "loss": 4.0144,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.47093220944166e-05,
+      "loss": 4.026,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470093614690608e-05,
+      "loss": 4.0287,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469256657819929e-05,
+      "loss": 4.0303,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468418063068877e-05,
+      "loss": 4.0224,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467579468317825e-05,
+      "loss": 4.0259,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4667408735667727e-05,
+      "loss": 4.0204,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4659022788157207e-05,
+      "loss": 4.0034,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4650636840646686e-05,
+      "loss": 4.031,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4642250893136166e-05,
+      "loss": 4.0229,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4633864945625646e-05,
+      "loss": 4.017,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4625478998115126e-05,
+      "loss": 4.0253,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.461710942940834e-05,
+      "loss": 4.0154,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.460872348189782e-05,
+      "loss": 4.033,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.46003375343873e-05,
+      "loss": 4.027,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.459195158687678e-05,
+      "loss": 4.0114,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.458356563936626e-05,
+      "loss": 4.023,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.457519607065947e-05,
+      "loss": 4.0182,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.456681012314895e-05,
+      "loss": 4.0289,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455842417563843e-05,
+      "loss": 4.006,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.455003822812791e-05,
+      "loss": 4.0281,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454165228061739e-05,
+      "loss": 4.0101,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45332827119106e-05,
+      "loss": 4.0227,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452489676440008e-05,
+      "loss": 4.0209,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451651081688957e-05,
+      "loss": 4.0286,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4508141248182776e-05,
+      "loss": 4.0132,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4499755300672256e-05,
+      "loss": 4.0089,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4491369353161736e-05,
+      "loss": 4.018,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4482983405651216e-05,
+      "loss": 4.0134,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4474597458140696e-05,
+      "loss": 4.0197,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4466211510630176e-05,
+      "loss": 4.0225,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445782556311965e-05,
+      "loss": 4.0188,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444943961560913e-05,
+      "loss": 4.0016,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444105366809861e-05,
+      "loss": 4.0168,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443266772058809e-05,
+      "loss": 4.017,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4424298151881305e-05,
+      "loss": 4.0092,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4415912204370785e-05,
+      "loss": 4.016,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4407526256860264e-05,
+      "loss": 4.0216,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4399140309349744e-05,
+      "loss": 4.0148,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4390754361839224e-05,
+      "loss": 4.0245,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4382368414328704e-05,
+      "loss": 4.0269,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4373982466818184e-05,
+      "loss": 4.0265,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4365596519307664e-05,
+      "loss": 4.0246,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.435722695060087e-05,
+      "loss": 4.0097,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434884100309035e-05,
+      "loss": 4.01,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434047143438356e-05,
+      "loss": 4.0189,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433208548687304e-05,
+      "loss": 4.026,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432369953936252e-05,
+      "loss": 4.0203,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4315313591852e-05,
+      "loss": 3.9982,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430692764434148e-05,
+      "loss": 4.0125,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42985580756347e-05,
+      "loss": 4.003,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429017212812418e-05,
+      "loss": 4.0244,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428178618061366e-05,
+      "loss": 4.0145,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427340023310314e-05,
+      "loss": 4.0269,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426503066439635e-05,
+      "loss": 4.0144,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425664471688583e-05,
+      "loss": 4.0158,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.424825876937531e-05,
+      "loss": 4.008,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4239889200668516e-05,
+      "loss": 4.0131,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4231503253157996e-05,
+      "loss": 4.0127,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4223117305647476e-05,
+      "loss": 4.0095,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4214731358136956e-05,
+      "loss": 4.0061,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4206345410626436e-05,
+      "loss": 4.0136,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419795946311592e-05,
+      "loss": 4.0165,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41895735156054e-05,
+      "loss": 4.0235,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418118756809488e-05,
+      "loss": 4.0192,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417280162058436e-05,
+      "loss": 4.0203,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4164415673073836e-05,
+      "loss": 4.0172,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156029725563316e-05,
+      "loss": 4.0175,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147643778052796e-05,
+      "loss": 4.0157,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413929058814974e-05,
+      "loss": 4.0238,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.413090464063922e-05,
+      "loss": 4.0148,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41225186931287e-05,
+      "loss": 4.0191,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.411413274561818e-05,
+      "loss": 4.0047,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.410574679810766e-05,
+      "loss": 4.0023,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.409736085059714e-05,
+      "loss": 4.0131,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4088991281890356e-05,
+      "loss": 4.0176,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4080605334379836e-05,
+      "loss": 4.0159,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.407221938686931e-05,
+      "loss": 3.9941,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.406383343935879e-05,
+      "loss": 3.9985,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.405544749184827e-05,
+      "loss": 4.0226,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4047077923141485e-05,
+      "loss": 4.0214,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403869197563096e-05,
+      "loss": 4.0224,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403030602812044e-05,
+      "loss": 4.0194,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402192008060992e-05,
+      "loss": 4.016,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40135341330994e-05,
+      "loss": 4.015,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400514818558888e-05,
+      "loss": 4.015,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399676223807836e-05,
+      "loss": 4.006,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3988376290567845e-05,
+      "loss": 4.0065,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3980006721861054e-05,
+      "loss": 4.0196,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3971620774350534e-05,
+      "loss": 4.0023,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3963234826840014e-05,
+      "loss": 4.0122,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3954848879329494e-05,
+      "loss": 4.0125,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39464793106227e-05,
+      "loss": 4.0173,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393809336311218e-05,
+      "loss": 4.0057,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392970741560166e-05,
+      "loss": 4.0161,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392133784689487e-05,
+      "loss": 4.0132,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391295189938435e-05,
+      "loss": 4.0155,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.390456595187383e-05,
+      "loss": 4.0217,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389618000436331e-05,
+      "loss": 4.0208,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38877940568528e-05,
+      "loss": 4.0124,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387940810934228e-05,
+      "loss": 4.0127,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387102216183176e-05,
+      "loss": 4.0139,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386263621432124e-05,
+      "loss": 4.0064,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385425026681072e-05,
+      "loss": 4.0257,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38458643193002e-05,
+      "loss": 4.0202,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.383749475059341e-05,
+      "loss": 4.0248,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382910880308289e-05,
+      "loss": 4.0171,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382072285557237e-05,
+      "loss": 4.0215,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3812353286865576e-05,
+      "loss": 4.0044,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3803967339355056e-05,
+      "loss": 4.0108,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3795581391844536e-05,
+      "loss": 4.0094,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3787195444334016e-05,
+      "loss": 4.0185,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3778809496823496e-05,
+      "loss": 4.0126,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3770423549312976e-05,
+      "loss": 3.9965,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3762037601802456e-05,
+      "loss": 4.0094,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.053101539611816,
+      "eval_runtime": 564.873,
+      "eval_samples_per_second": 675.534,
+      "eval_steps_per_second": 21.111,
+      "step": 992147
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3753651654291936e-05,
+      "loss": 4.0168,
+      "step": 992256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3745265706781416e-05,
+      "loss": 4.017,
+      "step": 992768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3736879759270896e-05,
+      "loss": 4.0201,
+      "step": 993280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3728493811760376e-05,
+      "loss": 4.0063,
+      "step": 993792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3720124243053585e-05,
+      "loss": 4.0168,
+      "step": 994304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3711738295543065e-05,
+      "loss": 4.0162,
+      "step": 994816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3703352348032545e-05,
+      "loss": 3.9968,
+      "step": 995328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3694966400522025e-05,
+      "loss": 4.0125,
+      "step": 995840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3686596831815234e-05,
+      "loss": 4.0074,
+      "step": 996352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.367822726310845e-05,
+      "loss": 4.005,
+      "step": 996864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.366984131559793e-05,
+      "loss": 4.0168,
+      "step": 997376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3661471746891146e-05,
+      "loss": 4.0117,
+      "step": 997888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.365308579938062e-05,
+      "loss": 4.0041,
+      "step": 998400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36446998518701e-05,
+      "loss": 4.0085,
+      "step": 998912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.363631390435958e-05,
+      "loss": 4.0131,
+      "step": 999424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.362792795684906e-05,
+      "loss": 4.0102,
+      "step": 999936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361954200933854e-05,
+      "loss": 4.0017,
+      "step": 1000448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.361115606182802e-05,
+      "loss": 4.0064,
+      "step": 1000960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.36027701143175e-05,
+      "loss": 4.0076,
+      "step": 1001472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.359438416680698e-05,
+      "loss": 4.0161,
+      "step": 1001984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.358599821929646e-05,
+      "loss": 4.0076,
+      "step": 1002496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.357761227178594e-05,
+      "loss": 4.0206,
+      "step": 1003008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.356922632427542e-05,
+      "loss": 4.0082,
+      "step": 1003520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.35608403767649e-05,
+      "loss": 4.0179,
+      "step": 1004032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3552454429254385e-05,
+      "loss": 4.0088,
+      "step": 1004544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.354406848174386e-05,
+      "loss": 4.0161,
+      "step": 1005056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.353568253423334e-05,
+      "loss": 4.0165,
+      "step": 1005568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3527312965526554e-05,
+      "loss": 4.0063,
+      "step": 1006080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351894339681976e-05,
+      "loss": 4.0098,
+      "step": 1006592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.351055744930924e-05,
+      "loss": 4.016,
+      "step": 1007104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.350217150179872e-05,
+      "loss": 3.9997,
+      "step": 1007616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.34937855542882e-05,
+      "loss": 4.0082,
+      "step": 1008128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.348541598558141e-05,
+      "loss": 4.0186,
+      "step": 1008640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.347703003807089e-05,
+      "loss": 4.0152,
+      "step": 1009152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346864409056037e-05,
+      "loss": 3.9988,
+      "step": 1009664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.346025814304985e-05,
+      "loss": 4.0122,
+      "step": 1010176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.345188857434307e-05,
+      "loss": 4.0143,
+      "step": 1010688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.344350262683255e-05,
+      "loss": 4.0131,
+      "step": 1011200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.343511667932203e-05,
+      "loss": 4.0042,
+      "step": 1011712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.342673073181151e-05,
+      "loss": 4.0111,
+      "step": 1012224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.341834478430098e-05,
+      "loss": 4.006,
+      "step": 1012736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340995883679046e-05,
+      "loss": 3.9877,
+      "step": 1013248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.340157288927994e-05,
+      "loss": 4.0125,
+      "step": 1013760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.339318694176942e-05,
+      "loss": 4.0081,
+      "step": 1014272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.33848009942589e-05,
+      "loss": 4.0027,
+      "step": 1014784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.337641504674838e-05,
+      "loss": 4.0071,
+      "step": 1015296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.336802909923786e-05,
+      "loss": 3.9986,
+      "step": 1015808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3359659530531077e-05,
+      "loss": 4.0185,
+      "step": 1016320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3351273583020557e-05,
+      "loss": 4.0111,
+      "step": 1016832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3342887635510037e-05,
+      "loss": 3.9963,
+      "step": 1017344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3334501687999516e-05,
+      "loss": 4.0081,
+      "step": 1017856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3326132119292726e-05,
+      "loss": 4.0036,
+      "step": 1018368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3317746171782206e-05,
+      "loss": 4.0113,
+      "step": 1018880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3309360224271685e-05,
+      "loss": 3.9875,
+      "step": 1019392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3300974276761165e-05,
+      "loss": 4.0134,
+      "step": 1019904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3292588329250645e-05,
+      "loss": 3.9933,
+      "step": 1020416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3284202381740125e-05,
+      "loss": 4.0065,
+      "step": 1020928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3275816434229605e-05,
+      "loss": 4.0099,
+      "step": 1021440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3267446865522814e-05,
+      "loss": 4.0078,
+      "step": 1021952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3259060918012294e-05,
+      "loss": 4.0028,
+      "step": 1022464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3250674970501774e-05,
+      "loss": 3.9926,
+      "step": 1022976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3242289022991254e-05,
+      "loss": 4.0037,
+      "step": 1023488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.323391945428447e-05,
+      "loss": 4.0002,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.322553350677395e-05,
+      "loss": 4.0015,
+      "step": 1024512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.321714755926343e-05,
+      "loss": 4.0097,
+      "step": 1025024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320876161175291e-05,
+      "loss": 3.9991,
+      "step": 1025536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.320037566424239e-05,
+      "loss": 3.9898,
+      "step": 1026048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.319198971673187e-05,
+      "loss": 3.9958,
+      "step": 1026560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.318360376922134e-05,
+      "loss": 4.005,
+      "step": 1027072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.317521782171082e-05,
+      "loss": 3.9895,
+      "step": 1027584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.316686463180777e-05,
+      "loss": 4.0,
+      "step": 1028096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315847868429725e-05,
+      "loss": 4.0116,
+      "step": 1028608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.315009273678673e-05,
+      "loss": 3.9996,
+      "step": 1029120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.314170678927621e-05,
+      "loss": 4.0053,
+      "step": 1029632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3133320841765695e-05,
+      "loss": 4.0131,
+      "step": 1030144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3124951273058904e-05,
+      "loss": 4.0114,
+      "step": 1030656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3116565325548384e-05,
+      "loss": 4.0115,
+      "step": 1031168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.3108179378037864e-05,
+      "loss": 3.9912,
+      "step": 1031680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309980980933107e-05,
+      "loss": 3.9902,
+      "step": 1032192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.309142386182055e-05,
+      "loss": 4.004,
+      "step": 1032704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.308303791431003e-05,
+      "loss": 4.009,
+      "step": 1033216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.307465196679951e-05,
+      "loss": 4.0075,
+      "step": 1033728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.306626601928899e-05,
+      "loss": 3.986,
+      "step": 1034240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30578964505822e-05,
+      "loss": 3.9979,
+      "step": 1034752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304951050307168e-05,
+      "loss": 3.9913,
+      "step": 1035264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.304112455556116e-05,
+      "loss": 4.0024,
+      "step": 1035776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.303273860805064e-05,
+      "loss": 4.0008,
+      "step": 1036288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.302435266054012e-05,
+      "loss": 4.0142,
+      "step": 1036800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.30159667130296e-05,
+      "loss": 4.0009,
+      "step": 1037312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.300758076551908e-05,
+      "loss": 4.0012,
+      "step": 1037824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299919481800856e-05,
+      "loss": 3.9928,
+      "step": 1038336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.299080887049804e-05,
+      "loss": 3.9949,
+      "step": 1038848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.298242292298752e-05,
+      "loss": 3.9974,
+      "step": 1039360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2974036975477e-05,
+      "loss": 4.0018,
+      "step": 1039872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.296565102796648e-05,
+      "loss": 3.9877,
+      "step": 1040384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.295728145925969e-05,
+      "loss": 3.9987,
+      "step": 1040896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294889551174917e-05,
+      "loss": 3.9998,
+      "step": 1041408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.294050956423865e-05,
+      "loss": 4.0071,
+      "step": 1041920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.293212361672813e-05,
+      "loss": 4.0073,
+      "step": 1042432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2923754048021346e-05,
+      "loss": 4.0069,
+      "step": 1042944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2915368100510826e-05,
+      "loss": 4.004,
+      "step": 1043456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2906982153000306e-05,
+      "loss": 4.0004,
+      "step": 1043968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2898612584293515e-05,
+      "loss": 3.9984,
+      "step": 1044480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2890226636782995e-05,
+      "loss": 4.0149,
+      "step": 1044992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2881840689272475e-05,
+      "loss": 3.9939,
+      "step": 1045504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2873454741761955e-05,
+      "loss": 4.0053,
+      "step": 1046016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2865068794251435e-05,
+      "loss": 3.9897,
+      "step": 1046528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2856682846740915e-05,
+      "loss": 3.9844,
+      "step": 1047040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2848296899230395e-05,
+      "loss": 3.9994,
+      "step": 1047552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2839910951719875e-05,
+      "loss": 4.0046,
+      "step": 1048064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2831541383013084e-05,
+      "loss": 4.001,
+      "step": 1048576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.28231718143063e-05,
+      "loss": 3.9782,
+      "step": 1049088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.281478586679578e-05,
+      "loss": 3.9889,
+      "step": 1049600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.280639991928526e-05,
+      "loss": 3.9998,
+      "step": 1050112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.279801397177474e-05,
+      "loss": 4.0085,
+      "step": 1050624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278964440306795e-05,
+      "loss": 4.0093,
+      "step": 1051136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.278125845555743e-05,
+      "loss": 4.0005,
+      "step": 1051648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.277287250804691e-05,
+      "loss": 4.0038,
+      "step": 1052160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.276448656053639e-05,
+      "loss": 3.9994,
+      "step": 1052672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.275610061302587e-05,
+      "loss": 3.998,
+      "step": 1053184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.274773104431908e-05,
+      "loss": 3.9968,
+      "step": 1053696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.273934509680856e-05,
+      "loss": 3.9931,
+      "step": 1054208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2730975528101773e-05,
+      "loss": 4.0031,
+      "step": 1054720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.272258958059125e-05,
+      "loss": 3.9862,
+      "step": 1055232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.271420363308073e-05,
+      "loss": 4.0,
+      "step": 1055744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.270581768557021e-05,
+      "loss": 3.9963,
+      "step": 1056256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.269743173805969e-05,
+      "loss": 4.007,
+      "step": 1056768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268904579054917e-05,
+      "loss": 3.986,
+      "step": 1057280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.268067622184238e-05,
+      "loss": 4.0014,
+      "step": 1057792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.267229027433186e-05,
+      "loss": 3.9977,
+      "step": 1058304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.266390432682134e-05,
+      "loss": 4.0026,
+      "step": 1058816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.265551837931082e-05,
+      "loss": 4.0079,
+      "step": 1059328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2647132431800295e-05,
+      "loss": 4.0078,
+      "step": 1059840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2638746484289775e-05,
+      "loss": 3.9934,
+      "step": 1060352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.263036053677926e-05,
+      "loss": 4.0032,
+      "step": 1060864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.262197458926874e-05,
+      "loss": 3.9988,
+      "step": 1061376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.261358864175822e-05,
+      "loss": 3.9921,
+      "step": 1061888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.260521907305143e-05,
+      "loss": 4.0121,
+      "step": 1062400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.259684950434465e-05,
+      "loss": 4.005,
+      "step": 1062912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.258846355683413e-05,
+      "loss": 4.0122,
+      "step": 1063424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.25800776093236e-05,
+      "loss": 3.999,
+      "step": 1063936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.257169166181308e-05,
+      "loss": 4.0065,
+      "step": 1064448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.256330571430256e-05,
+      "loss": 3.9895,
+      "step": 1064960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.255491976679204e-05,
+      "loss": 3.9993,
+      "step": 1065472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.254653381928152e-05,
+      "loss": 3.9923,
+      "step": 1065984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2538147871771e-05,
+      "loss": 4.0057,
+      "step": 1066496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.252976192426048e-05,
+      "loss": 3.999,
+      "step": 1067008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2521392355553696e-05,
+      "loss": 3.9831,
+      "step": 1067520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.2513022786846905e-05,
+      "loss": 3.9984,
+      "step": 1068032
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.044682502746582,
+      "eval_runtime": 566.1028,
+      "eval_samples_per_second": 674.067,
+      "eval_steps_per_second": 21.065,
+      "step": 1068466
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2504636839336385e-05,
+      "loss": 4.0089,
+      "step": 1068544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2496250891825865e-05,
+      "loss": 4.0092,
+      "step": 1069056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2487864944315345e-05,
+      "loss": 4.0003,
+      "step": 1069568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2479478996804825e-05,
+      "loss": 3.995,
+      "step": 1070080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2471093049294305e-05,
+      "loss": 4.0038,
+      "step": 1070592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2462707101783784e-05,
+      "loss": 4.0022,
+      "step": 1071104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2454321154273264e-05,
+      "loss": 3.9835,
+      "step": 1071616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2445935206762744e-05,
+      "loss": 3.9948,
+      "step": 1072128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2437549259252224e-05,
+      "loss": 3.9951,
+      "step": 1072640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2429163311741704e-05,
+      "loss": 3.9944,
+      "step": 1073152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2420777364231184e-05,
+      "loss": 4.002,
+      "step": 1073664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2412391416720664e-05,
+      "loss": 4.0008,
+      "step": 1074176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2404005469210144e-05,
+      "loss": 3.9887,
+      "step": 1074688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2395619521699624e-05,
+      "loss": 3.9978,
+      "step": 1075200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2387233574189104e-05,
+      "loss": 3.9953,
+      "step": 1075712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2378847626678584e-05,
+      "loss": 3.9923,
+      "step": 1076224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2370461679168064e-05,
+      "loss": 3.9922,
+      "step": 1076736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.236209211046127e-05,
+      "loss": 3.995,
+      "step": 1077248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.235370616295075e-05,
+      "loss": 3.9961,
+      "step": 1077760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.234532021544023e-05,
+      "loss": 3.9954,
+      "step": 1078272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.233693426792971e-05,
+      "loss": 3.9995,
+      "step": 1078784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232854832041919e-05,
+      "loss": 4.0005,
+      "step": 1079296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.232016237290867e-05,
+      "loss": 3.9935,
+      "step": 1079808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.231179280420188e-05,
+      "loss": 4.0017,
+      "step": 1080320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.230340685669136e-05,
+      "loss": 3.9985,
+      "step": 1080832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.229502090918085e-05,
+      "loss": 4.0045,
+      "step": 1081344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.228663496167033e-05,
+      "loss": 3.9927,
+      "step": 1081856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.227824901415981e-05,
+      "loss": 3.9975,
+      "step": 1082368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.226986306664929e-05,
+      "loss": 3.9916,
+      "step": 1082880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.22614934979425e-05,
+      "loss": 4.0048,
+      "step": 1083392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.225310755043198e-05,
+      "loss": 3.9858,
+      "step": 1083904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.224472160292146e-05,
+      "loss": 3.992,
+      "step": 1084416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.223633565541094e-05,
+      "loss": 4.0065,
+      "step": 1084928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.222794970790042e-05,
+      "loss": 3.9985,
+      "step": 1085440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.22195637603899e-05,
+      "loss": 3.9859,
+      "step": 1085952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2211194191683106e-05,
+      "loss": 3.9974,
+      "step": 1086464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2202824622976316e-05,
+      "loss": 4.0008,
+      "step": 1086976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21944386754658e-05,
+      "loss": 4.0001,
+      "step": 1087488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.218605272795528e-05,
+      "loss": 3.9929,
+      "step": 1088000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.217766678044476e-05,
+      "loss": 3.9955,
+      "step": 1088512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216928083293424e-05,
+      "loss": 3.9881,
+      "step": 1089024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.216089488542372e-05,
+      "loss": 3.9779,
+      "step": 1089536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.21525089379132e-05,
+      "loss": 3.9942,
+      "step": 1090048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.214413936920641e-05,
+      "loss": 3.9945,
+      "step": 1090560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.213575342169589e-05,
+      "loss": 3.9903,
+      "step": 1091072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.212736747418537e-05,
+      "loss": 3.99,
+      "step": 1091584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2118981526674844e-05,
+      "loss": 3.9878,
+      "step": 1092096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2110595579164324e-05,
+      "loss": 4.0018,
+      "step": 1092608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2102209631653804e-05,
+      "loss": 3.9986,
+      "step": 1093120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2093823684143284e-05,
+      "loss": 3.9838,
+      "step": 1093632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2085437736632764e-05,
+      "loss": 3.9952,
+      "step": 1094144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.207706816792598e-05,
+      "loss": 3.9881,
+      "step": 1094656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206868222041546e-05,
+      "loss": 4.001,
+      "step": 1095168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.206029627290494e-05,
+      "loss": 3.9782,
+      "step": 1095680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.205191032539442e-05,
+      "loss": 3.9989,
+      "step": 1096192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.20435243778839e-05,
+      "loss": 3.9782,
+      "step": 1096704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.203515480917711e-05,
+      "loss": 3.9895,
+      "step": 1097216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.202676886166659e-05,
+      "loss": 4.0009,
+      "step": 1097728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201838291415607e-05,
+      "loss": 3.9912,
+      "step": 1098240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.201001334544928e-05,
+      "loss": 3.9901,
+      "step": 1098752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.200162739793876e-05,
+      "loss": 3.9774,
+      "step": 1099264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.199324145042824e-05,
+      "loss": 3.9875,
+      "step": 1099776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.198485550291772e-05,
+      "loss": 3.9938,
+      "step": 1100288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1976469555407204e-05,
+      "loss": 3.9846,
+      "step": 1100800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1968083607896684e-05,
+      "loss": 3.9946,
+      "step": 1101312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1959697660386164e-05,
+      "loss": 3.9906,
+      "step": 1101824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1951311712875644e-05,
+      "loss": 3.9763,
+      "step": 1102336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1942942144168853e-05,
+      "loss": 3.9786,
+      "step": 1102848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1934556196658333e-05,
+      "loss": 3.9973,
+      "step": 1103360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.192617024914781e-05,
+      "loss": 3.9733,
+      "step": 1103872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.191778430163729e-05,
+      "loss": 3.9853,
+      "step": 1104384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190939835412677e-05,
+      "loss": 4.0008,
+      "step": 1104896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.190101240661625e-05,
+      "loss": 3.9796,
+      "step": 1105408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.189262645910573e-05,
+      "loss": 3.9986,
+      "step": 1105920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1884240511595206e-05,
+      "loss": 4.0019,
+      "step": 1106432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.187587094288842e-05,
+      "loss": 3.9959,
+      "step": 1106944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.18674849953779e-05,
+      "loss": 3.997,
+      "step": 1107456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185909904786739e-05,
+      "loss": 3.9795,
+      "step": 1107968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.185071310035686e-05,
+      "loss": 3.9841,
+      "step": 1108480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.184234353165008e-05,
+      "loss": 3.9868,
+      "step": 1108992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.183395758413956e-05,
+      "loss": 3.9952,
+      "step": 1109504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.182557163662903e-05,
+      "loss": 3.992,
+      "step": 1110016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181718568911851e-05,
+      "loss": 3.9749,
+      "step": 1110528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180881612041173e-05,
+      "loss": 3.9823,
+      "step": 1111040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.180043017290121e-05,
+      "loss": 3.9801,
+      "step": 1111552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1792060604194416e-05,
+      "loss": 3.9849,
+      "step": 1112064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1783674656683896e-05,
+      "loss": 3.9902,
+      "step": 1112576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1775288709173376e-05,
+      "loss": 3.9982,
+      "step": 1113088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1766902761662856e-05,
+      "loss": 3.9931,
+      "step": 1113600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.1758516814152336e-05,
+      "loss": 3.9871,
+      "step": 1114112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1750130866641816e-05,
+      "loss": 3.9792,
+      "step": 1114624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.174176129793503e-05,
+      "loss": 3.978,
+      "step": 1115136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1733375350424505e-05,
+      "loss": 3.9861,
+      "step": 1115648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1724989402913985e-05,
+      "loss": 3.9863,
+      "step": 1116160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1716603455403465e-05,
+      "loss": 3.9773,
+      "step": 1116672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.170823388669668e-05,
+      "loss": 3.9814,
+      "step": 1117184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1699847939186154e-05,
+      "loss": 3.9897,
+      "step": 1117696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1691461991675634e-05,
+      "loss": 3.9938,
+      "step": 1118208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1683076044165114e-05,
+      "loss": 3.9909,
+      "step": 1118720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1674690096654594e-05,
+      "loss": 3.9919,
+      "step": 1119232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.166630414914408e-05,
+      "loss": 3.9938,
+      "step": 1119744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.165791820163356e-05,
+      "loss": 3.9862,
+      "step": 1120256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164953225412304e-05,
+      "loss": 3.985,
+      "step": 1120768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.164117906421998e-05,
+      "loss": 4.0029,
+      "step": 1121280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.163279311670946e-05,
+      "loss": 3.9823,
+      "step": 1121792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.162440716919894e-05,
+      "loss": 3.996,
+      "step": 1122304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.161602122168842e-05,
+      "loss": 3.9782,
+      "step": 1122816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.16076352741779e-05,
+      "loss": 3.9728,
+      "step": 1123328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159924932666738e-05,
+      "loss": 3.9863,
+      "step": 1123840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.159086337915686e-05,
+      "loss": 3.9891,
+      "step": 1124352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.158247743164634e-05,
+      "loss": 3.9885,
+      "step": 1124864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.157410786293955e-05,
+      "loss": 3.9636,
+      "step": 1125376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1565721915429034e-05,
+      "loss": 3.9789,
+      "step": 1125888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1557335967918514e-05,
+      "loss": 3.9895,
+      "step": 1126400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.154896639921172e-05,
+      "loss": 3.9942,
+      "step": 1126912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.15405804517012e-05,
+      "loss": 3.992,
+      "step": 1127424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.153219450419068e-05,
+      "loss": 3.9898,
+      "step": 1127936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.152380855668016e-05,
+      "loss": 3.989,
+      "step": 1128448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.151542260916964e-05,
+      "loss": 3.9879,
+      "step": 1128960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.150703666165912e-05,
+      "loss": 3.9875,
+      "step": 1129472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14986507141486e-05,
+      "loss": 3.9861,
+      "step": 1129984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.149026476663808e-05,
+      "loss": 3.9762,
+      "step": 1130496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.148187881912756e-05,
+      "loss": 3.9908,
+      "step": 1131008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.147350925042077e-05,
+      "loss": 3.973,
+      "step": 1131520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.146512330291025e-05,
+      "loss": 3.9906,
+      "step": 1132032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.145673735539973e-05,
+      "loss": 3.9826,
+      "step": 1132544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.144835140788921e-05,
+      "loss": 3.9932,
+      "step": 1133056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.14399654603787e-05,
+      "loss": 3.9722,
+      "step": 1133568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.143157951286817e-05,
+      "loss": 3.9903,
+      "step": 1134080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.142320994416139e-05,
+      "loss": 3.9857,
+      "step": 1134592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.141482399665087e-05,
+      "loss": 3.989,
+      "step": 1135104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.140643804914034e-05,
+      "loss": 3.9947,
+      "step": 1135616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1398068480433556e-05,
+      "loss": 3.9903,
+      "step": 1136128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1389682532923036e-05,
+      "loss": 3.983,
+      "step": 1136640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1381312964216246e-05,
+      "loss": 3.995,
+      "step": 1137152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1372927016705726e-05,
+      "loss": 3.9824,
+      "step": 1137664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1364541069195205e-05,
+      "loss": 3.9803,
+      "step": 1138176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1356155121684685e-05,
+      "loss": 3.9994,
+      "step": 1138688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1347769174174165e-05,
+      "loss": 3.9941,
+      "step": 1139200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1339383226663645e-05,
+      "loss": 4.0007,
+      "step": 1139712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1330997279153125e-05,
+      "loss": 3.9873,
+      "step": 1140224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1322611331642605e-05,
+      "loss": 3.9935,
+      "step": 1140736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1314225384132085e-05,
+      "loss": 3.98,
+      "step": 1141248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1305839436621565e-05,
+      "loss": 3.9852,
+      "step": 1141760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1297453489111045e-05,
+      "loss": 3.9793,
+      "step": 1142272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1289067541600525e-05,
+      "loss": 3.9885,
+      "step": 1142784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1280681594090005e-05,
+      "loss": 3.9864,
+      "step": 1143296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1272312025383214e-05,
+      "loss": 3.9759,
+      "step": 1143808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1263926077872694e-05,
+      "loss": 3.9843,
+      "step": 1144320
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.037414073944092,
+      "eval_runtime": 544.5169,
+      "eval_samples_per_second": 700.788,
+      "eval_steps_per_second": 21.9,
+      "step": 1144785
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1255540130362174e-05,
+      "loss": 3.9814,
+      "step": 1144832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1247154182851654e-05,
+      "loss": 3.9976,
+      "step": 1145344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1238768235341134e-05,
+      "loss": 3.9893,
+      "step": 1145856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.123038228783062e-05,
+      "loss": 3.9821,
+      "step": 1146368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.12219963403201e-05,
+      "loss": 3.9904,
+      "step": 1146880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.121361039280958e-05,
+      "loss": 3.9863,
+      "step": 1147392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1205224445299054e-05,
+      "loss": 3.9716,
+      "step": 1147904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.119685487659227e-05,
+      "loss": 3.9807,
+      "step": 1148416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118846892908175e-05,
+      "loss": 3.9865,
+      "step": 1148928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.118008298157123e-05,
+      "loss": 3.9793,
+      "step": 1149440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.117171341286444e-05,
+      "loss": 3.9873,
+      "step": 1149952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.116332746535392e-05,
+      "loss": 3.9918,
+      "step": 1150464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.11549415178434e-05,
+      "loss": 3.9728,
+      "step": 1150976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.114657194913661e-05,
+      "loss": 3.9851,
+      "step": 1151488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.113818600162609e-05,
+      "loss": 3.983,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1129800054115574e-05,
+      "loss": 3.9786,
+      "step": 1152512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1121414106605054e-05,
+      "loss": 3.9834,
+      "step": 1153024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.111302815909453e-05,
+      "loss": 3.981,
+      "step": 1153536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.110464221158401e-05,
+      "loss": 3.9846,
+      "step": 1154048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.109625626407349e-05,
+      "loss": 3.9815,
+      "step": 1154560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.108787031656297e-05,
+      "loss": 3.9862,
+      "step": 1155072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.107948436905245e-05,
+      "loss": 3.9902,
+      "step": 1155584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1071114800345656e-05,
+      "loss": 3.9803,
+      "step": 1156096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1062728852835136e-05,
+      "loss": 3.9934,
+      "step": 1156608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1054342905324616e-05,
+      "loss": 3.9874,
+      "step": 1157120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.1045973336617825e-05,
+      "loss": 3.9933,
+      "step": 1157632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.103758738910731e-05,
+      "loss": 3.9781,
+      "step": 1158144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102920144159679e-05,
+      "loss": 3.9891,
+      "step": 1158656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.102081549408627e-05,
+      "loss": 3.9791,
+      "step": 1159168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.101242954657575e-05,
+      "loss": 3.991,
+      "step": 1159680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.100404359906523e-05,
+      "loss": 3.9727,
+      "step": 1160192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.099565765155471e-05,
+      "loss": 3.9798,
+      "step": 1160704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.098727170404419e-05,
+      "loss": 3.9934,
+      "step": 1161216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.09789021353374e-05,
+      "loss": 3.9886,
+      "step": 1161728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.097051618782688e-05,
+      "loss": 3.9756,
+      "step": 1162240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.096213024031636e-05,
+      "loss": 3.9823,
+      "step": 1162752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.095376067160957e-05,
+      "loss": 3.9859,
+      "step": 1163264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.094537472409905e-05,
+      "loss": 3.9877,
+      "step": 1163776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.093698877658853e-05,
+      "loss": 3.9842,
+      "step": 1164288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092860282907801e-05,
+      "loss": 3.985,
+      "step": 1164800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.092021688156749e-05,
+      "loss": 3.9728,
+      "step": 1165312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0911830934056977e-05,
+      "loss": 3.9732,
+      "step": 1165824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0903444986546456e-05,
+      "loss": 3.9796,
+      "step": 1166336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0895059039035936e-05,
+      "loss": 3.9853,
+      "step": 1166848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0886689470329146e-05,
+      "loss": 3.9796,
+      "step": 1167360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0878303522818625e-05,
+      "loss": 3.9755,
+      "step": 1167872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0869917575308105e-05,
+      "loss": 3.9764,
+      "step": 1168384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0861531627797585e-05,
+      "loss": 3.9842,
+      "step": 1168896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0853145680287065e-05,
+      "loss": 3.9935,
+      "step": 1169408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0844776111580274e-05,
+      "loss": 3.9743,
+      "step": 1169920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0836390164069754e-05,
+      "loss": 3.9793,
+      "step": 1170432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0828004216559234e-05,
+      "loss": 3.9771,
+      "step": 1170944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0819618269048714e-05,
+      "loss": 3.9893,
+      "step": 1171456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0811232321538194e-05,
+      "loss": 3.9704,
+      "step": 1171968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0802846374027674e-05,
+      "loss": 3.9866,
+      "step": 1172480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0794460426517154e-05,
+      "loss": 3.9695,
+      "step": 1172992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0786074479006634e-05,
+      "loss": 3.9771,
+      "step": 1173504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.077770491029984e-05,
+      "loss": 3.9861,
+      "step": 1174016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076933534159306e-05,
+      "loss": 3.9815,
+      "step": 1174528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.076094939408254e-05,
+      "loss": 3.9755,
+      "step": 1175040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.075256344657201e-05,
+      "loss": 3.9653,
+      "step": 1175552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.074417749906149e-05,
+      "loss": 3.9724,
+      "step": 1176064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.073579155155097e-05,
+      "loss": 3.9828,
+      "step": 1176576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.072740560404045e-05,
+      "loss": 3.9737,
+      "step": 1177088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071901965652993e-05,
+      "loss": 3.9757,
+      "step": 1177600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.071063370901941e-05,
+      "loss": 3.9783,
+      "step": 1178112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.070226414031263e-05,
+      "loss": 3.9715,
+      "step": 1178624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.069387819280211e-05,
+      "loss": 3.9641,
+      "step": 1179136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.068550862409532e-05,
+      "loss": 3.9866,
+      "step": 1179648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.06771226765848e-05,
+      "loss": 3.9653,
+      "step": 1180160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066873672907428e-05,
+      "loss": 3.9682,
+      "step": 1180672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.066035078156376e-05,
+      "loss": 3.9891,
+      "step": 1181184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.065196483405324e-05,
+      "loss": 3.9714,
+      "step": 1181696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.064357888654272e-05,
+      "loss": 3.9856,
+      "step": 1182208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0635209317835926e-05,
+      "loss": 3.9931,
+      "step": 1182720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0626823370325406e-05,
+      "loss": 3.9801,
+      "step": 1183232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0618437422814886e-05,
+      "loss": 3.9867,
+      "step": 1183744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0610051475304366e-05,
+      "loss": 3.9739,
+      "step": 1184256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.060166552779385e-05,
+      "loss": 3.9688,
+      "step": 1184768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.059327958028333e-05,
+      "loss": 3.9781,
+      "step": 1185280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.058489363277281e-05,
+      "loss": 3.9827,
+      "step": 1185792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.057650768526229e-05,
+      "loss": 3.9827,
+      "step": 1186304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.056812173775177e-05,
+      "loss": 3.9633,
+      "step": 1186816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055973579024125e-05,
+      "loss": 3.971,
+      "step": 1187328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.055136622153446e-05,
+      "loss": 3.9685,
+      "step": 1187840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.054298027402394e-05,
+      "loss": 3.9737,
+      "step": 1188352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.053459432651342e-05,
+      "loss": 3.9754,
+      "step": 1188864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.05262083790029e-05,
+      "loss": 3.989,
+      "step": 1189376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.051783881029611e-05,
+      "loss": 3.9812,
+      "step": 1189888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.0509452862785594e-05,
+      "loss": 3.9757,
+      "step": 1190400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0501066915275074e-05,
+      "loss": 3.9681,
+      "step": 1190912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0492697346568283e-05,
+      "loss": 3.9688,
+      "step": 1191424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0484311399057763e-05,
+      "loss": 3.9728,
+      "step": 1191936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0475925451547243e-05,
+      "loss": 3.9749,
+      "step": 1192448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0467539504036723e-05,
+      "loss": 3.9671,
+      "step": 1192960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.04591535565262e-05,
+      "loss": 3.9674,
+      "step": 1193472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.045076760901568e-05,
+      "loss": 3.9804,
+      "step": 1193984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.044238166150516e-05,
+      "loss": 3.981,
+      "step": 1194496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.043399571399464e-05,
+      "loss": 3.9812,
+      "step": 1195008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.042562614528785e-05,
+      "loss": 3.9753,
+      "step": 1195520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0417256576581067e-05,
+      "loss": 3.9873,
+      "step": 1196032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0408870629070547e-05,
+      "loss": 3.9752,
+      "step": 1196544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.040048468156002e-05,
+      "loss": 3.9738,
+      "step": 1197056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.03920987340495e-05,
+      "loss": 3.991,
+      "step": 1197568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.038371278653898e-05,
+      "loss": 3.9748,
+      "step": 1198080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0375343217832196e-05,
+      "loss": 3.9863,
+      "step": 1198592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0366957270321673e-05,
+      "loss": 3.9689,
+      "step": 1199104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0358571322811153e-05,
+      "loss": 3.9605,
+      "step": 1199616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0350185375300633e-05,
+      "loss": 3.9722,
+      "step": 1200128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0341799427790113e-05,
+      "loss": 3.977,
+      "step": 1200640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0333413480279593e-05,
+      "loss": 3.9799,
+      "step": 1201152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0325027532769073e-05,
+      "loss": 3.9571,
+      "step": 1201664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0316641585258553e-05,
+      "loss": 3.9641,
+      "step": 1202176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0308255637748036e-05,
+      "loss": 3.9747,
+      "step": 1202688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0299886069041245e-05,
+      "loss": 3.985,
+      "step": 1203200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0291500121530725e-05,
+      "loss": 3.9813,
+      "step": 1203712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0283114174020205e-05,
+      "loss": 3.9756,
+      "step": 1204224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0274728226509685e-05,
+      "loss": 3.98,
+      "step": 1204736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0266342278999165e-05,
+      "loss": 3.9808,
+      "step": 1205248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0257956331488645e-05,
+      "loss": 3.9745,
+      "step": 1205760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0249570383978125e-05,
+      "loss": 3.9722,
+      "step": 1206272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0241184436467608e-05,
+      "loss": 3.963,
+      "step": 1206784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0232814867760817e-05,
+      "loss": 3.9806,
+      "step": 1207296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0224428920250297e-05,
+      "loss": 3.9612,
+      "step": 1207808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0216042972739777e-05,
+      "loss": 3.9813,
+      "step": 1208320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.020767340403299e-05,
+      "loss": 3.9713,
+      "step": 1208832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019928745652247e-05,
+      "loss": 3.9809,
+      "step": 1209344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.019090150901195e-05,
+      "loss": 3.9615,
+      "step": 1209856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.018251556150143e-05,
+      "loss": 3.9807,
+      "step": 1210368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.017412961399091e-05,
+      "loss": 3.975,
+      "step": 1210880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.016576004528412e-05,
+      "loss": 3.9782,
+      "step": 1211392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.01573740977736e-05,
+      "loss": 3.9834,
+      "step": 1211904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.014898815026308e-05,
+      "loss": 3.979,
+      "step": 1212416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0140602202752562e-05,
+      "loss": 3.9748,
+      "step": 1212928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0132216255242035e-05,
+      "loss": 3.9811,
+      "step": 1213440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0123830307731515e-05,
+      "loss": 3.9688,
+      "step": 1213952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0115444360220995e-05,
+      "loss": 3.9735,
+      "step": 1214464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.010707479151421e-05,
+      "loss": 3.9855,
+      "step": 1214976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0098688844003687e-05,
+      "loss": 3.9875,
+      "step": 1215488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0090302896493167e-05,
+      "loss": 3.9841,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0081916948982647e-05,
+      "loss": 3.9792,
+      "step": 1216512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0073547380275856e-05,
+      "loss": 3.978,
+      "step": 1217024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0065177811569072e-05,
+      "loss": 3.9712,
+      "step": 1217536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0056791864058552e-05,
+      "loss": 3.9755,
+      "step": 1218048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.0048405916548032e-05,
+      "loss": 3.9686,
+      "step": 1218560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.004001996903751e-05,
+      "loss": 3.9756,
+      "step": 1219072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.003163402152699e-05,
+      "loss": 3.9779,
+      "step": 1219584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.002324807401647e-05,
+      "loss": 3.9629,
+      "step": 1220096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.001486212650595e-05,
+      "loss": 3.9745,
+      "step": 1220608
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.031482219696045,
+      "eval_runtime": 551.3537,
+      "eval_samples_per_second": 692.098,
+      "eval_steps_per_second": 21.629,
+      "step": 1221104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.000647617899543e-05,
+      "loss": 3.977,
+      "step": 1221120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.999809023148491e-05,
+      "loss": 3.9725,
+      "step": 1221632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9989704283974392e-05,
+      "loss": 3.97,
+      "step": 1222144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.998131833646387e-05,
+      "loss": 3.9831,
+      "step": 1222656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.997293238895335e-05,
+      "loss": 3.9793,
+      "step": 1223168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.996454644144283e-05,
+      "loss": 3.9876,
+      "step": 1223680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.995616049393231e-05,
+      "loss": 3.9745,
+      "step": 1224192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.994777454642179e-05,
+      "loss": 3.9693,
+      "step": 1224704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.993938859891127e-05,
+      "loss": 3.968,
+      "step": 1225216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9931002651400748e-05,
+      "loss": 3.9804,
+      "step": 1225728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9922616703890228e-05,
+      "loss": 3.9743,
+      "step": 1226240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9914230756379708e-05,
+      "loss": 3.9769,
+      "step": 1226752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9905844808869188e-05,
+      "loss": 3.9795,
+      "step": 1227264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9897475240162397e-05,
+      "loss": 3.9673,
+      "step": 1227776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9889089292651877e-05,
+      "loss": 3.9689,
+      "step": 1228288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.988070334514136e-05,
+      "loss": 3.9664,
+      "step": 1228800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.987231739763084e-05,
+      "loss": 3.9645,
+      "step": 1229312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.986393145012032e-05,
+      "loss": 3.9765,
+      "step": 1229824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.985556188141353e-05,
+      "loss": 3.9704,
+      "step": 1230336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.984717593390301e-05,
+      "loss": 3.9769,
+      "step": 1230848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983878998639249e-05,
+      "loss": 3.9896,
+      "step": 1231360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.983040403888197e-05,
+      "loss": 3.9713,
+      "step": 1231872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9822018091371453e-05,
+      "loss": 3.9772,
+      "step": 1232384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9813632143860932e-05,
+      "loss": 3.9731,
+      "step": 1232896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9805246196350412e-05,
+      "loss": 3.9876,
+      "step": 1233408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9796860248839892e-05,
+      "loss": 3.9712,
+      "step": 1233920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.9788474301329372e-05,
+      "loss": 3.9746,
+      "step": 1234432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978010473262258e-05,
+      "loss": 3.9649,
+      "step": 1234944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.977171878511206e-05,
+      "loss": 3.9745,
+      "step": 1235456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.976333283760154e-05,
+      "loss": 3.9681,
+      "step": 1235968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9754946890091025e-05,
+      "loss": 3.9723,
+      "step": 1236480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9746577321384234e-05,
+      "loss": 3.9796,
+      "step": 1236992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9738191373873714e-05,
+      "loss": 3.9755,
+      "step": 1237504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9729805426363194e-05,
+      "loss": 3.9782,
+      "step": 1238016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9721419478852674e-05,
+      "loss": 3.9757,
+      "step": 1238528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9713049910145886e-05,
+      "loss": 3.9731,
+      "step": 1239040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9704663962635366e-05,
+      "loss": 3.9808,
+      "step": 1239552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9696278015124846e-05,
+      "loss": 3.9623,
+      "step": 1240064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9687892067614326e-05,
+      "loss": 3.9741,
+      "step": 1240576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9679506120103806e-05,
+      "loss": 3.9553,
+      "step": 1241088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9671120172593286e-05,
+      "loss": 3.9676,
+      "step": 1241600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9662734225082762e-05,
+      "loss": 3.9638,
+      "step": 1242112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9654348277572242e-05,
+      "loss": 3.9831,
+      "step": 1242624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9645962330061722e-05,
+      "loss": 3.969,
+      "step": 1243136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.963759276135493e-05,
+      "loss": 3.9665,
+      "step": 1243648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962920681384441e-05,
+      "loss": 3.9723,
+      "step": 1244160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.962082086633389e-05,
+      "loss": 3.9728,
+      "step": 1244672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.961243491882337e-05,
+      "loss": 3.9676,
+      "step": 1245184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9604048971312855e-05,
+      "loss": 3.9542,
+      "step": 1245696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9595679402606064e-05,
+      "loss": 3.9651,
+      "step": 1246208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9587293455095544e-05,
+      "loss": 3.9857,
+      "step": 1246720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9578907507585024e-05,
+      "loss": 3.9817,
+      "step": 1247232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9570521560074504e-05,
+      "loss": 3.968,
+      "step": 1247744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9562135612563984e-05,
+      "loss": 3.9673,
+      "step": 1248256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9553749665053464e-05,
+      "loss": 3.9788,
+      "step": 1248768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9545380096346676e-05,
+      "loss": 3.9561,
+      "step": 1249280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9536994148836156e-05,
+      "loss": 3.982,
+      "step": 1249792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9528608201325636e-05,
+      "loss": 3.9544,
+      "step": 1250304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9520222253815116e-05,
+      "loss": 3.983,
+      "step": 1250816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9511836306304596e-05,
+      "loss": 3.9785,
+      "step": 1251328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.950346673759781e-05,
+      "loss": 3.9557,
+      "step": 1251840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.949508079008729e-05,
+      "loss": 3.9631,
+      "step": 1252352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9486694842576768e-05,
+      "loss": 3.9765,
+      "step": 1252864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9478308895066248e-05,
+      "loss": 3.947,
+      "step": 1253376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9469955705163186e-05,
+      "loss": 3.9597,
+      "step": 1253888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.946156975765267e-05,
+      "loss": 3.9784,
+      "step": 1254400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.945318381014215e-05,
+      "loss": 3.9703,
+      "step": 1254912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.944479786263163e-05,
+      "loss": 3.9678,
+      "step": 1255424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.943641191512111e-05,
+      "loss": 3.9507,
+      "step": 1255936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.942802596761059e-05,
+      "loss": 3.9635,
+      "step": 1256448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941964002010007e-05,
+      "loss": 3.9551,
+      "step": 1256960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.941125407258955e-05,
+      "loss": 3.9795,
+      "step": 1257472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.940286812507903e-05,
+      "loss": 3.9571,
+      "step": 1257984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9394498556372242e-05,
+      "loss": 3.9728,
+      "step": 1258496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9386112608861722e-05,
+      "loss": 3.9713,
+      "step": 1259008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9377726661351202e-05,
+      "loss": 3.9762,
+      "step": 1259520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9369340713840682e-05,
+      "loss": 3.962,
+      "step": 1260032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.936097114513389e-05,
+      "loss": 3.9685,
+      "step": 1260544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.935258519762337e-05,
+      "loss": 3.9732,
+      "step": 1261056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9344199250112854e-05,
+      "loss": 3.9693,
+      "step": 1261568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9335813302602334e-05,
+      "loss": 3.9661,
+      "step": 1262080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9327427355091814e-05,
+      "loss": 3.9648,
+      "step": 1262592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9319041407581294e-05,
+      "loss": 3.9721,
+      "step": 1263104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9310655460070767e-05,
+      "loss": 3.97,
+      "step": 1263616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9302269512560247e-05,
+      "loss": 3.9541,
+      "step": 1264128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9293899943853463e-05,
+      "loss": 3.9665,
+      "step": 1264640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9285513996342946e-05,
+      "loss": 3.9628,
+      "step": 1265152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.927712804883242e-05,
+      "loss": 3.97,
+      "step": 1265664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.92687421013219e-05,
+      "loss": 3.9655,
+      "step": 1266176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9260372532615115e-05,
+      "loss": 3.9666,
+      "step": 1266688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9251986585104592e-05,
+      "loss": 3.9603,
+      "step": 1267200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9243600637594072e-05,
+      "loss": 3.9545,
+      "step": 1267712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9235214690083552e-05,
+      "loss": 3.9599,
+      "step": 1268224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9226828742573032e-05,
+      "loss": 3.9733,
+      "step": 1268736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9218442795062512e-05,
+      "loss": 3.9658,
+      "step": 1269248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9210056847551992e-05,
+      "loss": 3.9728,
+      "step": 1269760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9201670900041472e-05,
+      "loss": 3.967,
+      "step": 1270272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9193301331334684e-05,
+      "loss": 3.9588,
+      "step": 1270784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9184915383824164e-05,
+      "loss": 3.9614,
+      "step": 1271296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9176545815117373e-05,
+      "loss": 3.974,
+      "step": 1271808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9168159867606853e-05,
+      "loss": 3.9664,
+      "step": 1272320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9159773920096333e-05,
+      "loss": 3.9743,
+      "step": 1272832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9151387972585813e-05,
+      "loss": 3.9616,
+      "step": 1273344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9143002025075293e-05,
+      "loss": 3.9595,
+      "step": 1273856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9134616077564773e-05,
+      "loss": 3.9799,
+      "step": 1274368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9126246508857986e-05,
+      "loss": 3.9541,
+      "step": 1274880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9117860561347466e-05,
+      "loss": 3.9595,
+      "step": 1275392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9109474613836945e-05,
+      "loss": 3.9668,
+      "step": 1275904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9101088666326425e-05,
+      "loss": 3.9624,
+      "step": 1276416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9092719097619635e-05,
+      "loss": 3.9579,
+      "step": 1276928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9084333150109118e-05,
+      "loss": 3.9605,
+      "step": 1277440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9075947202598598e-05,
+      "loss": 3.9584,
+      "step": 1277952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9067561255088078e-05,
+      "loss": 3.9671,
+      "step": 1278464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9059175307577558e-05,
+      "loss": 3.963,
+      "step": 1278976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.9050805738870767e-05,
+      "loss": 3.9626,
+      "step": 1279488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.904243617016398e-05,
+      "loss": 3.9721,
+      "step": 1280000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.903405022265346e-05,
+      "loss": 3.9826,
+      "step": 1280512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.902566427514294e-05,
+      "loss": 3.9692,
+      "step": 1281024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.901727832763242e-05,
+      "loss": 3.9604,
+      "step": 1281536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.90088923801219e-05,
+      "loss": 3.9597,
+      "step": 1282048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.900050643261138e-05,
+      "loss": 3.962,
+      "step": 1282560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.899212048510086e-05,
+      "loss": 3.9644,
+      "step": 1283072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.898373453759034e-05,
+      "loss": 3.9651,
+      "step": 1283584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.897534859007982e-05,
+      "loss": 3.9621,
+      "step": 1284096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8966962642569302e-05,
+      "loss": 3.9681,
+      "step": 1284608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8958576695058782e-05,
+      "loss": 3.9639,
+      "step": 1285120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8950190747548255e-05,
+      "loss": 3.9656,
+      "step": 1285632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.894182117884147e-05,
+      "loss": 3.9612,
+      "step": 1286144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.893343523133095e-05,
+      "loss": 3.9617,
+      "step": 1286656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8925065662624164e-05,
+      "loss": 3.959,
+      "step": 1287168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8916679715113644e-05,
+      "loss": 3.9737,
+      "step": 1287680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8908293767603124e-05,
+      "loss": 3.9746,
+      "step": 1288192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8899907820092604e-05,
+      "loss": 3.9718,
+      "step": 1288704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8891521872582077e-05,
+      "loss": 3.9543,
+      "step": 1289216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8883135925071557e-05,
+      "loss": 3.9692,
+      "step": 1289728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.887474997756104e-05,
+      "loss": 3.9654,
+      "step": 1290240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.886636403005052e-05,
+      "loss": 3.9693,
+      "step": 1290752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.885799446134373e-05,
+      "loss": 3.9652,
+      "step": 1291264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8849624892636945e-05,
+      "loss": 3.9795,
+      "step": 1291776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8841238945126425e-05,
+      "loss": 3.9624,
+      "step": 1292288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.88328529976159e-05,
+      "loss": 3.982,
+      "step": 1292800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.882446705010538e-05,
+      "loss": 3.962,
+      "step": 1293312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.881608110259486e-05,
+      "loss": 3.9735,
+      "step": 1293824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.880769515508434e-05,
+      "loss": 3.9637,
+      "step": 1294336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879932558637755e-05,
+      "loss": 3.958,
+      "step": 1294848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.879093963886703e-05,
+      "loss": 3.9671,
+      "step": 1295360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.878255369135651e-05,
+      "loss": 3.9526,
+      "step": 1295872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8774167743845994e-05,
+      "loss": 3.9638,
+      "step": 1296384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8765798175139203e-05,
+      "loss": 3.9672,
+      "step": 1296896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8757412227628683e-05,
+      "loss": 3.9639,
+      "step": 1297408
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.024557590484619,
+      "eval_runtime": 294.8174,
+      "eval_samples_per_second": 1294.33,
+      "eval_steps_per_second": 40.449,
+      "step": 1297424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8749026280118163e-05,
+      "loss": 3.9633,
+      "step": 1297920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8740640332607643e-05,
+      "loss": 3.9616,
+      "step": 1298432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8732254385097123e-05,
+      "loss": 3.9702,
+      "step": 1298944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8723868437586603e-05,
+      "loss": 3.9662,
+      "step": 1299456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8715482490076086e-05,
+      "loss": 3.9799,
+      "step": 1299968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8707096542565566e-05,
+      "loss": 3.9608,
+      "step": 1300480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8698710595055046e-05,
+      "loss": 3.9593,
+      "step": 1300992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8690324647544526e-05,
+      "loss": 3.9577,
+      "step": 1301504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8681938700034006e-05,
+      "loss": 3.9692,
+      "step": 1302016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8673552752523486e-05,
+      "loss": 3.9631,
+      "step": 1302528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8665183183816695e-05,
+      "loss": 3.9668,
+      "step": 1303040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8656797236306178e-05,
+      "loss": 3.9663,
+      "step": 1303552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8648427667599387e-05,
+      "loss": 3.9525,
+      "step": 1304064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8640041720088867e-05,
+      "loss": 3.964,
+      "step": 1304576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8631655772578347e-05,
+      "loss": 3.9534,
+      "step": 1305088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8623269825067827e-05,
+      "loss": 3.9576,
+      "step": 1305600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8614883877557307e-05,
+      "loss": 3.961,
+      "step": 1306112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8606497930046787e-05,
+      "loss": 3.9592,
+      "step": 1306624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8598111982536264e-05,
+      "loss": 3.9626,
+      "step": 1307136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858974241382948e-05,
+      "loss": 3.9838,
+      "step": 1307648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.858135646631896e-05,
+      "loss": 3.9598,
+      "step": 1308160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.857297051880844e-05,
+      "loss": 3.9647,
+      "step": 1308672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8564584571297916e-05,
+      "loss": 3.9637,
+      "step": 1309184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.855621500259113e-05,
+      "loss": 3.9714,
+      "step": 1309696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8547829055080612e-05,
+      "loss": 3.9619,
+      "step": 1310208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8539443107570085e-05,
+      "loss": 3.9637,
+      "step": 1310720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8531057160059565e-05,
+      "loss": 3.9545,
+      "step": 1311232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8522671212549045e-05,
+      "loss": 3.964,
+      "step": 1311744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.8514285265038525e-05,
+      "loss": 3.9534,
+      "step": 1312256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8505899317528005e-05,
+      "loss": 3.9644,
+      "step": 1312768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8497513370017488e-05,
+      "loss": 3.9676,
+      "step": 1313280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8489160180114433e-05,
+      "loss": 3.9635,
+      "step": 1313792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8480774232603913e-05,
+      "loss": 3.971,
+      "step": 1314304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8472388285093386e-05,
+      "loss": 3.9641,
+      "step": 1314816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8464035095190335e-05,
+      "loss": 3.9658,
+      "step": 1315328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8455649147679815e-05,
+      "loss": 3.9678,
+      "step": 1315840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8447263200169295e-05,
+      "loss": 3.9572,
+      "step": 1316352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8438877252658775e-05,
+      "loss": 3.9581,
+      "step": 1316864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8430491305148255e-05,
+      "loss": 3.9453,
+      "step": 1317376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8422105357637735e-05,
+      "loss": 3.9584,
+      "step": 1317888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.841371941012721e-05,
+      "loss": 3.9576,
+      "step": 1318400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.840533346261669e-05,
+      "loss": 3.9703,
+      "step": 1318912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.839694751510617e-05,
+      "loss": 3.9583,
+      "step": 1319424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838856156759565e-05,
+      "loss": 3.9601,
+      "step": 1319936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.838017562008513e-05,
+      "loss": 3.9612,
+      "step": 1320448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.837178967257461e-05,
+      "loss": 3.9603,
+      "step": 1320960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.836340372506409e-05,
+      "loss": 3.958,
+      "step": 1321472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.835501777755357e-05,
+      "loss": 3.9474,
+      "step": 1321984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8346648208846783e-05,
+      "loss": 3.95,
+      "step": 1322496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8338262261336263e-05,
+      "loss": 3.9767,
+      "step": 1323008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8329876313825743e-05,
+      "loss": 3.9721,
+      "step": 1323520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8321490366315223e-05,
+      "loss": 3.9536,
+      "step": 1324032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8313104418804703e-05,
+      "loss": 3.9564,
+      "step": 1324544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8304718471294183e-05,
+      "loss": 3.9685,
+      "step": 1325056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8296348902587396e-05,
+      "loss": 3.9474,
+      "step": 1325568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8287962955076875e-05,
+      "loss": 3.969,
+      "step": 1326080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8279593386370085e-05,
+      "loss": 3.9459,
+      "step": 1326592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8271207438859565e-05,
+      "loss": 3.9763,
+      "step": 1327104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8262821491349044e-05,
+      "loss": 3.9648,
+      "step": 1327616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8254435543838524e-05,
+      "loss": 3.9525,
+      "step": 1328128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8246049596328004e-05,
+      "loss": 3.9483,
+      "step": 1328640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8237663648817488e-05,
+      "loss": 3.967,
+      "step": 1329152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8229277701306968e-05,
+      "loss": 3.9387,
+      "step": 1329664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8220891753796448e-05,
+      "loss": 3.9507,
+      "step": 1330176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.821250580628592e-05,
+      "loss": 3.9675,
+      "step": 1330688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.82041198587754e-05,
+      "loss": 3.9614,
+      "step": 1331200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.819573391126488e-05,
+      "loss": 3.9552,
+      "step": 1331712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8187364342558097e-05,
+      "loss": 3.9427,
+      "step": 1332224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817899477385131e-05,
+      "loss": 3.9518,
+      "step": 1332736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.817060882634079e-05,
+      "loss": 3.9498,
+      "step": 1333248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.816222287883027e-05,
+      "loss": 3.9633,
+      "step": 1333760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8153836931319742e-05,
+      "loss": 3.9527,
+      "step": 1334272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8145450983809226e-05,
+      "loss": 3.9596,
+      "step": 1334784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8137065036298705e-05,
+      "loss": 3.964,
+      "step": 1335296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8128679088788185e-05,
+      "loss": 3.9685,
+      "step": 1335808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8120309520081395e-05,
+      "loss": 3.9506,
+      "step": 1336320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.811193995137461e-05,
+      "loss": 3.9603,
+      "step": 1336832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.810355400386409e-05,
+      "loss": 3.9605,
+      "step": 1337344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.809516805635357e-05,
+      "loss": 3.9595,
+      "step": 1337856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8086782108843047e-05,
+      "loss": 3.956,
+      "step": 1338368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8078396161332527e-05,
+      "loss": 3.954,
+      "step": 1338880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8070010213822007e-05,
+      "loss": 3.9618,
+      "step": 1339392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8061624266311487e-05,
+      "loss": 3.9591,
+      "step": 1339904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8053238318800967e-05,
+      "loss": 3.9455,
+      "step": 1340416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8044852371290447e-05,
+      "loss": 3.9519,
+      "step": 1340928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.8036466423779927e-05,
+      "loss": 3.9561,
+      "step": 1341440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.802809685507314e-05,
+      "loss": 3.9618,
+      "step": 1341952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.801971090756262e-05,
+      "loss": 3.9594,
+      "step": 1342464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.80113249600521e-05,
+      "loss": 3.9539,
+      "step": 1342976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.800293901254158e-05,
+      "loss": 3.9482,
+      "step": 1343488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7994569443834788e-05,
+      "loss": 3.9466,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.798618349632427e-05,
+      "loss": 3.9484,
+      "step": 1344512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.797779754881375e-05,
+      "loss": 3.9617,
+      "step": 1345024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796941160130323e-05,
+      "loss": 3.9578,
+      "step": 1345536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.796102565379271e-05,
+      "loss": 3.9628,
+      "step": 1346048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.795263970628219e-05,
+      "loss": 3.9604,
+      "step": 1346560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.794425375877167e-05,
+      "loss": 3.9468,
+      "step": 1347072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.793586781126115e-05,
+      "loss": 3.9529,
+      "step": 1347584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.792749824255436e-05,
+      "loss": 3.9602,
+      "step": 1348096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7919112295043844e-05,
+      "loss": 3.9616,
+      "step": 1348608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7910726347533324e-05,
+      "loss": 3.9607,
+      "step": 1349120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7902340400022803e-05,
+      "loss": 3.9551,
+      "step": 1349632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7893954452512283e-05,
+      "loss": 3.9459,
+      "step": 1350144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7885568505001757e-05,
+      "loss": 3.9706,
+      "step": 1350656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7877198936294972e-05,
+      "loss": 3.9454,
+      "step": 1351168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7868812988784452e-05,
+      "loss": 3.9545,
+      "step": 1351680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7860427041273936e-05,
+      "loss": 3.9528,
+      "step": 1352192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.785204109376341e-05,
+      "loss": 3.9532,
+      "step": 1352704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.784365514625289e-05,
+      "loss": 3.9451,
+      "step": 1353216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7835285577546105e-05,
+      "loss": 3.9527,
+      "step": 1353728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.782689963003558e-05,
+      "loss": 3.9468,
+      "step": 1354240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.781851368252506e-05,
+      "loss": 3.9551,
+      "step": 1354752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.781012773501454e-05,
+      "loss": 3.9525,
+      "step": 1355264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7801758166307757e-05,
+      "loss": 3.948,
+      "step": 1355776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.779337221879723e-05,
+      "loss": 3.9672,
+      "step": 1356288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.778498627128671e-05,
+      "loss": 3.9712,
+      "step": 1356800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.777660032377619e-05,
+      "loss": 3.9587,
+      "step": 1357312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7768230755069406e-05,
+      "loss": 3.9538,
+      "step": 1357824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7759844807558883e-05,
+      "loss": 3.9496,
+      "step": 1358336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7751458860048363e-05,
+      "loss": 3.952,
+      "step": 1358848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7743072912537843e-05,
+      "loss": 3.9542,
+      "step": 1359360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7734686965027323e-05,
+      "loss": 3.961,
+      "step": 1359872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7726317396320535e-05,
+      "loss": 3.9473,
+      "step": 1360384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7717931448810015e-05,
+      "loss": 3.9558,
+      "step": 1360896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7709545501299495e-05,
+      "loss": 3.9562,
+      "step": 1361408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7701159553788975e-05,
+      "loss": 3.9605,
+      "step": 1361920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7692789985082184e-05,
+      "loss": 3.9478,
+      "step": 1362432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7684404037571664e-05,
+      "loss": 3.9524,
+      "step": 1362944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7676018090061144e-05,
+      "loss": 3.9524,
+      "step": 1363456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7667632142550627e-05,
+      "loss": 3.9591,
+      "step": 1363968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7659246195040107e-05,
+      "loss": 3.9654,
+      "step": 1364480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7650860247529587e-05,
+      "loss": 3.9659,
+      "step": 1364992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7642507057626525e-05,
+      "loss": 3.9456,
+      "step": 1365504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7634121110116005e-05,
+      "loss": 3.9565,
+      "step": 1366016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.762573516260549e-05,
+      "loss": 3.9593,
+      "step": 1366528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.761734921509497e-05,
+      "loss": 3.9593,
+      "step": 1367040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760896326758445e-05,
+      "loss": 3.9553,
+      "step": 1367552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.760057732007393e-05,
+      "loss": 3.9654,
+      "step": 1368064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.759219137256341e-05,
+      "loss": 3.9544,
+      "step": 1368576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.758380542505289e-05,
+      "loss": 3.973,
+      "step": 1369088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.757541947754237e-05,
+      "loss": 3.9495,
+      "step": 1369600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.756703353003185e-05,
+      "loss": 3.966,
+      "step": 1370112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755864758252133e-05,
+      "loss": 3.9532,
+      "step": 1370624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.755026163501081e-05,
+      "loss": 3.9504,
+      "step": 1371136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.754187568750029e-05,
+      "loss": 3.9579,
+      "step": 1371648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.753352249759723e-05,
+      "loss": 3.9474,
+      "step": 1372160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.752513655008671e-05,
+      "loss": 3.9526,
+      "step": 1372672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7516766981379922e-05,
+      "loss": 3.9551,
+      "step": 1373184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.7508381033869402e-05,
+      "loss": 3.9518,
+      "step": 1373696
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.019606590270996,
+      "eval_runtime": 291.0313,
+      "eval_samples_per_second": 1311.168,
+      "eval_steps_per_second": 40.975,
+      "step": 1373744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7499995086358882e-05,
+      "loss": 3.9533,
+      "step": 1374208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7491609138848362e-05,
+      "loss": 3.9485,
+      "step": 1374720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7483223191337842e-05,
+      "loss": 3.9648,
+      "step": 1375232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7474837243827322e-05,
+      "loss": 3.9553,
+      "step": 1375744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7466451296316802e-05,
+      "loss": 3.9696,
+      "step": 1376256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7458065348806282e-05,
+      "loss": 3.9507,
+      "step": 1376768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7449679401295762e-05,
+      "loss": 3.9546,
+      "step": 1377280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.744129345378524e-05,
+      "loss": 3.9434,
+      "step": 1377792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.743290750627472e-05,
+      "loss": 3.9611,
+      "step": 1378304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.74245215587642e-05,
+      "loss": 3.9523,
+      "step": 1378816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.741613561125368e-05,
+      "loss": 3.9523,
+      "step": 1379328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.740774966374316e-05,
+      "loss": 3.9628,
+      "step": 1379840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739938009503637e-05,
+      "loss": 3.9424,
+      "step": 1380352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.739099414752585e-05,
+      "loss": 3.9566,
+      "step": 1380864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.738260820001533e-05,
+      "loss": 3.9411,
+      "step": 1381376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.737422225250481e-05,
+      "loss": 3.9489,
+      "step": 1381888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.736583630499429e-05,
+      "loss": 3.9493,
+      "step": 1382400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.735745035748377e-05,
+      "loss": 3.9527,
+      "step": 1382912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734906440997325e-05,
+      "loss": 3.9513,
+      "step": 1383424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.734067846246273e-05,
+      "loss": 3.9742,
+      "step": 1383936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7332308893755943e-05,
+      "loss": 3.9482,
+      "step": 1384448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7323922946245423e-05,
+      "loss": 3.9565,
+      "step": 1384960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7315536998734903e-05,
+      "loss": 3.9538,
+      "step": 1385472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7307151051224383e-05,
+      "loss": 3.9628,
+      "step": 1385984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7298765103713863e-05,
+      "loss": 3.9516,
+      "step": 1386496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7290379156203343e-05,
+      "loss": 3.9589,
+      "step": 1387008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7281993208692823e-05,
+      "loss": 3.9426,
+      "step": 1387520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7273607261182306e-05,
+      "loss": 3.9536,
+      "step": 1388032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7265237692475515e-05,
+      "loss": 3.9429,
+      "step": 1388544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7256851744964995e-05,
+      "loss": 3.9562,
+      "step": 1389056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7248465797454475e-05,
+      "loss": 3.9551,
+      "step": 1389568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7240096228747684e-05,
+      "loss": 3.955,
+      "step": 1390080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7231710281237168e-05,
+      "loss": 3.9645,
+      "step": 1390592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7223340712530377e-05,
+      "loss": 3.9515,
+      "step": 1391104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7214971143823586e-05,
+      "loss": 3.9584,
+      "step": 1391616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7206585196313066e-05,
+      "loss": 3.9565,
+      "step": 1392128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7198199248802546e-05,
+      "loss": 3.9505,
+      "step": 1392640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718981330129203e-05,
+      "loss": 3.9452,
+      "step": 1393152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.718142735378151e-05,
+      "loss": 3.94,
+      "step": 1393664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.717304140627099e-05,
+      "loss": 3.9445,
+      "step": 1394176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.716465545876047e-05,
+      "loss": 3.949,
+      "step": 1394688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.715626951124995e-05,
+      "loss": 3.9586,
+      "step": 1395200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.714788356373943e-05,
+      "loss": 3.9512,
+      "step": 1395712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7139497616228905e-05,
+      "loss": 3.9458,
+      "step": 1396224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7131111668718385e-05,
+      "loss": 3.9559,
+      "step": 1396736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7122725721207865e-05,
+      "loss": 3.9475,
+      "step": 1397248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7114339773697345e-05,
+      "loss": 3.9503,
+      "step": 1397760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7105953826186825e-05,
+      "loss": 3.9416,
+      "step": 1398272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7097567878676305e-05,
+      "loss": 3.9384,
+      "step": 1398784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7089181931165785e-05,
+      "loss": 3.967,
+      "step": 1399296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7080812362458998e-05,
+      "loss": 3.9596,
+      "step": 1399808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7072442793752207e-05,
+      "loss": 3.9452,
+      "step": 1400320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7064056846241687e-05,
+      "loss": 3.9483,
+      "step": 1400832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7055670898731167e-05,
+      "loss": 3.9588,
+      "step": 1401344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7047284951220647e-05,
+      "loss": 3.9371,
+      "step": 1401856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7038899003710126e-05,
+      "loss": 3.9564,
+      "step": 1402368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.703052943500334e-05,
+      "loss": 3.9365,
+      "step": 1402880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.702214348749282e-05,
+      "loss": 3.9636,
+      "step": 1403392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.70137575399823e-05,
+      "loss": 3.9578,
+      "step": 1403904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.700537159247178e-05,
+      "loss": 3.9444,
+      "step": 1404416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.699698564496126e-05,
+      "loss": 3.9387,
+      "step": 1404928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698859969745074e-05,
+      "loss": 3.9601,
+      "step": 1405440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.698023012874395e-05,
+      "loss": 3.9269,
+      "step": 1405952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.697186056003716e-05,
+      "loss": 3.9437,
+      "step": 1406464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.696347461252664e-05,
+      "loss": 3.9567,
+      "step": 1406976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.695508866501612e-05,
+      "loss": 3.9529,
+      "step": 1407488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.69467027175056e-05,
+      "loss": 3.9465,
+      "step": 1408000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.693831676999508e-05,
+      "loss": 3.9336,
+      "step": 1408512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6929947201288293e-05,
+      "loss": 3.941,
+      "step": 1409024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6921561253777773e-05,
+      "loss": 3.9398,
+      "step": 1409536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6913175306267253e-05,
+      "loss": 3.9533,
+      "step": 1410048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6904789358756732e-05,
+      "loss": 3.9425,
+      "step": 1410560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6896403411246212e-05,
+      "loss": 3.9521,
+      "step": 1411072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6888017463735692e-05,
+      "loss": 3.9539,
+      "step": 1411584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6879631516225172e-05,
+      "loss": 3.9571,
+      "step": 1412096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6871245568714652e-05,
+      "loss": 3.942,
+      "step": 1412608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6862876000007865e-05,
+      "loss": 3.9473,
+      "step": 1413120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6854490052497345e-05,
+      "loss": 3.9539,
+      "step": 1413632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6846104104986825e-05,
+      "loss": 3.9467,
+      "step": 1414144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6837718157476305e-05,
+      "loss": 3.9466,
+      "step": 1414656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6829332209965785e-05,
+      "loss": 3.9479,
+      "step": 1415168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.682094626245526e-05,
+      "loss": 3.9518,
+      "step": 1415680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6812576693748477e-05,
+      "loss": 3.9512,
+      "step": 1416192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6804190746237957e-05,
+      "loss": 3.9349,
+      "step": 1416704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6795804798727437e-05,
+      "loss": 3.9448,
+      "step": 1417216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.678741885121691e-05,
+      "loss": 3.943,
+      "step": 1417728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.677903290370639e-05,
+      "loss": 3.9523,
+      "step": 1418240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6770663334999606e-05,
+      "loss": 3.9482,
+      "step": 1418752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6762277387489086e-05,
+      "loss": 3.9453,
+      "step": 1419264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6753891439978563e-05,
+      "loss": 3.9401,
+      "step": 1419776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6745505492468042e-05,
+      "loss": 3.9335,
+      "step": 1420288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6737119544957522e-05,
+      "loss": 3.9409,
+      "step": 1420800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6728733597447002e-05,
+      "loss": 3.9482,
+      "step": 1421312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6720364028740215e-05,
+      "loss": 3.955,
+      "step": 1421824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6711978081229695e-05,
+      "loss": 3.954,
+      "step": 1422336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6703592133719175e-05,
+      "loss": 3.9473,
+      "step": 1422848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6695206186208655e-05,
+      "loss": 3.9426,
+      "step": 1423360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6686836617501864e-05,
+      "loss": 3.9413,
+      "step": 1423872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6678450669991344e-05,
+      "loss": 3.9478,
+      "step": 1424384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6670064722480824e-05,
+      "loss": 3.955,
+      "step": 1424896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6661678774970307e-05,
+      "loss": 3.9532,
+      "step": 1425408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6653292827459787e-05,
+      "loss": 3.9458,
+      "step": 1425920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6644906879949267e-05,
+      "loss": 3.9393,
+      "step": 1426432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6636520932438747e-05,
+      "loss": 3.9574,
+      "step": 1426944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6628134984928227e-05,
+      "loss": 3.9359,
+      "step": 1427456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6619749037417707e-05,
+      "loss": 3.9464,
+      "step": 1427968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6611379468710916e-05,
+      "loss": 3.9452,
+      "step": 1428480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.66029935212004e-05,
+      "loss": 3.9455,
+      "step": 1428992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.659460757368988e-05,
+      "loss": 3.9358,
+      "step": 1429504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.658622162617936e-05,
+      "loss": 3.9482,
+      "step": 1430016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.657783567866884e-05,
+      "loss": 3.9345,
+      "step": 1430528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.656944973115832e-05,
+      "loss": 3.9449,
+      "step": 1431040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.65610637836478e-05,
+      "loss": 3.9484,
+      "step": 1431552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6552677836137276e-05,
+      "loss": 3.937,
+      "step": 1432064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.654430826743049e-05,
+      "loss": 3.9578,
+      "step": 1432576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.653592231991997e-05,
+      "loss": 3.9616,
+      "step": 1433088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6527536372409445e-05,
+      "loss": 3.9515,
+      "step": 1433600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6519150424898925e-05,
+      "loss": 3.9438,
+      "step": 1434112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.651078085619214e-05,
+      "loss": 3.9391,
+      "step": 1434624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.650239490868162e-05,
+      "loss": 3.9457,
+      "step": 1435136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6494025339974833e-05,
+      "loss": 3.939,
+      "step": 1435648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6485639392464313e-05,
+      "loss": 3.9537,
+      "step": 1436160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6477253444953793e-05,
+      "loss": 3.9362,
+      "step": 1436672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6468867497443273e-05,
+      "loss": 3.9503,
+      "step": 1437184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6460497928736482e-05,
+      "loss": 3.9503,
+      "step": 1437696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6452111981225962e-05,
+      "loss": 3.9467,
+      "step": 1438208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6443726033715445e-05,
+      "loss": 3.939,
+      "step": 1438720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.643534008620492e-05,
+      "loss": 3.9434,
+      "step": 1439232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.64269541386944e-05,
+      "loss": 3.945,
+      "step": 1439744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6418568191183878e-05,
+      "loss": 3.9505,
+      "step": 1440256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6410182243673358e-05,
+      "loss": 3.9538,
+      "step": 1440768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6401796296162838e-05,
+      "loss": 3.957,
+      "step": 1441280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.639342672745605e-05,
+      "loss": 3.9335,
+      "step": 1441792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6385057158749267e-05,
+      "loss": 3.9509,
+      "step": 1442304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6376671211238746e-05,
+      "loss": 3.9542,
+      "step": 1442816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.636828526372822e-05,
+      "loss": 3.9458,
+      "step": 1443328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.63598993162177e-05,
+      "loss": 3.948,
+      "step": 1443840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6351513368707183e-05,
+      "loss": 3.9578,
+      "step": 1444352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6343127421196663e-05,
+      "loss": 3.9484,
+      "step": 1444864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6334741473686143e-05,
+      "loss": 3.9613,
+      "step": 1445376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6326355526175623e-05,
+      "loss": 3.9453,
+      "step": 1445888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6317969578665103e-05,
+      "loss": 3.9533,
+      "step": 1446400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6309600009958312e-05,
+      "loss": 3.9455,
+      "step": 1446912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6301214062447792e-05,
+      "loss": 3.9437,
+      "step": 1447424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6292828114937272e-05,
+      "loss": 3.9454,
+      "step": 1447936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.628447492503422e-05,
+      "loss": 3.9411,
+      "step": 1448448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6276088977523693e-05,
+      "loss": 3.9433,
+      "step": 1448960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6267703030013173e-05,
+      "loss": 3.9443,
+      "step": 1449472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.6259317082502653e-05,
+      "loss": 3.9425,
+      "step": 1449984
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0152082443237305,
+      "eval_runtime": 294.703,
+      "eval_samples_per_second": 1294.832,
+      "eval_steps_per_second": 40.464,
+      "step": 1450064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6250931134992137e-05,
+      "loss": 3.9518,
+      "step": 1450496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6242545187481617e-05,
+      "loss": 3.9387,
+      "step": 1451008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6234159239971097e-05,
+      "loss": 3.9549,
+      "step": 1451520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6225773292460577e-05,
+      "loss": 3.9429,
+      "step": 1452032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6217387344950056e-05,
+      "loss": 3.9661,
+      "step": 1452544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6209001397439536e-05,
+      "loss": 3.9393,
+      "step": 1453056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6200631828732746e-05,
+      "loss": 3.9461,
+      "step": 1453568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6192245881222225e-05,
+      "loss": 3.9348,
+      "step": 1454080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.618385993371171e-05,
+      "loss": 3.9517,
+      "step": 1454592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.617547398620119e-05,
+      "loss": 3.947,
+      "step": 1455104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6167104417494398e-05,
+      "loss": 3.9401,
+      "step": 1455616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6158718469983878e-05,
+      "loss": 3.9535,
+      "step": 1456128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6150348901277087e-05,
+      "loss": 3.9347,
+      "step": 1456640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.614196295376657e-05,
+      "loss": 3.9502,
+      "step": 1457152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.613357700625605e-05,
+      "loss": 3.9356,
+      "step": 1457664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.612519105874553e-05,
+      "loss": 3.9377,
+      "step": 1458176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.611680511123501e-05,
+      "loss": 3.9389,
+      "step": 1458688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.610843554252822e-05,
+      "loss": 3.9451,
+      "step": 1459200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.61000495950177e-05,
+      "loss": 3.9446,
+      "step": 1459712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.609166364750718e-05,
+      "loss": 3.962,
+      "step": 1460224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6083277699996662e-05,
+      "loss": 3.941,
+      "step": 1460736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6074891752486142e-05,
+      "loss": 3.9507,
+      "step": 1461248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6066505804975622e-05,
+      "loss": 3.9479,
+      "step": 1461760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6058119857465102e-05,
+      "loss": 3.9508,
+      "step": 1462272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6049733909954582e-05,
+      "loss": 3.9446,
+      "step": 1462784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6041347962444055e-05,
+      "loss": 3.9469,
+      "step": 1463296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6032994772541004e-05,
+      "loss": 3.9396,
+      "step": 1463808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6024608825030484e-05,
+      "loss": 3.9425,
+      "step": 1464320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.6016222877519964e-05,
+      "loss": 3.9348,
+      "step": 1464832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6007836930009444e-05,
+      "loss": 3.9479,
+      "step": 1465344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5999450982498924e-05,
+      "loss": 3.9511,
+      "step": 1465856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5991065034988404e-05,
+      "loss": 3.9448,
+      "step": 1466368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.598267908747788e-05,
+      "loss": 3.9578,
+      "step": 1466880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.597429313996736e-05,
+      "loss": 3.946,
+      "step": 1467392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5965923571260576e-05,
+      "loss": 3.9438,
+      "step": 1467904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5957537623750056e-05,
+      "loss": 3.9489,
+      "step": 1468416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5949168055043265e-05,
+      "loss": 3.9478,
+      "step": 1468928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5940782107532745e-05,
+      "loss": 3.9348,
+      "step": 1469440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5932396160022225e-05,
+      "loss": 3.9304,
+      "step": 1469952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.59240102125117e-05,
+      "loss": 3.9378,
+      "step": 1470464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.591562426500118e-05,
+      "loss": 3.9441,
+      "step": 1470976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.590723831749066e-05,
+      "loss": 3.9486,
+      "step": 1471488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.589885236998014e-05,
+      "loss": 3.9425,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.589046642246962e-05,
+      "loss": 3.9355,
+      "step": 1472512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.58820804749591e-05,
+      "loss": 3.9517,
+      "step": 1473024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.587372728505605e-05,
+      "loss": 3.9395,
+      "step": 1473536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5865341337545523e-05,
+      "loss": 3.9415,
+      "step": 1474048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5856955390035003e-05,
+      "loss": 3.935,
+      "step": 1474560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5848569442524483e-05,
+      "loss": 3.926,
+      "step": 1475072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5840183495013963e-05,
+      "loss": 3.9635,
+      "step": 1475584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5831797547503446e-05,
+      "loss": 3.9496,
+      "step": 1476096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5823411599992926e-05,
+      "loss": 3.9418,
+      "step": 1476608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5815025652482406e-05,
+      "loss": 3.9429,
+      "step": 1477120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5806639704971886e-05,
+      "loss": 3.9452,
+      "step": 1477632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5798253757461366e-05,
+      "loss": 3.9285,
+      "step": 1478144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5789867809950846e-05,
+      "loss": 3.9485,
+      "step": 1478656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5781481862440326e-05,
+      "loss": 3.9317,
+      "step": 1479168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.577311229373354e-05,
+      "loss": 3.948,
+      "step": 1479680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.576472634622302e-05,
+      "loss": 3.9547,
+      "step": 1480192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5756340398712498e-05,
+      "loss": 3.9364,
+      "step": 1480704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5747970830005707e-05,
+      "loss": 3.9297,
+      "step": 1481216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5739584882495187e-05,
+      "loss": 3.9559,
+      "step": 1481728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5731198934984667e-05,
+      "loss": 3.9214,
+      "step": 1482240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5722812987474147e-05,
+      "loss": 3.9326,
+      "step": 1482752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.571444341876736e-05,
+      "loss": 3.9482,
+      "step": 1483264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.570605747125684e-05,
+      "loss": 3.9456,
+      "step": 1483776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.569767152374632e-05,
+      "loss": 3.9399,
+      "step": 1484288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.56892855762358e-05,
+      "loss": 3.9298,
+      "step": 1484800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.568089962872528e-05,
+      "loss": 3.9289,
+      "step": 1485312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.567251368121476e-05,
+      "loss": 3.9289,
+      "step": 1485824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5664144112507972e-05,
+      "loss": 3.9474,
+      "step": 1486336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5655758164997452e-05,
+      "loss": 3.9325,
+      "step": 1486848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5647372217486932e-05,
+      "loss": 3.9464,
+      "step": 1487360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5638986269976412e-05,
+      "loss": 3.9459,
+      "step": 1487872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5630600322465885e-05,
+      "loss": 3.9524,
+      "step": 1488384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.562221437495537e-05,
+      "loss": 3.9344,
+      "step": 1488896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.561382842744485e-05,
+      "loss": 3.9373,
+      "step": 1489408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5605442479934328e-05,
+      "loss": 3.9493,
+      "step": 1489920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5597072911227537e-05,
+      "loss": 3.9359,
+      "step": 1490432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5588686963717017e-05,
+      "loss": 3.9456,
+      "step": 1490944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5580301016206497e-05,
+      "loss": 3.9378,
+      "step": 1491456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5571931447499713e-05,
+      "loss": 3.9399,
+      "step": 1491968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.556354549998919e-05,
+      "loss": 3.9453,
+      "step": 1492480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.555515955247867e-05,
+      "loss": 3.9245,
+      "step": 1492992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.554677360496815e-05,
+      "loss": 3.9391,
+      "step": 1493504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553838765745763e-05,
+      "loss": 3.9375,
+      "step": 1494016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.553000170994711e-05,
+      "loss": 3.9463,
+      "step": 1494528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.552161576243659e-05,
+      "loss": 3.9372,
+      "step": 1495040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.5513246193729802e-05,
+      "loss": 3.9382,
+      "step": 1495552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5504860246219282e-05,
+      "loss": 3.9333,
+      "step": 1496064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5496474298708762e-05,
+      "loss": 3.9339,
+      "step": 1496576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.548810473000197e-05,
+      "loss": 3.9294,
+      "step": 1497088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.547971878249145e-05,
+      "loss": 3.9394,
+      "step": 1497600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.547133283498093e-05,
+      "loss": 3.9509,
+      "step": 1498112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.546294688747041e-05,
+      "loss": 3.9471,
+      "step": 1498624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5454560939959894e-05,
+      "loss": 3.9396,
+      "step": 1499136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5446174992449374e-05,
+      "loss": 3.931,
+      "step": 1499648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5437789044938854e-05,
+      "loss": 3.9353,
+      "step": 1500160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5429403097428334e-05,
+      "loss": 3.9384,
+      "step": 1500672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5421017149917814e-05,
+      "loss": 3.9449,
+      "step": 1501184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5412631202407294e-05,
+      "loss": 3.9456,
+      "step": 1501696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5404245254896774e-05,
+      "loss": 3.9386,
+      "step": 1502208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.539585930738625e-05,
+      "loss": 3.9304,
+      "step": 1502720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5387489738679466e-05,
+      "loss": 3.9535,
+      "step": 1503232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5379103791168946e-05,
+      "loss": 3.9279,
+      "step": 1503744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5370717843658426e-05,
+      "loss": 3.9385,
+      "step": 1504256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.53623318961479e-05,
+      "loss": 3.9319,
+      "step": 1504768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.535394594863738e-05,
+      "loss": 3.9392,
+      "step": 1505280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5345560001126863e-05,
+      "loss": 3.9302,
+      "step": 1505792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.533719043242008e-05,
+      "loss": 3.9385,
+      "step": 1506304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5328804484909552e-05,
+      "loss": 3.9342,
+      "step": 1506816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5320418537399032e-05,
+      "loss": 3.9307,
+      "step": 1507328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5312032589888512e-05,
+      "loss": 3.9409,
+      "step": 1507840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5303663021181724e-05,
+      "loss": 3.9315,
+      "step": 1508352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5295277073671204e-05,
+      "loss": 3.9495,
+      "step": 1508864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5286891126160684e-05,
+      "loss": 3.9548,
+      "step": 1509376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5278505178650164e-05,
+      "loss": 3.9459,
+      "step": 1509888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5270119231139644e-05,
+      "loss": 3.9358,
+      "step": 1510400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5261733283629124e-05,
+      "loss": 3.9348,
+      "step": 1510912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5253347336118604e-05,
+      "loss": 3.9327,
+      "step": 1511424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5244961388608084e-05,
+      "loss": 3.9362,
+      "step": 1511936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5236591819901296e-05,
+      "loss": 3.9443,
+      "step": 1512448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5228205872390776e-05,
+      "loss": 3.9282,
+      "step": 1512960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5219836303683985e-05,
+      "loss": 3.9426,
+      "step": 1513472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5211450356173465e-05,
+      "loss": 3.9402,
+      "step": 1513984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5203064408662945e-05,
+      "loss": 3.9405,
+      "step": 1514496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5194678461152425e-05,
+      "loss": 3.935,
+      "step": 1515008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.518629251364191e-05,
+      "loss": 3.9368,
+      "step": 1515520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5177939323738847e-05,
+      "loss": 3.9358,
+      "step": 1516032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5169553376228327e-05,
+      "loss": 3.9431,
+      "step": 1516544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5161167428717807e-05,
+      "loss": 3.9446,
+      "step": 1517056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5152781481207287e-05,
+      "loss": 3.9489,
+      "step": 1517568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.514439553369677e-05,
+      "loss": 3.9286,
+      "step": 1518080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.513602596498998e-05,
+      "loss": 3.9424,
+      "step": 1518592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.512764001747946e-05,
+      "loss": 3.946,
+      "step": 1519104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511925406996894e-05,
+      "loss": 3.9383,
+      "step": 1519616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.511086812245842e-05,
+      "loss": 3.9395,
+      "step": 1520128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.51024821749479e-05,
+      "loss": 3.9504,
+      "step": 1520640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.509409622743738e-05,
+      "loss": 3.9393,
+      "step": 1521152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.508571027992686e-05,
+      "loss": 3.9543,
+      "step": 1521664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5077324332416342e-05,
+      "loss": 3.935,
+      "step": 1522176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5068938384905822e-05,
+      "loss": 3.9445,
+      "step": 1522688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5060552437395302e-05,
+      "loss": 3.9378,
+      "step": 1523200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5052166489884782e-05,
+      "loss": 3.9379,
+      "step": 1523712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5043780542374262e-05,
+      "loss": 3.9362,
+      "step": 1524224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.503541097366747e-05,
+      "loss": 3.9365,
+      "step": 1524736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.502702502615695e-05,
+      "loss": 3.9335,
+      "step": 1525248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5018655457450164e-05,
+      "loss": 3.9367,
+      "step": 1525760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.5010269509939644e-05,
+      "loss": 3.9366,
+      "step": 1526272
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.012029647827148,
+      "eval_runtime": 293.0853,
+      "eval_samples_per_second": 1301.979,
+      "eval_steps_per_second": 40.688,
+      "step": 1526384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5001883562429124e-05,
+      "loss": 3.94,
+      "step": 1526784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.49934976149186e-05,
+      "loss": 3.9317,
+      "step": 1527296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.498511166740808e-05,
+      "loss": 3.947,
+      "step": 1527808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4976725719897563e-05,
+      "loss": 3.9362,
+      "step": 1528320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4968339772387043e-05,
+      "loss": 3.9551,
+      "step": 1528832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4959953824876523e-05,
+      "loss": 3.9375,
+      "step": 1529344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4951567877366e-05,
+      "loss": 3.9354,
+      "step": 1529856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.494318192985548e-05,
+      "loss": 3.9256,
+      "step": 1530368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.493479598234496e-05,
+      "loss": 3.9392,
+      "step": 1530880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.492641003483444e-05,
+      "loss": 3.9397,
+      "step": 1531392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.491802408732392e-05,
+      "loss": 3.9374,
+      "step": 1531904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4909638139813403e-05,
+      "loss": 3.9413,
+      "step": 1532416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4901268571106612e-05,
+      "loss": 3.9305,
+      "step": 1532928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4892882623596092e-05,
+      "loss": 3.9405,
+      "step": 1533440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4884496676085572e-05,
+      "loss": 3.927,
+      "step": 1533952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4876110728575052e-05,
+      "loss": 3.9293,
+      "step": 1534464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4867724781064532e-05,
+      "loss": 3.9318,
+      "step": 1534976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4859338833554012e-05,
+      "loss": 3.9381,
+      "step": 1535488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4850952886043492e-05,
+      "loss": 3.9365,
+      "step": 1536000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4842566938532972e-05,
+      "loss": 3.9532,
+      "step": 1536512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4834180991022452e-05,
+      "loss": 3.9365,
+      "step": 1537024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4825795043511932e-05,
+      "loss": 3.9475,
+      "step": 1537536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4817409096001412e-05,
+      "loss": 3.9351,
+      "step": 1538048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4809023148490888e-05,
+      "loss": 3.9433,
+      "step": 1538560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4800653579784104e-05,
+      "loss": 3.9344,
+      "step": 1539072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4792267632273584e-05,
+      "loss": 3.9458,
+      "step": 1539584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4783898063566793e-05,
+      "loss": 3.9305,
+      "step": 1540096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4775512116056273e-05,
+      "loss": 3.9338,
+      "step": 1540608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4767126168545753e-05,
+      "loss": 3.9314,
+      "step": 1541120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4758740221035233e-05,
+      "loss": 3.9377,
+      "step": 1541632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4750354273524713e-05,
+      "loss": 3.9432,
+      "step": 1542144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4741968326014193e-05,
+      "loss": 3.9395,
+      "step": 1542656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4733598757307405e-05,
+      "loss": 3.9491,
+      "step": 1543168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4725212809796885e-05,
+      "loss": 3.9371,
+      "step": 1543680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4716843241090095e-05,
+      "loss": 3.9385,
+      "step": 1544192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4708457293579574e-05,
+      "loss": 3.9418,
+      "step": 1544704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4700071346069058e-05,
+      "loss": 3.9428,
+      "step": 1545216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4691685398558538e-05,
+      "loss": 3.9259,
+      "step": 1545728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4683299451048014e-05,
+      "loss": 3.9252,
+      "step": 1546240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4674913503537494e-05,
+      "loss": 3.9257,
+      "step": 1546752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4666527556026974e-05,
+      "loss": 3.9369,
+      "step": 1547264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4658141608516454e-05,
+      "loss": 3.9437,
+      "step": 1547776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4649755661005934e-05,
+      "loss": 3.9358,
+      "step": 1548288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4641369713495414e-05,
+      "loss": 3.9281,
+      "step": 1548800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4632983765984894e-05,
+      "loss": 3.9385,
+      "step": 1549312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4624614197278107e-05,
+      "loss": 3.9361,
+      "step": 1549824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4616228249767587e-05,
+      "loss": 3.9335,
+      "step": 1550336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4607842302257066e-05,
+      "loss": 3.9279,
+      "step": 1550848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4599456354746543e-05,
+      "loss": 3.9135,
+      "step": 1551360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4591070407236026e-05,
+      "loss": 3.9573,
+      "step": 1551872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4582684459725506e-05,
+      "loss": 3.9418,
+      "step": 1552384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.457431489101872e-05,
+      "loss": 3.9378,
+      "step": 1552896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4565928943508195e-05,
+      "loss": 3.9358,
+      "step": 1553408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4557542995997675e-05,
+      "loss": 3.9363,
+      "step": 1553920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4549157048487155e-05,
+      "loss": 3.9259,
+      "step": 1554432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4540787479780368e-05,
+      "loss": 3.9355,
+      "step": 1554944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4532401532269848e-05,
+      "loss": 3.929,
+      "step": 1555456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4524015584759328e-05,
+      "loss": 3.9368,
+      "step": 1555968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4515629637248808e-05,
+      "loss": 3.9486,
+      "step": 1556480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4507243689738288e-05,
+      "loss": 3.9297,
+      "step": 1556992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4498874121031497e-05,
+      "loss": 3.9209,
+      "step": 1557504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.449048817352098e-05,
+      "loss": 3.9469,
+      "step": 1558016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.448210222601046e-05,
+      "loss": 3.9156,
+      "step": 1558528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.447371627849994e-05,
+      "loss": 3.9242,
+      "step": 1559040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.446533033098942e-05,
+      "loss": 3.941,
+      "step": 1559552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.44569443834789e-05,
+      "loss": 3.936,
+      "step": 1560064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.444857481477211e-05,
+      "loss": 3.9329,
+      "step": 1560576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.444020524606532e-05,
+      "loss": 3.9236,
+      "step": 1561088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.44318192985548e-05,
+      "loss": 3.9146,
+      "step": 1561600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.442343335104428e-05,
+      "loss": 3.9233,
+      "step": 1562112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.441504740353376e-05,
+      "loss": 3.9408,
+      "step": 1562624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.440666145602324e-05,
+      "loss": 3.9273,
+      "step": 1563136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.439827550851272e-05,
+      "loss": 3.9349,
+      "step": 1563648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4389889561002198e-05,
+      "loss": 3.9392,
+      "step": 1564160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4381519992295414e-05,
+      "loss": 3.9481,
+      "step": 1564672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4373134044784894e-05,
+      "loss": 3.927,
+      "step": 1565184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4364748097274374e-05,
+      "loss": 3.9277,
+      "step": 1565696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.435636214976385e-05,
+      "loss": 3.9421,
+      "step": 1566208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.434797620225333e-05,
+      "loss": 3.923,
+      "step": 1566720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.433959025474281e-05,
+      "loss": 3.9407,
+      "step": 1567232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.433120430723229e-05,
+      "loss": 3.9326,
+      "step": 1567744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4322818359721773e-05,
+      "loss": 3.9316,
+      "step": 1568256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4314432412211253e-05,
+      "loss": 3.9366,
+      "step": 1568768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.430604646470073e-05,
+      "loss": 3.9199,
+      "step": 1569280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4297676895993942e-05,
+      "loss": 3.9315,
+      "step": 1569792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4289290948483422e-05,
+      "loss": 3.9277,
+      "step": 1570304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4280905000972902e-05,
+      "loss": 3.9373,
+      "step": 1570816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4272519053462382e-05,
+      "loss": 3.9266,
+      "step": 1571328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4264149484755595e-05,
+      "loss": 3.9343,
+      "step": 1571840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4255763537245075e-05,
+      "loss": 3.9214,
+      "step": 1572352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4247377589734555e-05,
+      "loss": 3.9252,
+      "step": 1572864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4239008021027764e-05,
+      "loss": 3.9219,
+      "step": 1573376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4230622073517244e-05,
+      "loss": 3.9293,
+      "step": 1573888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4222236126006727e-05,
+      "loss": 3.9463,
+      "step": 1574400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4213850178496204e-05,
+      "loss": 3.9368,
+      "step": 1574912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4205464230985684e-05,
+      "loss": 3.9343,
+      "step": 1575424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4197078283475164e-05,
+      "loss": 3.9237,
+      "step": 1575936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4188692335964643e-05,
+      "loss": 3.9284,
+      "step": 1576448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4180306388454123e-05,
+      "loss": 3.9296,
+      "step": 1576960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4171920440943603e-05,
+      "loss": 3.9374,
+      "step": 1577472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4163534493433083e-05,
+      "loss": 3.935,
+      "step": 1577984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4155148545922563e-05,
+      "loss": 3.9324,
+      "step": 1578496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4146762598412043e-05,
+      "loss": 3.9239,
+      "step": 1579008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4138409408508985e-05,
+      "loss": 3.9469,
+      "step": 1579520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4130023460998465e-05,
+      "loss": 3.917,
+      "step": 1580032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4121637513487945e-05,
+      "loss": 3.9336,
+      "step": 1580544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4113251565977428e-05,
+      "loss": 3.925,
+      "step": 1581056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4104865618466908e-05,
+      "loss": 3.932,
+      "step": 1581568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4096496049760117e-05,
+      "loss": 3.9203,
+      "step": 1582080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4088110102249597e-05,
+      "loss": 3.9366,
+      "step": 1582592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4079724154739077e-05,
+      "loss": 3.9264,
+      "step": 1583104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4071338207228557e-05,
+      "loss": 3.9271,
+      "step": 1583616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4062952259718037e-05,
+      "loss": 3.9303,
+      "step": 1584128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.405458269101125e-05,
+      "loss": 3.9292,
+      "step": 1584640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.404619674350073e-05,
+      "loss": 3.9389,
+      "step": 1585152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4037810795990206e-05,
+      "loss": 3.9479,
+      "step": 1585664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4029424848479686e-05,
+      "loss": 3.9344,
+      "step": 1586176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4021038900969166e-05,
+      "loss": 3.93,
+      "step": 1586688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.4012652953458646e-05,
+      "loss": 3.9286,
+      "step": 1587200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.400426700594813e-05,
+      "loss": 3.9275,
+      "step": 1587712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.399588105843761e-05,
+      "loss": 3.9259,
+      "step": 1588224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3987511489730818e-05,
+      "loss": 3.9387,
+      "step": 1588736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3979125542220298e-05,
+      "loss": 3.9181,
+      "step": 1589248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3970739594709778e-05,
+      "loss": 3.9352,
+      "step": 1589760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.396237002600299e-05,
+      "loss": 3.9295,
+      "step": 1590272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.395398407849247e-05,
+      "loss": 3.9352,
+      "step": 1590784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.394559813098195e-05,
+      "loss": 3.9255,
+      "step": 1591296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.393721218347143e-05,
+      "loss": 3.9308,
+      "step": 1591808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.392882623596091e-05,
+      "loss": 3.9262,
+      "step": 1592320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3920440288450387e-05,
+      "loss": 3.9405,
+      "step": 1592832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3912054340939867e-05,
+      "loss": 3.9354,
+      "step": 1593344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.390366839342935e-05,
+      "loss": 3.9426,
+      "step": 1593856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.389528244591883e-05,
+      "loss": 3.9206,
+      "step": 1594368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.388689649840831e-05,
+      "loss": 3.936,
+      "step": 1594880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387851055089779e-05,
+      "loss": 3.9411,
+      "step": 1595392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.387012460338727e-05,
+      "loss": 3.9285,
+      "step": 1595904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3861771413484212e-05,
+      "loss": 3.9307,
+      "step": 1596416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3853385465973692e-05,
+      "loss": 3.9429,
+      "step": 1596928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3844999518463172e-05,
+      "loss": 3.9349,
+      "step": 1597440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.383661357095265e-05,
+      "loss": 3.9505,
+      "step": 1597952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.382822762344213e-05,
+      "loss": 3.9268,
+      "step": 1598464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381985805473534e-05,
+      "loss": 3.9401,
+      "step": 1598976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.381147210722482e-05,
+      "loss": 3.9293,
+      "step": 1599488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.38030861597143e-05,
+      "loss": 3.9317,
+      "step": 1600000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3794700212203784e-05,
+      "loss": 3.9335,
+      "step": 1600512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3786314264693264e-05,
+      "loss": 3.9243,
+      "step": 1601024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3777944695986473e-05,
+      "loss": 3.9269,
+      "step": 1601536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3769558748475953e-05,
+      "loss": 3.9252,
+      "step": 1602048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.3761172800965433e-05,
+      "loss": 3.9314,
+      "step": 1602560
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.009164333343506,
+      "eval_runtime": 290.8687,
+      "eval_samples_per_second": 1311.901,
+      "eval_steps_per_second": 40.998,
+      "step": 1602704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3752786853454913e-05,
+      "loss": 3.9351,
+      "step": 1603072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3744417284748125e-05,
+      "loss": 3.9263,
+      "step": 1603584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3736031337237605e-05,
+      "loss": 3.9371,
+      "step": 1604096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3727645389727085e-05,
+      "loss": 3.9326,
+      "step": 1604608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3719259442216565e-05,
+      "loss": 3.9451,
+      "step": 1605120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3710873494706042e-05,
+      "loss": 3.9317,
+      "step": 1605632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3702503925999254e-05,
+      "loss": 3.9261,
+      "step": 1606144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3694117978488738e-05,
+      "loss": 3.923,
+      "step": 1606656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3685732030978218e-05,
+      "loss": 3.9306,
+      "step": 1607168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3677346083467694e-05,
+      "loss": 3.9327,
+      "step": 1607680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3668960135957174e-05,
+      "loss": 3.9317,
+      "step": 1608192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3660574188446654e-05,
+      "loss": 3.9315,
+      "step": 1608704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.36522209985436e-05,
+      "loss": 3.9255,
+      "step": 1609216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.364383505103308e-05,
+      "loss": 3.9295,
+      "step": 1609728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.363544910352256e-05,
+      "loss": 3.9243,
+      "step": 1610240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.362706315601204e-05,
+      "loss": 3.9171,
+      "step": 1610752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3618677208501516e-05,
+      "loss": 3.927,
+      "step": 1611264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3610291260990995e-05,
+      "loss": 3.9285,
+      "step": 1611776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3601905313480475e-05,
+      "loss": 3.9292,
+      "step": 1612288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.359351936596996e-05,
+      "loss": 3.9464,
+      "step": 1612800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.358513341845944e-05,
+      "loss": 3.9322,
+      "step": 1613312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3576763849752648e-05,
+      "loss": 3.9405,
+      "step": 1613824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.3568377902242128e-05,
+      "loss": 3.9252,
+      "step": 1614336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.356000833353534e-05,
+      "loss": 3.9335,
+      "step": 1614848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.355162238602482e-05,
+      "loss": 3.9292,
+      "step": 1615360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.35432364385143e-05,
+      "loss": 3.9385,
+      "step": 1615872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.353485049100378e-05,
+      "loss": 3.9271,
+      "step": 1616384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.352646454349326e-05,
+      "loss": 3.9239,
+      "step": 1616896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.351807859598274e-05,
+      "loss": 3.9243,
+      "step": 1617408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.350969264847222e-05,
+      "loss": 3.9314,
+      "step": 1617920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3501306700961697e-05,
+      "loss": 3.9311,
+      "step": 1618432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3492920753451177e-05,
+      "loss": 3.9341,
+      "step": 1618944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.348453480594066e-05,
+      "loss": 3.9394,
+      "step": 1619456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.347614885843014e-05,
+      "loss": 3.9306,
+      "step": 1619968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.346776291091962e-05,
+      "loss": 3.9282,
+      "step": 1620480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345939334221283e-05,
+      "loss": 3.9396,
+      "step": 1620992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.345100739470231e-05,
+      "loss": 3.9354,
+      "step": 1621504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.344262144719179e-05,
+      "loss": 3.9183,
+      "step": 1622016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.343423549968127e-05,
+      "loss": 3.9208,
+      "step": 1622528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3425849552170752e-05,
+      "loss": 3.9172,
+      "step": 1623040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.341746360466023e-05,
+      "loss": 3.9247,
+      "step": 1623552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.340907765714971e-05,
+      "loss": 3.939,
+      "step": 1624064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.340070808844292e-05,
+      "loss": 3.9284,
+      "step": 1624576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.33923221409324e-05,
+      "loss": 3.9231,
+      "step": 1625088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3383936193421878e-05,
+      "loss": 3.9335,
+      "step": 1625600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3375566624715094e-05,
+      "loss": 3.9237,
+      "step": 1626112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3367180677204573e-05,
+      "loss": 3.9317,
+      "step": 1626624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3358794729694053e-05,
+      "loss": 3.921,
+      "step": 1627136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.335040878218353e-05,
+      "loss": 3.9017,
+      "step": 1627648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3342039213476742e-05,
+      "loss": 3.9528,
+      "step": 1628160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3333653265966222e-05,
+      "loss": 3.9303,
+      "step": 1628672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3325283697259435e-05,
+      "loss": 3.933,
+      "step": 1629184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3316897749748915e-05,
+      "loss": 3.9248,
+      "step": 1629696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3308511802238395e-05,
+      "loss": 3.9308,
+      "step": 1630208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3300125854727875e-05,
+      "loss": 3.9207,
+      "step": 1630720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.329173990721735e-05,
+      "loss": 3.9235,
+      "step": 1631232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.328335395970683e-05,
+      "loss": 3.9258,
+      "step": 1631744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3274968012196315e-05,
+      "loss": 3.9308,
+      "step": 1632256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3266582064685795e-05,
+      "loss": 3.9412,
+      "step": 1632768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3258196117175275e-05,
+      "loss": 3.92,
+      "step": 1633280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3249826548468484e-05,
+      "loss": 3.9114,
+      "step": 1633792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3241440600957964e-05,
+      "loss": 3.9401,
+      "step": 1634304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3233054653447444e-05,
+      "loss": 3.9088,
+      "step": 1634816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3224668705936924e-05,
+      "loss": 3.9158,
+      "step": 1635328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3216282758426407e-05,
+      "loss": 3.9312,
+      "step": 1635840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3207896810915883e-05,
+      "loss": 3.9341,
+      "step": 1636352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3199510863405363e-05,
+      "loss": 3.9232,
+      "step": 1636864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3191124915894843e-05,
+      "loss": 3.9206,
+      "step": 1637376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3182738968384323e-05,
+      "loss": 3.9039,
+      "step": 1637888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3174369399677532e-05,
+      "loss": 3.9189,
+      "step": 1638400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3165999830970748e-05,
+      "loss": 3.9327,
+      "step": 1638912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3157613883460228e-05,
+      "loss": 3.9178,
+      "step": 1639424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3149227935949708e-05,
+      "loss": 3.9261,
+      "step": 1639936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3140841988439185e-05,
+      "loss": 3.9309,
+      "step": 1640448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3132456040928665e-05,
+      "loss": 3.9421,
+      "step": 1640960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3124070093418145e-05,
+      "loss": 3.9196,
+      "step": 1641472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3115684145907625e-05,
+      "loss": 3.9238,
+      "step": 1641984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3107298198397108e-05,
+      "loss": 3.9313,
+      "step": 1642496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3098928629690317e-05,
+      "loss": 3.9159,
+      "step": 1643008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.309055906098353e-05,
+      "loss": 3.9336,
+      "step": 1643520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3082173113473006e-05,
+      "loss": 3.9231,
+      "step": 1644032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.3073787165962486e-05,
+      "loss": 3.9265,
+      "step": 1644544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.306540121845197e-05,
+      "loss": 3.933,
+      "step": 1645056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.305701527094145e-05,
+      "loss": 3.9134,
+      "step": 1645568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304862932343093e-05,
+      "loss": 3.9163,
+      "step": 1646080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.304024337592041e-05,
+      "loss": 3.922,
+      "step": 1646592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.303185742840989e-05,
+      "loss": 3.933,
+      "step": 1647104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.30234878597031e-05,
+      "loss": 3.919,
+      "step": 1647616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.301511829099631e-05,
+      "loss": 3.927,
+      "step": 1648128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.300673234348579e-05,
+      "loss": 3.9144,
+      "step": 1648640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.299834639597527e-05,
+      "loss": 3.9163,
+      "step": 1649152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.298996044846475e-05,
+      "loss": 3.9163,
+      "step": 1649664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.298157450095423e-05,
+      "loss": 3.9239,
+      "step": 1650176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.297318855344371e-05,
+      "loss": 3.9371,
+      "step": 1650688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.296480260593319e-05,
+      "loss": 3.927,
+      "step": 1651200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.295641665842267e-05,
+      "loss": 3.9281,
+      "step": 1651712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.294803071091215e-05,
+      "loss": 3.9167,
+      "step": 1652224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.293964476340163e-05,
+      "loss": 3.9197,
+      "step": 1652736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.293125881589111e-05,
+      "loss": 3.9244,
+      "step": 1653248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.292287286838059e-05,
+      "loss": 3.9298,
+      "step": 1653760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2914486920870067e-05,
+      "loss": 3.9275,
+      "step": 1654272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.290611735216328e-05,
+      "loss": 3.9239,
+      "step": 1654784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2897747783456492e-05,
+      "loss": 3.9216,
+      "step": 1655296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2889361835945972e-05,
+      "loss": 3.9359,
+      "step": 1655808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2880975888435452e-05,
+      "loss": 3.9117,
+      "step": 1656320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2872589940924932e-05,
+      "loss": 3.9245,
+      "step": 1656832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.286422037221814e-05,
+      "loss": 3.9144,
+      "step": 1657344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2855834424707624e-05,
+      "loss": 3.9264,
+      "step": 1657856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2847448477197104e-05,
+      "loss": 3.9153,
+      "step": 1658368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2839062529686584e-05,
+      "loss": 3.9268,
+      "step": 1658880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2830676582176064e-05,
+      "loss": 3.9152,
+      "step": 1659392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.282229063466554e-05,
+      "loss": 3.922,
+      "step": 1659904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.281390468715502e-05,
+      "loss": 3.9222,
+      "step": 1660416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.28055187396445e-05,
+      "loss": 3.9195,
+      "step": 1660928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2797149170937716e-05,
+      "loss": 3.9299,
+      "step": 1661440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2788779602230925e-05,
+      "loss": 3.9394,
+      "step": 1661952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2780393654720405e-05,
+      "loss": 3.9267,
+      "step": 1662464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2772024086013615e-05,
+      "loss": 3.9266,
+      "step": 1662976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2763638138503094e-05,
+      "loss": 3.9174,
+      "step": 1663488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2755252190992578e-05,
+      "loss": 3.9193,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2746866243482058e-05,
+      "loss": 3.9236,
+      "step": 1664512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2738480295971538e-05,
+      "loss": 3.932,
+      "step": 1665024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2730094348461014e-05,
+      "loss": 3.9069,
+      "step": 1665536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2721708400950494e-05,
+      "loss": 3.9267,
+      "step": 1666048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2713338832243707e-05,
+      "loss": 3.9242,
+      "step": 1666560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2704952884733187e-05,
+      "loss": 3.9263,
+      "step": 1667072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2696566937222667e-05,
+      "loss": 3.9191,
+      "step": 1667584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2688180989712147e-05,
+      "loss": 3.922,
+      "step": 1668096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2679795042201627e-05,
+      "loss": 3.9209,
+      "step": 1668608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.267142547349484e-05,
+      "loss": 3.9291,
+      "step": 1669120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2663039525984316e-05,
+      "loss": 3.9327,
+      "step": 1669632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.26546535784738e-05,
+      "loss": 3.9326,
+      "step": 1670144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.264626763096328e-05,
+      "loss": 3.9137,
+      "step": 1670656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.263788168345276e-05,
+      "loss": 3.9286,
+      "step": 1671168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.262949573594224e-05,
+      "loss": 3.932,
+      "step": 1671680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.262110978843172e-05,
+      "loss": 3.9226,
+      "step": 1672192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2612723840921195e-05,
+      "loss": 3.9241,
+      "step": 1672704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2604354272214408e-05,
+      "loss": 3.9351,
+      "step": 1673216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2595968324703888e-05,
+      "loss": 3.9261,
+      "step": 1673728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.258758237719337e-05,
+      "loss": 3.9444,
+      "step": 1674240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2579196429682848e-05,
+      "loss": 3.9219,
+      "step": 1674752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2570810482172328e-05,
+      "loss": 3.9267,
+      "step": 1675264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2562424534661808e-05,
+      "loss": 3.9181,
+      "step": 1675776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2554038587151288e-05,
+      "loss": 3.9271,
+      "step": 1676288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.2545652639640768e-05,
+      "loss": 3.924,
+      "step": 1676800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.253728307093398e-05,
+      "loss": 3.9182,
+      "step": 1677312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252889712342346e-05,
+      "loss": 3.9239,
+      "step": 1677824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.252052755471667e-05,
+      "loss": 3.9104,
+      "step": 1678336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.251214160720615e-05,
+      "loss": 3.9281,
+      "step": 1678848
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0061869621276855,
+      "eval_runtime": 291.2785,
+      "eval_samples_per_second": 1310.055,
+      "eval_steps_per_second": 40.94,
+      "step": 1679024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.250375565969563e-05,
+      "loss": 3.9279,
+      "step": 1679360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.249536971218511e-05,
+      "loss": 3.9213,
+      "step": 1679872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2486983764674592e-05,
+      "loss": 3.9279,
+      "step": 1680384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2478597817164072e-05,
+      "loss": 3.9267,
+      "step": 1680896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2470211869653552e-05,
+      "loss": 3.9379,
+      "step": 1681408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.246182592214303e-05,
+      "loss": 3.9217,
+      "step": 1681920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.245343997463251e-05,
+      "loss": 3.9258,
+      "step": 1682432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.244505402712199e-05,
+      "loss": 3.9095,
+      "step": 1682944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.243666807961147e-05,
+      "loss": 3.9239,
+      "step": 1683456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.242828213210095e-05,
+      "loss": 3.9267,
+      "step": 1683968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2419896184590432e-05,
+      "loss": 3.9246,
+      "step": 1684480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2411510237079912e-05,
+      "loss": 3.9248,
+      "step": 1684992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.240314066837312e-05,
+      "loss": 3.9174,
+      "step": 1685504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.23947547208626e-05,
+      "loss": 3.9245,
+      "step": 1686016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.238636877335208e-05,
+      "loss": 3.9193,
+      "step": 1686528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2377982825841557e-05,
+      "loss": 3.9112,
+      "step": 1687040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.236959687833104e-05,
+      "loss": 3.9206,
+      "step": 1687552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.236121093082052e-05,
+      "loss": 3.9202,
+      "step": 1688064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.235282498331e-05,
+      "loss": 3.9238,
+      "step": 1688576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.234443903579948e-05,
+      "loss": 3.9346,
+      "step": 1689088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.233605308828896e-05,
+      "loss": 3.927,
+      "step": 1689600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.232766714077844e-05,
+      "loss": 3.931,
+      "step": 1690112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.231929757207165e-05,
+      "loss": 3.9189,
+      "step": 1690624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2310911624561133e-05,
+      "loss": 3.9293,
+      "step": 1691136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2302525677050613e-05,
+      "loss": 3.9235,
+      "step": 1691648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.229413972954009e-05,
+      "loss": 3.9313,
+      "step": 1692160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.228575378202957e-05,
+      "loss": 3.9157,
+      "step": 1692672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2277384213322782e-05,
+      "loss": 3.9199,
+      "step": 1693184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2268998265812262e-05,
+      "loss": 3.9191,
+      "step": 1693696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.2260612318301742e-05,
+      "loss": 3.922,
+      "step": 1694208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2252226370791222e-05,
+      "loss": 3.9231,
+      "step": 1694720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2243840423280702e-05,
+      "loss": 3.9305,
+      "step": 1695232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2235454475770182e-05,
+      "loss": 3.9293,
+      "step": 1695744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2227068528259662e-05,
+      "loss": 3.9266,
+      "step": 1696256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.221868258074914e-05,
+      "loss": 3.9256,
+      "step": 1696768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2210313012042354e-05,
+      "loss": 3.9261,
+      "step": 1697280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2201943443335563e-05,
+      "loss": 3.9259,
+      "step": 1697792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2193557495825043e-05,
+      "loss": 3.9154,
+      "step": 1698304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2185171548314523e-05,
+      "loss": 3.9134,
+      "step": 1698816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2176785600804003e-05,
+      "loss": 3.9115,
+      "step": 1699328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2168399653293483e-05,
+      "loss": 3.9166,
+      "step": 1699840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2160013705782963e-05,
+      "loss": 3.9283,
+      "step": 1700352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2151627758272443e-05,
+      "loss": 3.9224,
+      "step": 1700864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2143258189565655e-05,
+      "loss": 3.9192,
+      "step": 1701376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2134872242055135e-05,
+      "loss": 3.9212,
+      "step": 1701888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2126486294544615e-05,
+      "loss": 3.9224,
+      "step": 1702400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2118100347034095e-05,
+      "loss": 3.9226,
+      "step": 1702912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2109714399523572e-05,
+      "loss": 3.9134,
+      "step": 1703424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2101328452013055e-05,
+      "loss": 3.8954,
+      "step": 1703936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2092942504502535e-05,
+      "loss": 3.9427,
+      "step": 1704448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2084556556992015e-05,
+      "loss": 3.9217,
+      "step": 1704960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2076186988285224e-05,
+      "loss": 3.9315,
+      "step": 1705472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2067801040774704e-05,
+      "loss": 3.9178,
+      "step": 1705984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2059415093264184e-05,
+      "loss": 3.9208,
+      "step": 1706496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2051029145753664e-05,
+      "loss": 3.9156,
+      "step": 1707008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2042659577046877e-05,
+      "loss": 3.9169,
+      "step": 1707520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.203429000834009e-05,
+      "loss": 3.9213,
+      "step": 1708032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.202590406082957e-05,
+      "loss": 3.9189,
+      "step": 1708544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2017518113319046e-05,
+      "loss": 3.9401,
+      "step": 1709056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2009132165808526e-05,
+      "loss": 3.9122,
+      "step": 1709568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.200076259710174e-05,
+      "loss": 3.9082,
+      "step": 1710080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1992376649591218e-05,
+      "loss": 3.9349,
+      "step": 1710592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1983990702080698e-05,
+      "loss": 3.9003,
+      "step": 1711104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1975604754570178e-05,
+      "loss": 3.9082,
+      "step": 1711616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1967218807059658e-05,
+      "loss": 3.9286,
+      "step": 1712128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1958832859549138e-05,
+      "loss": 3.926,
+      "step": 1712640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1950446912038618e-05,
+      "loss": 3.9131,
+      "step": 1713152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.194207734333183e-05,
+      "loss": 3.9181,
+      "step": 1713664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.193369139582131e-05,
+      "loss": 3.8988,
+      "step": 1714176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.192530544831079e-05,
+      "loss": 3.9157,
+      "step": 1714688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1916935879604e-05,
+      "loss": 3.922,
+      "step": 1715200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.190854993209348e-05,
+      "loss": 3.9118,
+      "step": 1715712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1900163984582963e-05,
+      "loss": 3.9191,
+      "step": 1716224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1891778037072443e-05,
+      "loss": 3.9234,
+      "step": 1716736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1883392089561922e-05,
+      "loss": 3.9362,
+      "step": 1717248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18750061420514e-05,
+      "loss": 3.9132,
+      "step": 1717760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.186663657334461e-05,
+      "loss": 3.9141,
+      "step": 1718272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.185825062583409e-05,
+      "loss": 3.9282,
+      "step": 1718784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.184986467832357e-05,
+      "loss": 3.9111,
+      "step": 1719296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.184147873081305e-05,
+      "loss": 3.9228,
+      "step": 1719808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.183309278330253e-05,
+      "loss": 3.9208,
+      "step": 1720320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.182470683579201e-05,
+      "loss": 3.9173,
+      "step": 1720832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1816337267085224e-05,
+      "loss": 3.9288,
+      "step": 1721344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.18079513195747e-05,
+      "loss": 3.9061,
+      "step": 1721856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.179956537206418e-05,
+      "loss": 3.9101,
+      "step": 1722368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1791179424553664e-05,
+      "loss": 3.9195,
+      "step": 1722880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1782793477043144e-05,
+      "loss": 3.9247,
+      "step": 1723392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1774407529532624e-05,
+      "loss": 3.9131,
+      "step": 1723904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1766037960825833e-05,
+      "loss": 3.9185,
+      "step": 1724416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1757652013315313e-05,
+      "loss": 3.9057,
+      "step": 1724928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1749266065804793e-05,
+      "loss": 3.9141,
+      "step": 1725440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1740880118294273e-05,
+      "loss": 3.9068,
+      "step": 1725952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1732494170783756e-05,
+      "loss": 3.9176,
+      "step": 1726464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1724108223273232e-05,
+      "loss": 3.9331,
+      "step": 1726976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1715722275762712e-05,
+      "loss": 3.9193,
+      "step": 1727488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1707336328252192e-05,
+      "loss": 3.9205,
+      "step": 1728000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1698950380741672e-05,
+      "loss": 3.9124,
+      "step": 1728512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1690564433231152e-05,
+      "loss": 3.9125,
+      "step": 1729024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1682178485720632e-05,
+      "loss": 3.9155,
+      "step": 1729536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1673808917013845e-05,
+      "loss": 3.9253,
+      "step": 1730048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1665422969503325e-05,
+      "loss": 3.9218,
+      "step": 1730560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1657037021992805e-05,
+      "loss": 3.9234,
+      "step": 1731072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1648651074482285e-05,
+      "loss": 3.912,
+      "step": 1731584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.164026512697176e-05,
+      "loss": 3.9277,
+      "step": 1732096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.163187917946124e-05,
+      "loss": 3.9089,
+      "step": 1732608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.162349323195072e-05,
+      "loss": 3.9197,
+      "step": 1733120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1615123663243937e-05,
+      "loss": 3.9044,
+      "step": 1733632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1606737715733414e-05,
+      "loss": 3.9222,
+      "step": 1734144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1598368147026626e-05,
+      "loss": 3.9089,
+      "step": 1734656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1589982199516106e-05,
+      "loss": 3.9153,
+      "step": 1735168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1581596252005586e-05,
+      "loss": 3.9108,
+      "step": 1735680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1573210304495066e-05,
+      "loss": 3.9165,
+      "step": 1736192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1564824356984546e-05,
+      "loss": 3.9131,
+      "step": 1736704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1556438409474026e-05,
+      "loss": 3.9133,
+      "step": 1737216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1548052461963506e-05,
+      "loss": 3.9246,
+      "step": 1737728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1539666514452986e-05,
+      "loss": 3.9316,
+      "step": 1738240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1531280566942466e-05,
+      "loss": 3.9227,
+      "step": 1738752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1522894619431942e-05,
+      "loss": 3.9209,
+      "step": 1739264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1514508671921426e-05,
+      "loss": 3.9122,
+      "step": 1739776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1506139103214638e-05,
+      "loss": 3.9117,
+      "step": 1740288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1497753155704118e-05,
+      "loss": 3.9177,
+      "step": 1740800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1489383586997327e-05,
+      "loss": 3.9233,
+      "step": 1741312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1480997639486807e-05,
+      "loss": 3.9006,
+      "step": 1741824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1472611691976287e-05,
+      "loss": 3.9225,
+      "step": 1742336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1464225744465767e-05,
+      "loss": 3.9206,
+      "step": 1742848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1455839796955247e-05,
+      "loss": 3.9171,
+      "step": 1743360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1447453849444727e-05,
+      "loss": 3.9136,
+      "step": 1743872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1439067901934207e-05,
+      "loss": 3.9188,
+      "step": 1744384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1430681954423687e-05,
+      "loss": 3.9135,
+      "step": 1744896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1422296006913167e-05,
+      "loss": 3.9256,
+      "step": 1745408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1413910059402647e-05,
+      "loss": 3.9224,
+      "step": 1745920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1405524111892127e-05,
+      "loss": 3.9223,
+      "step": 1746432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1397138164381607e-05,
+      "loss": 3.9107,
+      "step": 1746944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.138876859567482e-05,
+      "loss": 3.9239,
+      "step": 1747456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.13803826481643e-05,
+      "loss": 3.9259,
+      "step": 1747968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1371996700653776e-05,
+      "loss": 3.9123,
+      "step": 1748480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1363627131946988e-05,
+      "loss": 3.9208,
+      "step": 1748992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1355241184436468e-05,
+      "loss": 3.9255,
+      "step": 1749504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1346855236925948e-05,
+      "loss": 3.9211,
+      "step": 1750016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1338469289415428e-05,
+      "loss": 3.9368,
+      "step": 1750528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1330083341904908e-05,
+      "loss": 3.9127,
+      "step": 1751040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1321697394394388e-05,
+      "loss": 3.9255,
+      "step": 1751552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1313311446883868e-05,
+      "loss": 3.9172,
+      "step": 1752064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.1304925499373348e-05,
+      "loss": 3.9208,
+      "step": 1752576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.129655593066656e-05,
+      "loss": 3.9139,
+      "step": 1753088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.128816998315604e-05,
+      "loss": 3.9123,
+      "step": 1753600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127978403564552e-05,
+      "loss": 3.9164,
+      "step": 1754112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.127141446693873e-05,
+      "loss": 3.9062,
+      "step": 1754624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.126302851942821e-05,
+      "loss": 3.9248,
+      "step": 1755136
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.003749370574951,
+      "eval_runtime": 288.6213,
+      "eval_samples_per_second": 1322.117,
+      "eval_steps_per_second": 41.317,
+      "step": 1755344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.125464257191769e-05,
+      "loss": 3.9207,
+      "step": 1755648
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1246256624407173e-05,
+      "loss": 3.9132,
+      "step": 1756160
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1237870676896652e-05,
+      "loss": 3.9235,
+      "step": 1756672
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.122948472938613e-05,
+      "loss": 3.9193,
+      "step": 1757184
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.122109878187561e-05,
+      "loss": 3.9271,
+      "step": 1757696
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.121272921316882e-05,
+      "loss": 3.9182,
+      "step": 1758208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.12043432656583e-05,
+      "loss": 3.9192,
+      "step": 1758720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.119595731814778e-05,
+      "loss": 3.9106,
+      "step": 1759232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.118757137063726e-05,
+      "loss": 3.9153,
+      "step": 1759744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.117918542312674e-05,
+      "loss": 3.9215,
+      "step": 1760256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.117079947561622e-05,
+      "loss": 3.9147,
+      "step": 1760768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.11624135281057e-05,
+      "loss": 3.9154,
+      "step": 1761280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.115404395939891e-05,
+      "loss": 3.9193,
+      "step": 1761792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.114565801188839e-05,
+      "loss": 3.9119,
+      "step": 1762304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1137272064377874e-05,
+      "loss": 3.9088,
+      "step": 1762816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1128886116867354e-05,
+      "loss": 3.9075,
+      "step": 1763328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1120500169356834e-05,
+      "loss": 3.9095,
+      "step": 1763840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.111211422184631e-05,
+      "loss": 3.9198,
+      "step": 1764352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.110372827433579e-05,
+      "loss": 3.9117,
+      "step": 1764864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.109534232682527e-05,
+      "loss": 3.9326,
+      "step": 1765376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1086972758118482e-05,
+      "loss": 3.9219,
+      "step": 1765888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1078586810607962e-05,
+      "loss": 3.9271,
+      "step": 1766400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1070200863097442e-05,
+      "loss": 3.9117,
+      "step": 1766912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1061814915586922e-05,
+      "loss": 3.9189,
+      "step": 1767424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1053428968076402e-05,
+      "loss": 3.9183,
+      "step": 1767936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1045043020565882e-05,
+      "loss": 3.9227,
+      "step": 1768448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1036657073055362e-05,
+      "loss": 3.9127,
+      "step": 1768960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1028271125544842e-05,
+      "loss": 3.9159,
+      "step": 1769472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1019885178034322e-05,
+      "loss": 3.9104,
+      "step": 1769984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1011515609327535e-05,
+      "loss": 3.914,
+      "step": 1770496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.1003129661817015e-05,
+      "loss": 3.9144,
+      "step": 1771008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.099474371430649e-05,
+      "loss": 3.9262,
+      "step": 1771520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.098635776679597e-05,
+      "loss": 3.9239,
+      "step": 1772032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.097797181928545e-05,
+      "loss": 3.9224,
+      "step": 1772544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0969585871774934e-05,
+      "loss": 3.916,
+      "step": 1773056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0961216303068143e-05,
+      "loss": 3.9175,
+      "step": 1773568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0952830355557623e-05,
+      "loss": 3.926,
+      "step": 1774080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0944444408047103e-05,
+      "loss": 3.9101,
+      "step": 1774592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0936058460536583e-05,
+      "loss": 3.9075,
+      "step": 1775104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0927672513026063e-05,
+      "loss": 3.9024,
+      "step": 1775616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0919286565515543e-05,
+      "loss": 3.9117,
+      "step": 1776128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0910916996808756e-05,
+      "loss": 3.9174,
+      "step": 1776640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0902531049298236e-05,
+      "loss": 3.9162,
+      "step": 1777152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0894145101787716e-05,
+      "loss": 3.9137,
+      "step": 1777664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0885759154277196e-05,
+      "loss": 3.913,
+      "step": 1778176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0877389585570405e-05,
+      "loss": 3.9149,
+      "step": 1778688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0869003638059885e-05,
+      "loss": 3.9204,
+      "step": 1779200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0860617690549368e-05,
+      "loss": 3.9057,
+      "step": 1779712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0852231743038848e-05,
+      "loss": 3.8889,
+      "step": 1780224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0843845795528325e-05,
+      "loss": 3.9299,
+      "step": 1780736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0835459848017804e-05,
+      "loss": 3.9219,
+      "step": 1781248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0827090279311017e-05,
+      "loss": 3.9242,
+      "step": 1781760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0818704331800497e-05,
+      "loss": 3.91,
+      "step": 1782272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0810318384289977e-05,
+      "loss": 3.9165,
+      "step": 1782784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0801932436779457e-05,
+      "loss": 3.9085,
+      "step": 1783296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0793546489268937e-05,
+      "loss": 3.9098,
+      "step": 1783808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0785176920562146e-05,
+      "loss": 3.916,
+      "step": 1784320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0776790973051626e-05,
+      "loss": 3.908,
+      "step": 1784832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0768405025541106e-05,
+      "loss": 3.9345,
+      "step": 1785344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.076001907803059e-05,
+      "loss": 3.9113,
+      "step": 1785856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.075163313052007e-05,
+      "loss": 3.8977,
+      "step": 1786368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.074324718300955e-05,
+      "loss": 3.9271,
+      "step": 1786880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0734877614302758e-05,
+      "loss": 3.8976,
+      "step": 1787392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0726491666792238e-05,
+      "loss": 3.8974,
+      "step": 1787904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0718105719281718e-05,
+      "loss": 3.9251,
+      "step": 1788416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0709719771771198e-05,
+      "loss": 3.9187,
+      "step": 1788928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0701333824260678e-05,
+      "loss": 3.9048,
+      "step": 1789440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0692947876750158e-05,
+      "loss": 3.9121,
+      "step": 1789952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0684561929239638e-05,
+      "loss": 3.8948,
+      "step": 1790464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0676175981729118e-05,
+      "loss": 3.9081,
+      "step": 1790976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0667806413022327e-05,
+      "loss": 3.9125,
+      "step": 1791488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0659436844315543e-05,
+      "loss": 3.9068,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0651050896805023e-05,
+      "loss": 3.9166,
+      "step": 1792512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.06426649492945e-05,
+      "loss": 3.9144,
+      "step": 1793024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.063427900178398e-05,
+      "loss": 3.9275,
+      "step": 1793536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0625909433077192e-05,
+      "loss": 3.9074,
+      "step": 1794048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0617523485566672e-05,
+      "loss": 3.9115,
+      "step": 1794560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060913753805615e-05,
+      "loss": 3.9196,
+      "step": 1795072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.060075159054563e-05,
+      "loss": 3.9058,
+      "step": 1795584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0592382021838844e-05,
+      "loss": 3.9182,
+      "step": 1796096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0583996074328324e-05,
+      "loss": 3.9115,
+      "step": 1796608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.05756101268178e-05,
+      "loss": 3.909,
+      "step": 1797120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.056722417930728e-05,
+      "loss": 3.9262,
+      "step": 1797632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.055883823179676e-05,
+      "loss": 3.8966,
+      "step": 1798144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0550452284286244e-05,
+      "loss": 3.9018,
+      "step": 1798656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0542066336775724e-05,
+      "loss": 3.9166,
+      "step": 1799168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0533696768068933e-05,
+      "loss": 3.9175,
+      "step": 1799680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0525310820558413e-05,
+      "loss": 3.9083,
+      "step": 1800192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.0516941251851622e-05,
+      "loss": 3.917,
+      "step": 1800704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0508555304341105e-05,
+      "loss": 3.8947,
+      "step": 1801216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0500169356830585e-05,
+      "loss": 3.9086,
+      "step": 1801728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0491783409320065e-05,
+      "loss": 3.9058,
+      "step": 1802240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0483413840613274e-05,
+      "loss": 3.9108,
+      "step": 1802752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0475027893102754e-05,
+      "loss": 3.9224,
+      "step": 1803264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0466641945592234e-05,
+      "loss": 3.9149,
+      "step": 1803776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0458255998081714e-05,
+      "loss": 3.9159,
+      "step": 1804288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0449870050571198e-05,
+      "loss": 3.9054,
+      "step": 1804800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0441500481864407e-05,
+      "loss": 3.9046,
+      "step": 1805312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0433114534353887e-05,
+      "loss": 3.9091,
+      "step": 1805824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0424728586843367e-05,
+      "loss": 3.9176,
+      "step": 1806336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0416342639332847e-05,
+      "loss": 3.9153,
+      "step": 1806848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.040797307062606e-05,
+      "loss": 3.9206,
+      "step": 1807360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.039958712311554e-05,
+      "loss": 3.9038,
+      "step": 1807872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.039120117560502e-05,
+      "loss": 3.9179,
+      "step": 1808384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.03828152280945e-05,
+      "loss": 3.908,
+      "step": 1808896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.037442928058398e-05,
+      "loss": 3.9122,
+      "step": 1809408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0366043333073455e-05,
+      "loss": 3.8983,
+      "step": 1809920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0357657385562935e-05,
+      "loss": 3.9161,
+      "step": 1810432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0349271438052415e-05,
+      "loss": 3.9022,
+      "step": 1810944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0340901869345628e-05,
+      "loss": 3.9109,
+      "step": 1811456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0332515921835108e-05,
+      "loss": 3.9023,
+      "step": 1811968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0324129974324588e-05,
+      "loss": 3.9138,
+      "step": 1812480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0315744026814068e-05,
+      "loss": 3.9054,
+      "step": 1812992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0307374458107277e-05,
+      "loss": 3.9046,
+      "step": 1813504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.029898851059676e-05,
+      "loss": 3.9198,
+      "step": 1814016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.029060256308624e-05,
+      "loss": 3.9242,
+      "step": 1814528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.028221661557572e-05,
+      "loss": 3.9199,
+      "step": 1815040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.02738306680652e-05,
+      "loss": 3.9154,
+      "step": 1815552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.026544472055468e-05,
+      "loss": 3.9055,
+      "step": 1816064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.025707515184789e-05,
+      "loss": 3.9053,
+      "step": 1816576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.024868920433737e-05,
+      "loss": 3.9133,
+      "step": 1817088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0240303256826852e-05,
+      "loss": 3.9137,
+      "step": 1817600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0231917309316332e-05,
+      "loss": 3.8996,
+      "step": 1818112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.022353136180581e-05,
+      "loss": 3.9178,
+      "step": 1818624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.021514541429529e-05,
+      "loss": 3.918,
+      "step": 1819136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.02067758455885e-05,
+      "loss": 3.9113,
+      "step": 1819648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.019838989807798e-05,
+      "loss": 3.9072,
+      "step": 1820160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.019000395056746e-05,
+      "loss": 3.9088,
+      "step": 1820672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.018161800305694e-05,
+      "loss": 3.9043,
+      "step": 1821184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.017323205554642e-05,
+      "loss": 3.9213,
+      "step": 1821696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.01648461080359e-05,
+      "loss": 3.9181,
+      "step": 1822208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.015646016052538e-05,
+      "loss": 3.9162,
+      "step": 1822720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.014807421301486e-05,
+      "loss": 3.8988,
+      "step": 1823232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.013970464430807e-05,
+      "loss": 3.9197,
+      "step": 1823744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0131318696797553e-05,
+      "loss": 3.9193,
+      "step": 1824256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0122932749287033e-05,
+      "loss": 3.9083,
+      "step": 1824768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0114563180580242e-05,
+      "loss": 3.9185,
+      "step": 1825280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0106177233069722e-05,
+      "loss": 3.9113,
+      "step": 1825792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0097791285559202e-05,
+      "loss": 3.9181,
+      "step": 1826304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0089405338048682e-05,
+      "loss": 3.9292,
+      "step": 1826816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0081019390538162e-05,
+      "loss": 3.9095,
+      "step": 1827328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0072649821831375e-05,
+      "loss": 3.9158,
+      "step": 1827840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0064280253124584e-05,
+      "loss": 3.9176,
+      "step": 1828352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0055894305614064e-05,
+      "loss": 3.9169,
+      "step": 1828864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0047508358103544e-05,
+      "loss": 3.9041,
+      "step": 1829376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0039122410593024e-05,
+      "loss": 3.9044,
+      "step": 1829888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0030736463082507e-05,
+      "loss": 3.9111,
+      "step": 1830400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0022366894375716e-05,
+      "loss": 3.9015,
+      "step": 1830912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.0013980946865196e-05,
+      "loss": 3.9168,
+      "step": 1831424
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.0009846687316895,
+      "eval_runtime": 287.847,
+      "eval_samples_per_second": 1325.673,
+      "eval_steps_per_second": 41.428,
+      "step": 1831664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0005594999354676e-05,
+      "loss": 3.9165,
+      "step": 1831936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9997209051844156e-05,
+      "loss": 3.9069,
+      "step": 1832448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9988823104333636e-05,
+      "loss": 3.9118,
+      "step": 1832960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9980437156823116e-05,
+      "loss": 3.9132,
+      "step": 1833472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9972051209312596e-05,
+      "loss": 3.9236,
+      "step": 1833984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9963665261802076e-05,
+      "loss": 3.9107,
+      "step": 1834496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9955295693095285e-05,
+      "loss": 3.9176,
+      "step": 1835008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9946909745584765e-05,
+      "loss": 3.9031,
+      "step": 1835520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9938523798074245e-05,
+      "loss": 3.9055,
+      "step": 1836032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9930137850563725e-05,
+      "loss": 3.919,
+      "step": 1836544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9921751903053208e-05,
+      "loss": 3.9093,
+      "step": 1837056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9913365955542688e-05,
+      "loss": 3.9088,
+      "step": 1837568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9904996386835897e-05,
+      "loss": 3.9164,
+      "step": 1838080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9896610439325377e-05,
+      "loss": 3.9021,
+      "step": 1838592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9888224491814857e-05,
+      "loss": 3.9036,
+      "step": 1839104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9879838544304337e-05,
+      "loss": 3.8999,
+      "step": 1839616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9871452596793817e-05,
+      "loss": 3.9074,
+      "step": 1840128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9863066649283297e-05,
+      "loss": 3.9128,
+      "step": 1840640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9854680701772777e-05,
+      "loss": 3.9058,
+      "step": 1841152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9846294754262257e-05,
+      "loss": 3.9212,
+      "step": 1841664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9837908806751737e-05,
+      "loss": 3.9179,
+      "step": 1842176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9829522859241217e-05,
+      "loss": 3.9179,
+      "step": 1842688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9821136911730697e-05,
+      "loss": 3.9086,
+      "step": 1843200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9812750964220177e-05,
+      "loss": 3.9113,
+      "step": 1843712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9804365016709657e-05,
+      "loss": 3.9125,
+      "step": 1844224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9795979069199137e-05,
+      "loss": 3.9164,
+      "step": 1844736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.978760950049235e-05,
+      "loss": 3.9061,
+      "step": 1845248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9779223552981826e-05,
+      "loss": 3.9105,
+      "step": 1845760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9770837605471306e-05,
+      "loss": 3.9008,
+      "step": 1846272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.9762451657960786e-05,
+      "loss": 3.9133,
+      "step": 1846784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.975406571045027e-05,
+      "loss": 3.9061,
+      "step": 1847296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.974567976293975e-05,
+      "loss": 3.9177,
+      "step": 1847808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.973729381542923e-05,
+      "loss": 3.918,
+      "step": 1848320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9728924246722438e-05,
+      "loss": 3.9203,
+      "step": 1848832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9720538299211918e-05,
+      "loss": 3.9073,
+      "step": 1849344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.971216873050513e-05,
+      "loss": 3.913,
+      "step": 1849856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.970378278299461e-05,
+      "loss": 3.9186,
+      "step": 1850368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.969539683548409e-05,
+      "loss": 3.9024,
+      "step": 1850880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.968701088797357e-05,
+      "loss": 3.9015,
+      "step": 1851392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.967862494046305e-05,
+      "loss": 3.8988,
+      "step": 1851904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.967023899295253e-05,
+      "loss": 3.9003,
+      "step": 1852416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9661853045442007e-05,
+      "loss": 3.9084,
+      "step": 1852928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9653467097931487e-05,
+      "loss": 3.9132,
+      "step": 1853440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.964508115042097e-05,
+      "loss": 3.9086,
+      "step": 1853952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.963669520291045e-05,
+      "loss": 3.9067,
+      "step": 1854464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.962830925539993e-05,
+      "loss": 3.9093,
+      "step": 1854976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.961992330788941e-05,
+      "loss": 3.9147,
+      "step": 1855488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.961153736037889e-05,
+      "loss": 3.8986,
+      "step": 1856000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9603151412868366e-05,
+      "loss": 3.8849,
+      "step": 1856512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9594765465357846e-05,
+      "loss": 3.9228,
+      "step": 1857024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9586395896651062e-05,
+      "loss": 3.9149,
+      "step": 1857536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.957800994914054e-05,
+      "loss": 3.9176,
+      "step": 1858048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.956962400163002e-05,
+      "loss": 3.9064,
+      "step": 1858560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.95612380541195e-05,
+      "loss": 3.9078,
+      "step": 1859072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.955285210660898e-05,
+      "loss": 3.9073,
+      "step": 1859584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.954446615909846e-05,
+      "loss": 3.9015,
+      "step": 1860096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.953608021158794e-05,
+      "loss": 3.9141,
+      "step": 1860608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.952769426407742e-05,
+      "loss": 3.8968,
+      "step": 1861120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951932469537063e-05,
+      "loss": 3.9282,
+      "step": 1861632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951095512666384e-05,
+      "loss": 3.9062,
+      "step": 1862144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.950256917915332e-05,
+      "loss": 3.8905,
+      "step": 1862656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.94941832316428e-05,
+      "loss": 3.9169,
+      "step": 1863168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.948579728413228e-05,
+      "loss": 3.8979,
+      "step": 1863680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9477427715425493e-05,
+      "loss": 3.8914,
+      "step": 1864192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9469041767914972e-05,
+      "loss": 3.9167,
+      "step": 1864704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9460688578011914e-05,
+      "loss": 3.9127,
+      "step": 1865216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9452302630501394e-05,
+      "loss": 3.9021,
+      "step": 1865728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9443916682990877e-05,
+      "loss": 3.903,
+      "step": 1866240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9435530735480357e-05,
+      "loss": 3.889,
+      "step": 1866752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9427144787969834e-05,
+      "loss": 3.9013,
+      "step": 1867264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9418758840459314e-05,
+      "loss": 3.909,
+      "step": 1867776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9410372892948794e-05,
+      "loss": 3.9027,
+      "step": 1868288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9401986945438274e-05,
+      "loss": 3.9095,
+      "step": 1868800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9393600997927754e-05,
+      "loss": 3.9058,
+      "step": 1869312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9385215050417234e-05,
+      "loss": 3.9212,
+      "step": 1869824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9376829102906717e-05,
+      "loss": 3.9042,
+      "step": 1870336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9368459534199926e-05,
+      "loss": 3.9086,
+      "step": 1870848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9360073586689406e-05,
+      "loss": 3.9082,
+      "step": 1871360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9351687639178886e-05,
+      "loss": 3.9016,
+      "step": 1871872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9343301691668366e-05,
+      "loss": 3.9104,
+      "step": 1872384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9334915744157846e-05,
+      "loss": 3.9047,
+      "step": 1872896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9326529796647326e-05,
+      "loss": 3.9052,
+      "step": 1873408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9318143849136806e-05,
+      "loss": 3.9197,
+      "step": 1873920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9309757901626286e-05,
+      "loss": 3.892,
+      "step": 1874432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9301371954115766e-05,
+      "loss": 3.8986,
+      "step": 1874944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9292986006605246e-05,
+      "loss": 3.9098,
+      "step": 1875456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9284600059094722e-05,
+      "loss": 3.9092,
+      "step": 1875968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9276230490387938e-05,
+      "loss": 3.9014,
+      "step": 1876480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9267860921681147e-05,
+      "loss": 3.9122,
+      "step": 1876992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.925949135297436e-05,
+      "loss": 3.8945,
+      "step": 1877504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.925110540546384e-05,
+      "loss": 3.9002,
+      "step": 1878016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9242719457953316e-05,
+      "loss": 3.8973,
+      "step": 1878528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.92343335104428e-05,
+      "loss": 3.9004,
+      "step": 1879040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.922594756293228e-05,
+      "loss": 3.9219,
+      "step": 1879552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.921757799422549e-05,
+      "loss": 3.9057,
+      "step": 1880064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.920919204671497e-05,
+      "loss": 3.9102,
+      "step": 1880576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.920080609920445e-05,
+      "loss": 3.9009,
+      "step": 1881088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.919242015169393e-05,
+      "loss": 3.9002,
+      "step": 1881600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.918403420418341e-05,
+      "loss": 3.9029,
+      "step": 1882112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.917564825667289e-05,
+      "loss": 3.9125,
+      "step": 1882624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9167262309162372e-05,
+      "loss": 3.9074,
+      "step": 1883136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.915887636165185e-05,
+      "loss": 3.911,
+      "step": 1883648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.915049041414133e-05,
+      "loss": 3.9042,
+      "step": 1884160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9142104466630808e-05,
+      "loss": 3.9067,
+      "step": 1884672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.913373489792402e-05,
+      "loss": 3.9031,
+      "step": 1885184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.91253489504135e-05,
+      "loss": 3.9045,
+      "step": 1885696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.911696300290298e-05,
+      "loss": 3.8913,
+      "step": 1886208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.910857705539246e-05,
+      "loss": 3.9121,
+      "step": 1886720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.910020748668567e-05,
+      "loss": 3.896,
+      "step": 1887232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.909182153917515e-05,
+      "loss": 3.9041,
+      "step": 1887744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.908343559166463e-05,
+      "loss": 3.8979,
+      "step": 1888256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.907504964415411e-05,
+      "loss": 3.9053,
+      "step": 1888768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9066663696643593e-05,
+      "loss": 3.9037,
+      "step": 1889280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9058294127936802e-05,
+      "loss": 3.8977,
+      "step": 1889792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9049908180426282e-05,
+      "loss": 3.9152,
+      "step": 1890304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9041522232915762e-05,
+      "loss": 3.9199,
+      "step": 1890816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9033136285405242e-05,
+      "loss": 3.9141,
+      "step": 1891328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9024750337894722e-05,
+      "loss": 3.9045,
+      "step": 1891840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9016380769187934e-05,
+      "loss": 3.9007,
+      "step": 1892352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.9007994821677414e-05,
+      "loss": 3.9,
+      "step": 1892864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8999608874166894e-05,
+      "loss": 3.9088,
+      "step": 1893376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8991222926656374e-05,
+      "loss": 3.9095,
+      "step": 1893888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.898283697914585e-05,
+      "loss": 3.8953,
+      "step": 1894400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.897445103163533e-05,
+      "loss": 3.9071,
+      "step": 1894912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8966081462928547e-05,
+      "loss": 3.9098,
+      "step": 1895424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8957695515418027e-05,
+      "loss": 3.9052,
+      "step": 1895936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8949309567907503e-05,
+      "loss": 3.9024,
+      "step": 1896448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8940923620396983e-05,
+      "loss": 3.9074,
+      "step": 1896960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8932537672886463e-05,
+      "loss": 3.8976,
+      "step": 1897472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8924168104179676e-05,
+      "loss": 3.9139,
+      "step": 1897984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8915782156669155e-05,
+      "loss": 3.9088,
+      "step": 1898496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8907396209158635e-05,
+      "loss": 3.9153,
+      "step": 1899008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8899026640451848e-05,
+      "loss": 3.8958,
+      "step": 1899520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8890640692941324e-05,
+      "loss": 3.9083,
+      "step": 1900032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8882254745430804e-05,
+      "loss": 3.9151,
+      "step": 1900544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8873868797920284e-05,
+      "loss": 3.9067,
+      "step": 1901056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8865482850409764e-05,
+      "loss": 3.9098,
+      "step": 1901568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8857096902899248e-05,
+      "loss": 3.9082,
+      "step": 1902080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8848710955388728e-05,
+      "loss": 3.9096,
+      "step": 1902592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8840325007878208e-05,
+      "loss": 3.9268,
+      "step": 1903104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8831939060367684e-05,
+      "loss": 3.9023,
+      "step": 1903616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8823553112857164e-05,
+      "loss": 3.9124,
+      "step": 1904128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8815167165346644e-05,
+      "loss": 3.9107,
+      "step": 1904640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8806797596639857e-05,
+      "loss": 3.9055,
+      "step": 1905152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8798411649129337e-05,
+      "loss": 3.9022,
+      "step": 1905664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.879004208042255e-05,
+      "loss": 3.8962,
+      "step": 1906176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.878165613291203e-05,
+      "loss": 3.9049,
+      "step": 1906688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8773270185401506e-05,
+      "loss": 3.9008,
+      "step": 1907200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.8764884237890985e-05,
+      "loss": 3.9081,
+      "step": 1907712
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9988596439361572,
+      "eval_runtime": 297.816,
+      "eval_samples_per_second": 1281.298,
+      "eval_steps_per_second": 40.041,
+      "step": 1907984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8756498290380465e-05,
+      "loss": 3.905,
+      "step": 1908224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.874812872167368e-05,
+      "loss": 3.9039,
+      "step": 1908736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8739742774163158e-05,
+      "loss": 3.9064,
+      "step": 1909248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.873137320545637e-05,
+      "loss": 3.9079,
+      "step": 1909760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.872298725794585e-05,
+      "loss": 3.9192,
+      "step": 1910272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8714601310435327e-05,
+      "loss": 3.9102,
+      "step": 1910784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.870621536292481e-05,
+      "loss": 3.9091,
+      "step": 1911296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.869782941541429e-05,
+      "loss": 3.8971,
+      "step": 1911808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.868944346790377e-05,
+      "loss": 3.9002,
+      "step": 1912320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.868107389919698e-05,
+      "loss": 3.912,
+      "step": 1912832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.867268795168646e-05,
+      "loss": 3.9046,
+      "step": 1913344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.866430200417594e-05,
+      "loss": 3.904,
+      "step": 1913856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8655932435469155e-05,
+      "loss": 3.9106,
+      "step": 1914368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.864754648795863e-05,
+      "loss": 3.8981,
+      "step": 1914880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.863916054044811e-05,
+      "loss": 3.9014,
+      "step": 1915392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.863077459293759e-05,
+      "loss": 3.8885,
+      "step": 1915904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.862238864542707e-05,
+      "loss": 3.9054,
+      "step": 1916416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.861400269791655e-05,
+      "loss": 3.9095,
+      "step": 1916928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.860561675040603e-05,
+      "loss": 3.8966,
+      "step": 1917440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.859723080289551e-05,
+      "loss": 3.9162,
+      "step": 1917952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.858884485538499e-05,
+      "loss": 3.9141,
+      "step": 1918464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.858045890787447e-05,
+      "loss": 3.9127,
+      "step": 1918976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.857207296036395e-05,
+      "loss": 3.9049,
+      "step": 1919488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.856368701285343e-05,
+      "loss": 3.9095,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.855530106534291e-05,
+      "loss": 3.9095,
+      "step": 1920512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.854691511783239e-05,
+      "loss": 3.9076,
+      "step": 1921024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853852917032187e-05,
+      "loss": 3.8984,
+      "step": 1921536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.853014322281135e-05,
+      "loss": 3.9035,
+      "step": 1922048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8521773654104563e-05,
+      "loss": 3.8989,
+      "step": 1922560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.8513404085397773e-05,
+      "loss": 3.904,
+      "step": 1923072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.850503451669098e-05,
+      "loss": 3.906,
+      "step": 1923584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8496648569180465e-05,
+      "loss": 3.9071,
+      "step": 1924096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8488262621669945e-05,
+      "loss": 3.9108,
+      "step": 1924608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8479876674159425e-05,
+      "loss": 3.9182,
+      "step": 1925120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8471490726648905e-05,
+      "loss": 3.9015,
+      "step": 1925632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8463121157942114e-05,
+      "loss": 3.9053,
+      "step": 1926144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8454735210431594e-05,
+      "loss": 3.9155,
+      "step": 1926656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8446349262921074e-05,
+      "loss": 3.8951,
+      "step": 1927168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8437963315410557e-05,
+      "loss": 3.8966,
+      "step": 1927680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8429577367900037e-05,
+      "loss": 3.8967,
+      "step": 1928192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8421191420389517e-05,
+      "loss": 3.8963,
+      "step": 1928704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8412805472878994e-05,
+      "loss": 3.9027,
+      "step": 1929216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8404419525368474e-05,
+      "loss": 3.9127,
+      "step": 1929728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8396033577857954e-05,
+      "loss": 3.8993,
+      "step": 1930240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8387647630347434e-05,
+      "loss": 3.9041,
+      "step": 1930752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8379278061640646e-05,
+      "loss": 3.8998,
+      "step": 1931264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8370892114130126e-05,
+      "loss": 3.911,
+      "step": 1931776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8362506166619606e-05,
+      "loss": 3.8925,
+      "step": 1932288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8354120219109086e-05,
+      "loss": 3.8848,
+      "step": 1932800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8345734271598566e-05,
+      "loss": 3.9147,
+      "step": 1933312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8337348324088046e-05,
+      "loss": 3.9105,
+      "step": 1933824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8328962376577526e-05,
+      "loss": 3.9186,
+      "step": 1934336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8320592807870738e-05,
+      "loss": 3.897,
+      "step": 1934848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8312206860360218e-05,
+      "loss": 3.9023,
+      "step": 1935360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8303820912849698e-05,
+      "loss": 3.902,
+      "step": 1935872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8295434965339175e-05,
+      "loss": 3.8954,
+      "step": 1936384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8287049017828655e-05,
+      "loss": 3.9124,
+      "step": 1936896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8278663070318135e-05,
+      "loss": 3.8904,
+      "step": 1937408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8270277122807618e-05,
+      "loss": 3.9236,
+      "step": 1937920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8261891175297098e-05,
+      "loss": 3.9032,
+      "step": 1938432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8253521606590307e-05,
+      "loss": 3.8825,
+      "step": 1938944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8245135659079787e-05,
+      "loss": 3.9153,
+      "step": 1939456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8236766090372996e-05,
+      "loss": 3.895,
+      "step": 1939968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.822838014286248e-05,
+      "loss": 3.8839,
+      "step": 1940480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.821999419535196e-05,
+      "loss": 3.9099,
+      "step": 1940992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.821160824784144e-05,
+      "loss": 3.9099,
+      "step": 1941504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.820323867913465e-05,
+      "loss": 3.8962,
+      "step": 1942016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.819485273162413e-05,
+      "loss": 3.8998,
+      "step": 1942528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.818646678411361e-05,
+      "loss": 3.8814,
+      "step": 1943040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.817809721540682e-05,
+      "loss": 3.8977,
+      "step": 1943552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.81697112678963e-05,
+      "loss": 3.9009,
+      "step": 1944064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.816132532038578e-05,
+      "loss": 3.8973,
+      "step": 1944576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.815293937287526e-05,
+      "loss": 3.9064,
+      "step": 1945088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.814455342536474e-05,
+      "loss": 3.897,
+      "step": 1945600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.813616747785422e-05,
+      "loss": 3.919,
+      "step": 1946112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.81277815303437e-05,
+      "loss": 3.9011,
+      "step": 1946624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8119411961636913e-05,
+      "loss": 3.9001,
+      "step": 1947136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8111026014126393e-05,
+      "loss": 3.9034,
+      "step": 1947648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8102640066615873e-05,
+      "loss": 3.8991,
+      "step": 1948160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.809425411910535e-05,
+      "loss": 3.9068,
+      "step": 1948672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.808586817159483e-05,
+      "loss": 3.9011,
+      "step": 1949184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.807748222408431e-05,
+      "loss": 3.9015,
+      "step": 1949696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.806909627657379e-05,
+      "loss": 3.9129,
+      "step": 1950208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8060710329063273e-05,
+      "loss": 3.8907,
+      "step": 1950720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8052340760356482e-05,
+      "loss": 3.8913,
+      "step": 1951232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8043954812845962e-05,
+      "loss": 3.9016,
+      "step": 1951744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8035568865335442e-05,
+      "loss": 3.9065,
+      "step": 1952256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8027182917824922e-05,
+      "loss": 3.8935,
+      "step": 1952768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.8018813349118134e-05,
+      "loss": 3.9049,
+      "step": 1953280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8010427401607614e-05,
+      "loss": 3.8916,
+      "step": 1953792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.8002041454097094e-05,
+      "loss": 3.892,
+      "step": 1954304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7993655506586574e-05,
+      "loss": 3.8925,
+      "step": 1954816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7985269559076054e-05,
+      "loss": 3.8982,
+      "step": 1955328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.797688361156553e-05,
+      "loss": 3.9131,
+      "step": 1955840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796849766405501e-05,
+      "loss": 3.9035,
+      "step": 1956352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.796011171654449e-05,
+      "loss": 3.904,
+      "step": 1956864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7951742147837706e-05,
+      "loss": 3.8957,
+      "step": 1957376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7943356200327183e-05,
+      "loss": 3.8933,
+      "step": 1957888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7934986631620395e-05,
+      "loss": 3.8985,
+      "step": 1958400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7926600684109875e-05,
+      "loss": 3.9068,
+      "step": 1958912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7918214736599355e-05,
+      "loss": 3.907,
+      "step": 1959424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7909828789088835e-05,
+      "loss": 3.9044,
+      "step": 1959936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7901459220382048e-05,
+      "loss": 3.9009,
+      "step": 1960448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7893073272871528e-05,
+      "loss": 3.8975,
+      "step": 1960960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7884687325361004e-05,
+      "loss": 3.902,
+      "step": 1961472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7876301377850484e-05,
+      "loss": 3.897,
+      "step": 1961984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7867915430339964e-05,
+      "loss": 3.8917,
+      "step": 1962496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7859529482829444e-05,
+      "loss": 3.8981,
+      "step": 1963008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7851159914122657e-05,
+      "loss": 3.8957,
+      "step": 1963520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7842773966612137e-05,
+      "loss": 3.8952,
+      "step": 1964032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7834388019101617e-05,
+      "loss": 3.8962,
+      "step": 1964544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7826002071591097e-05,
+      "loss": 3.8949,
+      "step": 1965056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7817616124080576e-05,
+      "loss": 3.8976,
+      "step": 1965568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7809230176570056e-05,
+      "loss": 3.8958,
+      "step": 1966080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7800844229059536e-05,
+      "loss": 3.9103,
+      "step": 1966592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.779247466035275e-05,
+      "loss": 3.9103,
+      "step": 1967104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7784105091645958e-05,
+      "loss": 3.9106,
+      "step": 1967616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7775719144135438e-05,
+      "loss": 3.9018,
+      "step": 1968128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7767333196624918e-05,
+      "loss": 3.8964,
+      "step": 1968640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7758947249114398e-05,
+      "loss": 3.8952,
+      "step": 1969152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.775056130160388e-05,
+      "loss": 3.9041,
+      "step": 1969664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.774217535409336e-05,
+      "loss": 3.9067,
+      "step": 1970176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7733789406582838e-05,
+      "loss": 3.8877,
+      "step": 1970688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.772541983787605e-05,
+      "loss": 3.9033,
+      "step": 1971200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.771703389036553e-05,
+      "loss": 3.9018,
+      "step": 1971712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.770864794285501e-05,
+      "loss": 3.9025,
+      "step": 1972224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.770026199534449e-05,
+      "loss": 3.8968,
+      "step": 1972736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.769187604783397e-05,
+      "loss": 3.8982,
+      "step": 1973248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.768349010032345e-05,
+      "loss": 3.8912,
+      "step": 1973760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.767510415281293e-05,
+      "loss": 3.912,
+      "step": 1974272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.766673458410614e-05,
+      "loss": 3.9053,
+      "step": 1974784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.765836501539935e-05,
+      "loss": 3.9113,
+      "step": 1975296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7649979067888835e-05,
+      "loss": 3.8914,
+      "step": 1975808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.764159312037831e-05,
+      "loss": 3.9018,
+      "step": 1976320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.763320717286779e-05,
+      "loss": 3.9121,
+      "step": 1976832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.762482122535727e-05,
+      "loss": 3.8977,
+      "step": 1977344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.761643527784675e-05,
+      "loss": 3.9046,
+      "step": 1977856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.760804933033623e-05,
+      "loss": 3.9027,
+      "step": 1978368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759966338282571e-05,
+      "loss": 3.9076,
+      "step": 1978880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.759127743531519e-05,
+      "loss": 3.9158,
+      "step": 1979392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7582907866608404e-05,
+      "loss": 3.9025,
+      "step": 1979904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7574521919097884e-05,
+      "loss": 3.9059,
+      "step": 1980416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7566135971587364e-05,
+      "loss": 3.9043,
+      "step": 1980928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.755775002407684e-05,
+      "loss": 3.9043,
+      "step": 1981440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.754936407656632e-05,
+      "loss": 3.8938,
+      "step": 1981952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7540978129055803e-05,
+      "loss": 3.892,
+      "step": 1982464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7532592181545283e-05,
+      "loss": 3.897,
+      "step": 1982976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7524238991642225e-05,
+      "loss": 3.897,
+      "step": 1983488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.7515853044131705e-05,
+      "loss": 3.9019,
+      "step": 1984000
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.996859073638916,
+      "eval_runtime": 300.1706,
+      "eval_samples_per_second": 1271.247,
+      "eval_steps_per_second": 39.727,
+      "step": 1984304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7507467096621185e-05,
+      "loss": 3.8996,
+      "step": 1984512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7499081149110665e-05,
+      "loss": 3.8966,
+      "step": 1985024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7490695201600145e-05,
+      "loss": 3.9012,
+      "step": 1985536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7482309254089625e-05,
+      "loss": 3.907,
+      "step": 1986048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7473923306579105e-05,
+      "loss": 3.9126,
+      "step": 1986560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7465537359068585e-05,
+      "loss": 3.9016,
+      "step": 1987072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7457151411558065e-05,
+      "loss": 3.906,
+      "step": 1987584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7448765464047545e-05,
+      "loss": 3.8948,
+      "step": 1988096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.744037951653702e-05,
+      "loss": 3.8939,
+      "step": 1988608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7431993569026505e-05,
+      "loss": 3.9015,
+      "step": 1989120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7423607621515984e-05,
+      "loss": 3.9043,
+      "step": 1989632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7415221674005464e-05,
+      "loss": 3.8984,
+      "step": 1990144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7406868484102406e-05,
+      "loss": 3.9059,
+      "step": 1990656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7398482536591886e-05,
+      "loss": 3.8898,
+      "step": 1991168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7390096589081366e-05,
+      "loss": 3.8941,
+      "step": 1991680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7381710641570846e-05,
+      "loss": 3.881,
+      "step": 1992192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7373324694060326e-05,
+      "loss": 3.9039,
+      "step": 1992704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7364938746549806e-05,
+      "loss": 3.9031,
+      "step": 1993216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7356552799039286e-05,
+      "loss": 3.8924,
+      "step": 1993728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7348166851528766e-05,
+      "loss": 3.9087,
+      "step": 1994240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7339780904018246e-05,
+      "loss": 3.9109,
+      "step": 1994752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7331394956507726e-05,
+      "loss": 3.9037,
+      "step": 1995264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7323009008997206e-05,
+      "loss": 3.9031,
+      "step": 1995776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7314623061486686e-05,
+      "loss": 3.9007,
+      "step": 1996288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7306253492779898e-05,
+      "loss": 3.9036,
+      "step": 1996800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7297867545269378e-05,
+      "loss": 3.9078,
+      "step": 1997312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7289481597758855e-05,
+      "loss": 3.8885,
+      "step": 1997824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7281095650248335e-05,
+      "loss": 3.9004,
+      "step": 1998336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7272709702737814e-05,
+      "loss": 3.8945,
+      "step": 1998848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7264340134031027e-05,
+      "loss": 3.8972,
+      "step": 1999360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7255954186520507e-05,
+      "loss": 3.9017,
+      "step": 1999872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7247568239009987e-05,
+      "loss": 3.9037,
+      "step": 2000384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.72391986703032e-05,
+      "loss": 3.9022,
+      "step": 2000896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7230812722792676e-05,
+      "loss": 3.9135,
+      "step": 2001408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.722242677528216e-05,
+      "loss": 3.9,
+      "step": 2001920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7214057206575372e-05,
+      "loss": 3.9022,
+      "step": 2002432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7205671259064852e-05,
+      "loss": 3.9099,
+      "step": 2002944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7197285311554328e-05,
+      "loss": 3.8885,
+      "step": 2003456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7188899364043808e-05,
+      "loss": 3.8921,
+      "step": 2003968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7180513416533288e-05,
+      "loss": 3.8937,
+      "step": 2004480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7172127469022768e-05,
+      "loss": 3.8912,
+      "step": 2004992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.716374152151225e-05,
+      "loss": 3.8983,
+      "step": 2005504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.715535557400173e-05,
+      "loss": 3.9094,
+      "step": 2006016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7146969626491208e-05,
+      "loss": 3.8955,
+      "step": 2006528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7138583678980688e-05,
+      "loss": 3.8985,
+      "step": 2007040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7130197731470168e-05,
+      "loss": 3.8956,
+      "step": 2007552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7121811783959648e-05,
+      "loss": 3.9037,
+      "step": 2008064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7113425836449128e-05,
+      "loss": 3.8887,
+      "step": 2008576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7105039888938608e-05,
+      "loss": 3.8781,
+      "step": 2009088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.709665394142809e-05,
+      "loss": 3.9092,
+      "step": 2009600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.70882843727213e-05,
+      "loss": 3.9026,
+      "step": 2010112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.707989842521078e-05,
+      "loss": 3.9133,
+      "step": 2010624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.707151247770026e-05,
+      "loss": 3.8921,
+      "step": 2011136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7063126530189737e-05,
+      "loss": 3.8954,
+      "step": 2011648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.705474058267922e-05,
+      "loss": 3.8965,
+      "step": 2012160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.70463546351687e-05,
+      "loss": 3.8907,
+      "step": 2012672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.703796868765818e-05,
+      "loss": 3.9072,
+      "step": 2013184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.702959911895139e-05,
+      "loss": 3.8831,
+      "step": 2013696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.702121317144087e-05,
+      "loss": 3.9214,
+      "step": 2014208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.701282722393035e-05,
+      "loss": 3.9008,
+      "step": 2014720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.700444127641983e-05,
+      "loss": 3.8756,
+      "step": 2015232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.699607170771304e-05,
+      "loss": 3.9047,
+      "step": 2015744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.698768576020252e-05,
+      "loss": 3.8928,
+      "step": 2016256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6979299812692e-05,
+      "loss": 3.8816,
+      "step": 2016768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.697091386518148e-05,
+      "loss": 3.9012,
+      "step": 2017280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.696252791767096e-05,
+      "loss": 3.9051,
+      "step": 2017792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.695414197016044e-05,
+      "loss": 3.8886,
+      "step": 2018304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6945772401453654e-05,
+      "loss": 3.9008,
+      "step": 2018816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6937386453943134e-05,
+      "loss": 3.8769,
+      "step": 2019328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6929000506432614e-05,
+      "loss": 3.8877,
+      "step": 2019840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6920614558922094e-05,
+      "loss": 3.8964,
+      "step": 2020352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.691222861141157e-05,
+      "loss": 3.8921,
+      "step": 2020864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6903859042704783e-05,
+      "loss": 3.9001,
+      "step": 2021376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6895473095194263e-05,
+      "loss": 3.894,
+      "step": 2021888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6887087147683746e-05,
+      "loss": 3.9123,
+      "step": 2022400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6878701200173222e-05,
+      "loss": 3.8929,
+      "step": 2022912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6870315252662702e-05,
+      "loss": 3.9018,
+      "step": 2023424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6861929305152182e-05,
+      "loss": 3.895,
+      "step": 2023936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6853543357641662e-05,
+      "loss": 3.8921,
+      "step": 2024448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6845173788934875e-05,
+      "loss": 3.8975,
+      "step": 2024960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6836787841424355e-05,
+      "loss": 3.8946,
+      "step": 2025472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6828401893913835e-05,
+      "loss": 3.8974,
+      "step": 2025984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6820015946403315e-05,
+      "loss": 3.9061,
+      "step": 2026496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6811629998892795e-05,
+      "loss": 3.8896,
+      "step": 2027008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6803276808989736e-05,
+      "loss": 3.8826,
+      "step": 2027520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6794890861479216e-05,
+      "loss": 3.896,
+      "step": 2028032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6786504913968696e-05,
+      "loss": 3.9021,
+      "step": 2028544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6778118966458176e-05,
+      "loss": 3.889,
+      "step": 2029056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.676974939775139e-05,
+      "loss": 3.9015,
+      "step": 2029568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6761363450240865e-05,
+      "loss": 3.8877,
+      "step": 2030080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6752977502730345e-05,
+      "loss": 3.8846,
+      "step": 2030592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.674459155521983e-05,
+      "loss": 3.8884,
+      "step": 2031104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.673620560770931e-05,
+      "loss": 3.8922,
+      "step": 2031616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6727836039002518e-05,
+      "loss": 3.9087,
+      "step": 2032128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6719450091491997e-05,
+      "loss": 3.8969,
+      "step": 2032640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6711064143981477e-05,
+      "loss": 3.8984,
+      "step": 2033152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6702678196470957e-05,
+      "loss": 3.8923,
+      "step": 2033664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6694292248960437e-05,
+      "loss": 3.8873,
+      "step": 2034176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.668592268025365e-05,
+      "loss": 3.8929,
+      "step": 2034688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.667753673274313e-05,
+      "loss": 3.899,
+      "step": 2035200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.666915078523261e-05,
+      "loss": 3.9027,
+      "step": 2035712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.666078121652582e-05,
+      "loss": 3.898,
+      "step": 2036224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.66523952690153e-05,
+      "loss": 3.8988,
+      "step": 2036736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.664400932150478e-05,
+      "loss": 3.8889,
+      "step": 2037248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6635623373994262e-05,
+      "loss": 3.8977,
+      "step": 2037760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6627237426483742e-05,
+      "loss": 3.8954,
+      "step": 2038272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6618851478973222e-05,
+      "loss": 3.8877,
+      "step": 2038784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.66104655314627e-05,
+      "loss": 3.8944,
+      "step": 2039296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.660207958395218e-05,
+      "loss": 3.8908,
+      "step": 2039808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.659369363644166e-05,
+      "loss": 3.8964,
+      "step": 2040320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.658530768893114e-05,
+      "loss": 3.8839,
+      "step": 2040832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6576921741420622e-05,
+      "loss": 3.8909,
+      "step": 2041344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.656855217271383e-05,
+      "loss": 3.8943,
+      "step": 2041856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.656016622520331e-05,
+      "loss": 3.8925,
+      "step": 2042368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.655178027769279e-05,
+      "loss": 3.8998,
+      "step": 2042880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.654339433018227e-05,
+      "loss": 3.9045,
+      "step": 2043392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.653500838267175e-05,
+      "loss": 3.9072,
+      "step": 2043904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6526638813964963e-05,
+      "loss": 3.8981,
+      "step": 2044416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6518252866454443e-05,
+      "loss": 3.8947,
+      "step": 2044928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6509866918943923e-05,
+      "loss": 3.8867,
+      "step": 2045440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6501497350237132e-05,
+      "loss": 3.904,
+      "step": 2045952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6493111402726612e-05,
+      "loss": 3.8966,
+      "step": 2046464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6484725455216092e-05,
+      "loss": 3.886,
+      "step": 2046976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6476339507705575e-05,
+      "loss": 3.892,
+      "step": 2047488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6467953560195052e-05,
+      "loss": 3.9024,
+      "step": 2048000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6459567612684532e-05,
+      "loss": 3.8968,
+      "step": 2048512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6451181665174012e-05,
+      "loss": 3.8916,
+      "step": 2049024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6442795717663492e-05,
+      "loss": 3.8931,
+      "step": 2049536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.64344261489567e-05,
+      "loss": 3.8889,
+      "step": 2050048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6426040201446184e-05,
+      "loss": 3.9033,
+      "step": 2050560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6417654253935664e-05,
+      "loss": 3.9016,
+      "step": 2051072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6409268306425144e-05,
+      "loss": 3.904,
+      "step": 2051584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6400882358914624e-05,
+      "loss": 3.8903,
+      "step": 2052096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6392496411404104e-05,
+      "loss": 3.8983,
+      "step": 2052608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6384110463893584e-05,
+      "loss": 3.907,
+      "step": 2053120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6375740895186793e-05,
+      "loss": 3.8916,
+      "step": 2053632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6367354947676277e-05,
+      "loss": 3.8987,
+      "step": 2054144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6358969000165756e-05,
+      "loss": 3.9011,
+      "step": 2054656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6350583052655233e-05,
+      "loss": 3.8997,
+      "step": 2055168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6342213483948446e-05,
+      "loss": 3.9109,
+      "step": 2055680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6333827536437926e-05,
+      "loss": 3.898,
+      "step": 2056192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6325441588927405e-05,
+      "loss": 3.8985,
+      "step": 2056704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6317055641416885e-05,
+      "loss": 3.9042,
+      "step": 2057216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6308686072710098e-05,
+      "loss": 3.8961,
+      "step": 2057728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6300300125199578e-05,
+      "loss": 3.8905,
+      "step": 2058240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6291914177689058e-05,
+      "loss": 3.8894,
+      "step": 2058752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6283528230178534e-05,
+      "loss": 3.8908,
+      "step": 2059264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6275142282668014e-05,
+      "loss": 3.8972,
+      "step": 2059776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.626678909276496e-05,
+      "loss": 3.8948,
+      "step": 2060288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.9948482513427734,
+      "eval_runtime": 295.9048,
+      "eval_samples_per_second": 1289.574,
+      "eval_steps_per_second": 40.3,
+      "step": 2060624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.625840314525444e-05,
+      "loss": 3.8955,
+      "step": 2060800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.625001719774392e-05,
+      "loss": 3.8934,
+      "step": 2061312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.62416312502334e-05,
+      "loss": 3.8971,
+      "step": 2061824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.623324530272288e-05,
+      "loss": 3.8981,
+      "step": 2062336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6224859355212356e-05,
+      "loss": 3.9091,
+      "step": 2062848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.621647340770184e-05,
+      "loss": 3.899,
+      "step": 2063360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.620810383899505e-05,
+      "loss": 3.8999,
+      "step": 2063872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.619973427028826e-05,
+      "loss": 3.8913,
+      "step": 2064384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.619134832277774e-05,
+      "loss": 3.8875,
+      "step": 2064896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.618296237526722e-05,
+      "loss": 3.8936,
+      "step": 2065408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.61745764277567e-05,
+      "loss": 3.9053,
+      "step": 2065920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.616619048024618e-05,
+      "loss": 3.8909,
+      "step": 2066432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6157820911539393e-05,
+      "loss": 3.9002,
+      "step": 2066944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6149434964028873e-05,
+      "loss": 3.8854,
+      "step": 2067456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6141049016518353e-05,
+      "loss": 3.8898,
+      "step": 2067968
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.613266306900783e-05,
+      "loss": 3.8791,
+      "step": 2068480
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.612427712149731e-05,
+      "loss": 3.8995,
+      "step": 2068992
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6115891173986793e-05,
+      "loss": 3.8974,
+      "step": 2069504
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6107505226476273e-05,
+      "loss": 3.8927,
+      "step": 2070016
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6099119278965753e-05,
+      "loss": 3.898,
+      "step": 2070528
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6090749710258962e-05,
+      "loss": 3.9075,
+      "step": 2071040
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6082363762748442e-05,
+      "loss": 3.901,
+      "step": 2071552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.607397781523792e-05,
+      "loss": 3.893,
+      "step": 2072064
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.60655918677274e-05,
+      "loss": 3.8967,
+      "step": 2072576
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6057205920216885e-05,
+      "loss": 3.9025,
+      "step": 2073088
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.604881997270636e-05,
+      "loss": 3.8997,
+      "step": 2073600
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.604043402519584e-05,
+      "loss": 3.8853,
+      "step": 2074112
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.603204807768532e-05,
+      "loss": 3.8962,
+      "step": 2074624
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6023678508978534e-05,
+      "loss": 3.8939,
+      "step": 2075136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.601529256146801e-05,
+      "loss": 3.8876,
+      "step": 2075648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.6006906613957494e-05,
+      "loss": 3.8982,
+      "step": 2076160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5998520666446974e-05,
+      "loss": 3.8981,
+      "step": 2076672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5990134718936454e-05,
+      "loss": 3.9012,
+      "step": 2077184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5981748771425934e-05,
+      "loss": 3.9075,
+      "step": 2077696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5973362823915414e-05,
+      "loss": 3.8924,
+      "step": 2078208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5964993255208623e-05,
+      "loss": 3.9019,
+      "step": 2078720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5956607307698103e-05,
+      "loss": 3.904,
+      "step": 2079232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5948221360187586e-05,
+      "loss": 3.8833,
+      "step": 2079744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5939835412677066e-05,
+      "loss": 3.8907,
+      "step": 2080256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5931449465166543e-05,
+      "loss": 3.8851,
+      "step": 2080768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5923063517656023e-05,
+      "loss": 3.8844,
+      "step": 2081280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5914677570145502e-05,
+      "loss": 3.8949,
+      "step": 2081792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5906291622634982e-05,
+      "loss": 3.9064,
+      "step": 2082304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5897922053928195e-05,
+      "loss": 3.8913,
+      "step": 2082816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5889552485221407e-05,
+      "loss": 3.8904,
+      "step": 2083328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5881166537710887e-05,
+      "loss": 3.8943,
+      "step": 2083840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5872780590200364e-05,
+      "loss": 3.8966,
+      "step": 2084352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5864394642689844e-05,
+      "loss": 3.8857,
+      "step": 2084864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5856008695179324e-05,
+      "loss": 3.8739,
+      "step": 2085376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5847622747668807e-05,
+      "loss": 3.9018,
+      "step": 2085888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5839236800158287e-05,
+      "loss": 3.9033,
+      "step": 2086400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5830867231451496e-05,
+      "loss": 3.906,
+      "step": 2086912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5822481283940976e-05,
+      "loss": 3.8921,
+      "step": 2087424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5814095336430456e-05,
+      "loss": 3.8891,
+      "step": 2087936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5805709388919936e-05,
+      "loss": 3.8952,
+      "step": 2088448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.579733982021315e-05,
+      "loss": 3.882,
+      "step": 2088960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.578895387270263e-05,
+      "loss": 3.9046,
+      "step": 2089472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.578056792519211e-05,
+      "loss": 3.8783,
+      "step": 2089984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.577218197768159e-05,
+      "loss": 3.917,
+      "step": 2090496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.576379603017107e-05,
+      "loss": 3.8941,
+      "step": 2091008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5755410082660545e-05,
+      "loss": 3.8771,
+      "step": 2091520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5747024135150025e-05,
+      "loss": 3.8983,
+      "step": 2092032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5738638187639508e-05,
+      "loss": 3.8886,
+      "step": 2092544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.573026861893272e-05,
+      "loss": 3.8754,
+      "step": 2093056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.572189905022593e-05,
+      "loss": 3.8951,
+      "step": 2093568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.571351310271541e-05,
+      "loss": 3.9019,
+      "step": 2094080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.570512715520489e-05,
+      "loss": 3.8865,
+      "step": 2094592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.569674120769437e-05,
+      "loss": 3.8947,
+      "step": 2095104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.568835526018385e-05,
+      "loss": 3.8702,
+      "step": 2095616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5679985691477062e-05,
+      "loss": 3.887,
+      "step": 2096128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5671599743966542e-05,
+      "loss": 3.888,
+      "step": 2096640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.566321379645602e-05,
+      "loss": 3.8935,
+      "step": 2097152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.56548278489455e-05,
+      "loss": 3.893,
+      "step": 2097664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.564644190143498e-05,
+      "loss": 3.8869,
+      "step": 2098176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5638072332728194e-05,
+      "loss": 3.9106,
+      "step": 2098688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.562968638521767e-05,
+      "loss": 3.8907,
+      "step": 2099200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.562130043770715e-05,
+      "loss": 3.898,
+      "step": 2099712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.561291449019663e-05,
+      "loss": 3.8901,
+      "step": 2100224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.560452854268611e-05,
+      "loss": 3.8867,
+      "step": 2100736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.559614259517559e-05,
+      "loss": 3.8991,
+      "step": 2101248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.558775664766507e-05,
+      "loss": 3.889,
+      "step": 2101760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.557937070015455e-05,
+      "loss": 3.8898,
+      "step": 2102272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5571017510251492e-05,
+      "loss": 3.9051,
+      "step": 2102784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5562631562740972e-05,
+      "loss": 3.8842,
+      "step": 2103296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5554245615230452e-05,
+      "loss": 3.8795,
+      "step": 2103808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5545859667719932e-05,
+      "loss": 3.8896,
+      "step": 2104320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5537473720209416e-05,
+      "loss": 3.8963,
+      "step": 2104832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5529087772698896e-05,
+      "loss": 3.8848,
+      "step": 2105344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5520734582795837e-05,
+      "loss": 3.8982,
+      "step": 2105856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5512348635285317e-05,
+      "loss": 3.8868,
+      "step": 2106368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5503962687774794e-05,
+      "loss": 3.8806,
+      "step": 2106880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5495576740264277e-05,
+      "loss": 3.886,
+      "step": 2107392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5487190792753757e-05,
+      "loss": 3.8866,
+      "step": 2107904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5478804845243237e-05,
+      "loss": 3.9034,
+      "step": 2108416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5470418897732717e-05,
+      "loss": 3.8934,
+      "step": 2108928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5462032950222197e-05,
+      "loss": 3.897,
+      "step": 2109440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5453647002711673e-05,
+      "loss": 3.8865,
+      "step": 2109952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5445261055201153e-05,
+      "loss": 3.8818,
+      "step": 2110464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5436891486494366e-05,
+      "loss": 3.8909,
+      "step": 2110976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.542850553898385e-05,
+      "loss": 3.8939,
+      "step": 2111488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5420119591473326e-05,
+      "loss": 3.8962,
+      "step": 2112000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5411733643962806e-05,
+      "loss": 3.895,
+      "step": 2112512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5403347696452286e-05,
+      "loss": 3.8928,
+      "step": 2113024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5394961748941766e-05,
+      "loss": 3.8855,
+      "step": 2113536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5386592180234978e-05,
+      "loss": 3.8967,
+      "step": 2114048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5378206232724458e-05,
+      "loss": 3.8882,
+      "step": 2114560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5369820285213938e-05,
+      "loss": 3.8851,
+      "step": 2115072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5361434337703418e-05,
+      "loss": 3.8878,
+      "step": 2115584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5353064768996627e-05,
+      "loss": 3.8873,
+      "step": 2116096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5344678821486107e-05,
+      "loss": 3.8889,
+      "step": 2116608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5336292873975587e-05,
+      "loss": 3.8814,
+      "step": 2117120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.532790692646507e-05,
+      "loss": 3.8847,
+      "step": 2117632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.531952097895455e-05,
+      "loss": 3.8931,
+      "step": 2118144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.531115141024776e-05,
+      "loss": 3.89,
+      "step": 2118656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.530276546273724e-05,
+      "loss": 3.8941,
+      "step": 2119168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.529437951522672e-05,
+      "loss": 3.8978,
+      "step": 2119680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.52859935677162e-05,
+      "loss": 3.9084,
+      "step": 2120192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.527760762020568e-05,
+      "loss": 3.8897,
+      "step": 2120704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5269238051498892e-05,
+      "loss": 3.8906,
+      "step": 2121216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5260852103988372e-05,
+      "loss": 3.886,
+      "step": 2121728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.525248253528158e-05,
+      "loss": 3.8958,
+      "step": 2122240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5244096587771062e-05,
+      "loss": 3.8906,
+      "step": 2122752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5235710640260542e-05,
+      "loss": 3.8884,
+      "step": 2123264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5227324692750022e-05,
+      "loss": 3.8884,
+      "step": 2123776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5218938745239502e-05,
+      "loss": 3.8972,
+      "step": 2124288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.521055279772898e-05,
+      "loss": 3.8921,
+      "step": 2124800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.520216685021846e-05,
+      "loss": 3.8852,
+      "step": 2125312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.519378090270794e-05,
+      "loss": 3.8908,
+      "step": 2125824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.518539495519742e-05,
+      "loss": 3.8873,
+      "step": 2126336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5177025386490631e-05,
+      "loss": 3.8937,
+      "step": 2126848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5168639438980111e-05,
+      "loss": 3.8975,
+      "step": 2127360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5160269870273325e-05,
+      "loss": 3.9001,
+      "step": 2127872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5151883922762802e-05,
+      "loss": 3.887,
+      "step": 2128384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5143514354056016e-05,
+      "loss": 3.8928,
+      "step": 2128896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5135128406545496e-05,
+      "loss": 3.8991,
+      "step": 2129408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5126742459034976e-05,
+      "loss": 3.8874,
+      "step": 2129920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5118356511524454e-05,
+      "loss": 3.8905,
+      "step": 2130432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5109970564013934e-05,
+      "loss": 3.8978,
+      "step": 2130944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5101584616503414e-05,
+      "loss": 3.9044,
+      "step": 2131456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5093198668992894e-05,
+      "loss": 3.9025,
+      "step": 2131968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5084812721482374e-05,
+      "loss": 3.8917,
+      "step": 2132480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5076426773971856e-05,
+      "loss": 3.8965,
+      "step": 2132992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5068040826461332e-05,
+      "loss": 3.8975,
+      "step": 2133504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5059654878950812e-05,
+      "loss": 3.8933,
+      "step": 2134016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5051268931440294e-05,
+      "loss": 3.886,
+      "step": 2134528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5042882983929774e-05,
+      "loss": 3.8884,
+      "step": 2135040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5034497036419254e-05,
+      "loss": 3.8838,
+      "step": 2135552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5026127467712465e-05,
+      "loss": 3.89,
+      "step": 2136064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5017757899005677e-05,
+      "loss": 3.891,
+      "step": 2136576
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9934945106506348,
+      "eval_runtime": 296.8979,
+      "eval_samples_per_second": 1285.26,
+      "eval_steps_per_second": 40.165,
+      "step": 2136944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5009371951495157e-05,
+      "loss": 3.8987,
+      "step": 2137088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5000986003984635e-05,
+      "loss": 3.8881,
+      "step": 2137600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4992600056474115e-05,
+      "loss": 3.8895,
+      "step": 2138112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4984214108963595e-05,
+      "loss": 3.8998,
+      "step": 2138624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4975828161453075e-05,
+      "loss": 3.9006,
+      "step": 2139136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4967442213942557e-05,
+      "loss": 3.8976,
+      "step": 2139648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4959056266432037e-05,
+      "loss": 3.8997,
+      "step": 2140160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4950670318921513e-05,
+      "loss": 3.8862,
+      "step": 2140672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4942284371410995e-05,
+      "loss": 3.8845,
+      "step": 2141184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4933898423900475e-05,
+      "loss": 3.8922,
+      "step": 2141696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4925512476389955e-05,
+      "loss": 3.897,
+      "step": 2142208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4917126528879435e-05,
+      "loss": 3.8897,
+      "step": 2142720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4908756960172646e-05,
+      "loss": 3.9012,
+      "step": 2143232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4900371012662126e-05,
+      "loss": 3.8775,
+      "step": 2143744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4891985065151606e-05,
+      "loss": 3.8877,
+      "step": 2144256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4883599117641087e-05,
+      "loss": 3.8737,
+      "step": 2144768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4875213170130567e-05,
+      "loss": 3.8964,
+      "step": 2145280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4866827222620047e-05,
+      "loss": 3.8928,
+      "step": 2145792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4858441275109525e-05,
+      "loss": 3.8886,
+      "step": 2146304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4850055327599005e-05,
+      "loss": 3.8918,
+      "step": 2146816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4841669380088485e-05,
+      "loss": 3.9041,
+      "step": 2147328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4833283432577965e-05,
+      "loss": 3.8952,
+      "step": 2147840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4824897485067445e-05,
+      "loss": 3.8925,
+      "step": 2148352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4816511537556927e-05,
+      "loss": 3.8915,
+      "step": 2148864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4808141968850136e-05,
+      "loss": 3.9014,
+      "step": 2149376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4799756021339618e-05,
+      "loss": 3.8914,
+      "step": 2149888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4791386452632827e-05,
+      "loss": 3.8871,
+      "step": 2150400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4783000505122307e-05,
+      "loss": 3.8899,
+      "step": 2150912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4774614557611788e-05,
+      "loss": 3.8836,
+      "step": 2151424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4766228610101268e-05,
+      "loss": 3.8845,
+      "step": 2151936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4757842662590748e-05,
+      "loss": 3.898,
+      "step": 2152448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4749456715080228e-05,
+      "loss": 3.8912,
+      "step": 2152960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4741087146373439e-05,
+      "loss": 3.8987,
+      "step": 2153472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4732701198862919e-05,
+      "loss": 3.9037,
+      "step": 2153984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4724315251352399e-05,
+      "loss": 3.889,
+      "step": 2154496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.471594568264561e-05,
+      "loss": 3.8966,
+      "step": 2155008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.470755973513509e-05,
+      "loss": 3.9023,
+      "step": 2155520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4699173787624571e-05,
+      "loss": 3.8825,
+      "step": 2156032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4690787840114051e-05,
+      "loss": 3.8881,
+      "step": 2156544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4682401892603528e-05,
+      "loss": 3.878,
+      "step": 2157056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.467401594509301e-05,
+      "loss": 3.8805,
+      "step": 2157568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.466562999758249e-05,
+      "loss": 3.8922,
+      "step": 2158080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.465724405007197e-05,
+      "loss": 3.9011,
+      "step": 2158592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.464885810256145e-05,
+      "loss": 3.8841,
+      "step": 2159104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.464047215505093e-05,
+      "loss": 3.8893,
+      "step": 2159616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.463210258634414e-05,
+      "loss": 3.8894,
+      "step": 2160128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.462371663883362e-05,
+      "loss": 3.891,
+      "step": 2160640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4615330691323102e-05,
+      "loss": 3.8832,
+      "step": 2161152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4606944743812582e-05,
+      "loss": 3.8699,
+      "step": 2161664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4598558796302058e-05,
+      "loss": 3.8946,
+      "step": 2162176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4590172848791538e-05,
+      "loss": 3.9018,
+      "step": 2162688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4581803280084752e-05,
+      "loss": 3.9031,
+      "step": 2163200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4573417332574232e-05,
+      "loss": 3.8822,
+      "step": 2163712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.456503138506371e-05,
+      "loss": 3.889,
+      "step": 2164224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.455664543755319e-05,
+      "loss": 3.8885,
+      "step": 2164736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.454825949004267e-05,
+      "loss": 3.8785,
+      "step": 2165248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4539889921335881e-05,
+      "loss": 3.9034,
+      "step": 2165760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4531503973825361e-05,
+      "loss": 3.8775,
+      "step": 2166272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4523118026314841e-05,
+      "loss": 3.9092,
+      "step": 2166784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4514732078804321e-05,
+      "loss": 3.8894,
+      "step": 2167296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4506346131293803e-05,
+      "loss": 3.8751,
+      "step": 2167808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4497976562587012e-05,
+      "loss": 3.8934,
+      "step": 2168320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4489590615076492e-05,
+      "loss": 3.8843,
+      "step": 2168832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4481204667565973e-05,
+      "loss": 3.8712,
+      "step": 2169344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4472818720055453e-05,
+      "loss": 3.8888,
+      "step": 2169856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4464432772544933e-05,
+      "loss": 3.8966,
+      "step": 2170368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4456063203838144e-05,
+      "loss": 3.8828,
+      "step": 2170880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4447677256327624e-05,
+      "loss": 3.891,
+      "step": 2171392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4439291308817104e-05,
+      "loss": 3.866,
+      "step": 2171904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4430905361306584e-05,
+      "loss": 3.8853,
+      "step": 2172416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4422519413796062e-05,
+      "loss": 3.8813,
+      "step": 2172928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4414133466285542e-05,
+      "loss": 3.8919,
+      "step": 2173440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4405747518775022e-05,
+      "loss": 3.8846,
+      "step": 2173952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4397377950068236e-05,
+      "loss": 3.8874,
+      "step": 2174464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4388992002557713e-05,
+      "loss": 3.904,
+      "step": 2174976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4380606055047195e-05,
+      "loss": 3.8919,
+      "step": 2175488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4372220107536675e-05,
+      "loss": 3.8898,
+      "step": 2176000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4363834160026155e-05,
+      "loss": 3.8858,
+      "step": 2176512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4355448212515634e-05,
+      "loss": 3.8823,
+      "step": 2177024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4347062265005114e-05,
+      "loss": 3.8952,
+      "step": 2177536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4338692696298325e-05,
+      "loss": 3.8825,
+      "step": 2178048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4330306748787805e-05,
+      "loss": 3.8876,
+      "step": 2178560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4321920801277285e-05,
+      "loss": 3.8982,
+      "step": 2179072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4313534853766767e-05,
+      "loss": 3.8855,
+      "step": 2179584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4305148906256243e-05,
+      "loss": 3.8699,
+      "step": 2180096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4296762958745723e-05,
+      "loss": 3.8877,
+      "step": 2180608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4288377011235205e-05,
+      "loss": 3.8907,
+      "step": 2181120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4280007442528417e-05,
+      "loss": 3.8815,
+      "step": 2181632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4271637873821628e-05,
+      "loss": 3.8916,
+      "step": 2182144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4263251926311108e-05,
+      "loss": 3.8862,
+      "step": 2182656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4254865978800588e-05,
+      "loss": 3.8761,
+      "step": 2183168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4246480031290068e-05,
+      "loss": 3.8823,
+      "step": 2183680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4238094083779546e-05,
+      "loss": 3.8841,
+      "step": 2184192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4229708136269026e-05,
+      "loss": 3.8965,
+      "step": 2184704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4221338567562239e-05,
+      "loss": 3.8909,
+      "step": 2185216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4212952620051717e-05,
+      "loss": 3.8957,
+      "step": 2185728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4204566672541197e-05,
+      "loss": 3.8827,
+      "step": 2186240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4196180725030677e-05,
+      "loss": 3.8806,
+      "step": 2186752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4187794777520159e-05,
+      "loss": 3.8849,
+      "step": 2187264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4179408830009639e-05,
+      "loss": 3.8933,
+      "step": 2187776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4171022882499119e-05,
+      "loss": 3.893,
+      "step": 2188288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4162636934988598e-05,
+      "loss": 3.8916,
+      "step": 2188800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4154250987478077e-05,
+      "loss": 3.8887,
+      "step": 2189312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.414588141877129e-05,
+      "loss": 3.8829,
+      "step": 2189824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.413749547126077e-05,
+      "loss": 3.8922,
+      "step": 2190336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4129109523750247e-05,
+      "loss": 3.8862,
+      "step": 2190848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4120723576239727e-05,
+      "loss": 3.8807,
+      "step": 2191360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4112337628729207e-05,
+      "loss": 3.8862,
+      "step": 2191872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4103951681218689e-05,
+      "loss": 3.883,
+      "step": 2192384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4095582112511898e-05,
+      "loss": 3.8842,
+      "step": 2192896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.408719616500138e-05,
+      "loss": 3.8795,
+      "step": 2193408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.407881021749086e-05,
+      "loss": 3.8776,
+      "step": 2193920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.407042426998034e-05,
+      "loss": 3.8893,
+      "step": 2194432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.406205470127355e-05,
+      "loss": 3.8854,
+      "step": 2194944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.405366875376303e-05,
+      "loss": 3.8858,
+      "step": 2195456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.404528280625251e-05,
+      "loss": 3.8964,
+      "step": 2195968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.403689685874199e-05,
+      "loss": 3.9018,
+      "step": 2196480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.402851091123147e-05,
+      "loss": 3.8853,
+      "step": 2196992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4020124963720952e-05,
+      "loss": 3.8879,
+      "step": 2197504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4011739016210429e-05,
+      "loss": 3.8802,
+      "step": 2198016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4003353068699908e-05,
+      "loss": 3.8926,
+      "step": 2198528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3994983499993123e-05,
+      "loss": 3.8853,
+      "step": 2199040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3986597552482603e-05,
+      "loss": 3.8812,
+      "step": 2199552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3978211604972081e-05,
+      "loss": 3.8851,
+      "step": 2200064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.396982565746156e-05,
+      "loss": 3.8951,
+      "step": 2200576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3961456088754773e-05,
+      "loss": 3.8844,
+      "step": 2201088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3953070141244253e-05,
+      "loss": 3.8876,
+      "step": 2201600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3944684193733732e-05,
+      "loss": 3.8885,
+      "step": 2202112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3936298246223211e-05,
+      "loss": 3.8858,
+      "step": 2202624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3927928677516424e-05,
+      "loss": 3.8857,
+      "step": 2203136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3919542730005902e-05,
+      "loss": 3.8976,
+      "step": 2203648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3911173161299115e-05,
+      "loss": 3.8935,
+      "step": 2204160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3902787213788596e-05,
+      "loss": 3.884,
+      "step": 2204672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3894401266278076e-05,
+      "loss": 3.8888,
+      "step": 2205184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3886015318767553e-05,
+      "loss": 3.8984,
+      "step": 2205696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3877629371257035e-05,
+      "loss": 3.8827,
+      "step": 2206208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3869259802550247e-05,
+      "loss": 3.8896,
+      "step": 2206720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3860873855039727e-05,
+      "loss": 3.8925,
+      "step": 2207232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3852487907529205e-05,
+      "loss": 3.9011,
+      "step": 2207744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3844101960018685e-05,
+      "loss": 3.8964,
+      "step": 2208256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3835716012508165e-05,
+      "loss": 3.8934,
+      "step": 2208768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3827330064997645e-05,
+      "loss": 3.8873,
+      "step": 2209280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3818944117487127e-05,
+      "loss": 3.8947,
+      "step": 2209792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3810574548780336e-05,
+      "loss": 3.892,
+      "step": 2210304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3802188601269816e-05,
+      "loss": 3.8816,
+      "step": 2210816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3793819032563027e-05,
+      "loss": 3.8868,
+      "step": 2211328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3785433085052507e-05,
+      "loss": 3.8771,
+      "step": 2211840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3777047137541988e-05,
+      "loss": 3.8897,
+      "step": 2212352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3768661190031468e-05,
+      "loss": 3.8871,
+      "step": 2212864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.992324113845825,
+      "eval_runtime": 302.6268,
+      "eval_samples_per_second": 1260.929,
+      "eval_steps_per_second": 39.405,
+      "step": 2213264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3760275242520948e-05,
+      "loss": 3.89,
+      "step": 2213376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3751889295010428e-05,
+      "loss": 3.8844,
+      "step": 2213888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3743503347499908e-05,
+      "loss": 3.8847,
+      "step": 2214400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3735117399989386e-05,
+      "loss": 3.8966,
+      "step": 2214912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3726747831282599e-05,
+      "loss": 3.8927,
+      "step": 2215424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3718361883772079e-05,
+      "loss": 3.8992,
+      "step": 2215936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3709975936261557e-05,
+      "loss": 3.8889,
+      "step": 2216448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3701589988751037e-05,
+      "loss": 3.8832,
+      "step": 2216960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3693204041240517e-05,
+      "loss": 3.8792,
+      "step": 2217472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3684834472533731e-05,
+      "loss": 3.8899,
+      "step": 2217984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3676448525023208e-05,
+      "loss": 3.8898,
+      "step": 2218496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.366806257751269e-05,
+      "loss": 3.8874,
+      "step": 2219008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3659693008805902e-05,
+      "loss": 3.8988,
+      "step": 2219520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3651307061295382e-05,
+      "loss": 3.8729,
+      "step": 2220032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.364292111378486e-05,
+      "loss": 3.8833,
+      "step": 2220544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.363453516627434e-05,
+      "loss": 3.8737,
+      "step": 2221056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.362614921876382e-05,
+      "loss": 3.8927,
+      "step": 2221568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.36177632712533e-05,
+      "loss": 3.8842,
+      "step": 2222080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.360939370254651e-05,
+      "loss": 3.8848,
+      "step": 2222592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.360100775503599e-05,
+      "loss": 3.8935,
+      "step": 2223104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.359262180752547e-05,
+      "loss": 3.8954,
+      "step": 2223616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3584235860014952e-05,
+      "loss": 3.8933,
+      "step": 2224128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3575849912504432e-05,
+      "loss": 3.8867,
+      "step": 2224640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3567463964993912e-05,
+      "loss": 3.8894,
+      "step": 2225152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.355907801748339e-05,
+      "loss": 3.898,
+      "step": 2225664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3550708448776603e-05,
+      "loss": 3.8889,
+      "step": 2226176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3542322501266083e-05,
+      "loss": 3.8777,
+      "step": 2226688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3533936553755561e-05,
+      "loss": 3.8896,
+      "step": 2227200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3525550606245041e-05,
+      "loss": 3.8774,
+      "step": 2227712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3517164658734521e-05,
+      "loss": 3.8846,
+      "step": 2228224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3508778711224001e-05,
+      "loss": 3.8924,
+      "step": 2228736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3500392763713483e-05,
+      "loss": 3.8858,
+      "step": 2229248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3492006816202963e-05,
+      "loss": 3.8918,
+      "step": 2229760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3483620868692443e-05,
+      "loss": 3.899,
+      "step": 2230272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.347523492118192e-05,
+      "loss": 3.8859,
+      "step": 2230784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3466881731278862e-05,
+      "loss": 3.8946,
+      "step": 2231296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3458495783768344e-05,
+      "loss": 3.8955,
+      "step": 2231808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3450109836257824e-05,
+      "loss": 3.8752,
+      "step": 2232320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3441723888747304e-05,
+      "loss": 3.8869,
+      "step": 2232832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3433337941236784e-05,
+      "loss": 3.8706,
+      "step": 2233344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3424951993726264e-05,
+      "loss": 3.8814,
+      "step": 2233856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3416582425019475e-05,
+      "loss": 3.8842,
+      "step": 2234368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3408196477508955e-05,
+      "loss": 3.9006,
+      "step": 2234880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3399810529998436e-05,
+      "loss": 3.8793,
+      "step": 2235392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3391424582487916e-05,
+      "loss": 3.8855,
+      "step": 2235904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3383038634977393e-05,
+      "loss": 3.8849,
+      "step": 2236416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3374652687466874e-05,
+      "loss": 3.8875,
+      "step": 2236928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3366266739956354e-05,
+      "loss": 3.8825,
+      "step": 2237440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3357880792445834e-05,
+      "loss": 3.8654,
+      "step": 2237952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3349494844935314e-05,
+      "loss": 3.8872,
+      "step": 2238464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3341108897424794e-05,
+      "loss": 3.9009,
+      "step": 2238976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3332722949914276e-05,
+      "loss": 3.8943,
+      "step": 2239488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3324353381207485e-05,
+      "loss": 3.8791,
+      "step": 2240000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3315967433696967e-05,
+      "loss": 3.8827,
+      "step": 2240512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3307581486186447e-05,
+      "loss": 3.8882,
+      "step": 2241024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3299195538675923e-05,
+      "loss": 3.8732,
+      "step": 2241536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3290809591165405e-05,
+      "loss": 3.9011,
+      "step": 2242048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3282423643654885e-05,
+      "loss": 3.8704,
+      "step": 2242560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3274037696144365e-05,
+      "loss": 3.9035,
+      "step": 2243072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3265651748633845e-05,
+      "loss": 3.8879,
+      "step": 2243584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3257282179927055e-05,
+      "loss": 3.8714,
+      "step": 2244096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3248896232416535e-05,
+      "loss": 3.8836,
+      "step": 2244608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3240510284906015e-05,
+      "loss": 3.8848,
+      "step": 2245120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3232124337395495e-05,
+      "loss": 3.8693,
+      "step": 2245632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3223754768688706e-05,
+      "loss": 3.8809,
+      "step": 2246144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.321538519998192e-05,
+      "loss": 3.8973,
+      "step": 2246656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3206999252471397e-05,
+      "loss": 3.88,
+      "step": 2247168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3198613304960877e-05,
+      "loss": 3.8862,
+      "step": 2247680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3190227357450358e-05,
+      "loss": 3.863,
+      "step": 2248192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3181841409939838e-05,
+      "loss": 3.8814,
+      "step": 2248704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3173455462429318e-05,
+      "loss": 3.8774,
+      "step": 2249216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3165069514918798e-05,
+      "loss": 3.8874,
+      "step": 2249728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3156699946212009e-05,
+      "loss": 3.8796,
+      "step": 2250240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3148313998701489e-05,
+      "loss": 3.8872,
+      "step": 2250752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3139928051190969e-05,
+      "loss": 3.8944,
+      "step": 2251264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3131542103680449e-05,
+      "loss": 3.8883,
+      "step": 2251776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.312317253497366e-05,
+      "loss": 3.8844,
+      "step": 2252288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.311478658746314e-05,
+      "loss": 3.8862,
+      "step": 2252800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3106400639952621e-05,
+      "loss": 3.882,
+      "step": 2253312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3098014692442101e-05,
+      "loss": 3.8897,
+      "step": 2253824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3089628744931578e-05,
+      "loss": 3.8788,
+      "step": 2254336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3081259176224792e-05,
+      "loss": 3.8879,
+      "step": 2254848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3072873228714272e-05,
+      "loss": 3.8961,
+      "step": 2255360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3064487281203752e-05,
+      "loss": 3.8837,
+      "step": 2255872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.305610133369323e-05,
+      "loss": 3.8626,
+      "step": 2256384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.304771538618271e-05,
+      "loss": 3.8864,
+      "step": 2256896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.303932943867219e-05,
+      "loss": 3.8855,
+      "step": 2257408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.303094349116167e-05,
+      "loss": 3.8762,
+      "step": 2257920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3022573922454881e-05,
+      "loss": 3.8851,
+      "step": 2258432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.3014187974944361e-05,
+      "loss": 3.8822,
+      "step": 2258944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.3005802027433841e-05,
+      "loss": 3.8721,
+      "step": 2259456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2997416079923323e-05,
+      "loss": 3.8773,
+      "step": 2259968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2989030132412802e-05,
+      "loss": 3.8807,
+      "step": 2260480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2980644184902282e-05,
+      "loss": 3.894,
+      "step": 2260992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.297225823739176e-05,
+      "loss": 3.8858,
+      "step": 2261504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2963888668684973e-05,
+      "loss": 3.891,
+      "step": 2262016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2955502721174453e-05,
+      "loss": 3.8773,
+      "step": 2262528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2947116773663933e-05,
+      "loss": 3.8771,
+      "step": 2263040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2938730826153411e-05,
+      "loss": 3.8828,
+      "step": 2263552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2930361257446624e-05,
+      "loss": 3.8888,
+      "step": 2264064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2921975309936104e-05,
+      "loss": 3.885,
+      "step": 2264576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2913605741229315e-05,
+      "loss": 3.8868,
+      "step": 2265088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2905219793718795e-05,
+      "loss": 3.8873,
+      "step": 2265600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2896833846208276e-05,
+      "loss": 3.8754,
+      "step": 2266112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2888447898697756e-05,
+      "loss": 3.8901,
+      "step": 2266624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2880061951187233e-05,
+      "loss": 3.8799,
+      "step": 2267136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2871676003676714e-05,
+      "loss": 3.8798,
+      "step": 2267648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2863290056166194e-05,
+      "loss": 3.8812,
+      "step": 2268160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2854904108655674e-05,
+      "loss": 3.8794,
+      "step": 2268672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2846534539948885e-05,
+      "loss": 3.8795,
+      "step": 2269184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2838148592438365e-05,
+      "loss": 3.8788,
+      "step": 2269696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2829762644927845e-05,
+      "loss": 3.8733,
+      "step": 2270208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2821393076221056e-05,
+      "loss": 3.8853,
+      "step": 2270720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2813007128710536e-05,
+      "loss": 3.8847,
+      "step": 2271232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2804621181200016e-05,
+      "loss": 3.8788,
+      "step": 2271744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2796235233689496e-05,
+      "loss": 3.8948,
+      "step": 2272256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2787849286178977e-05,
+      "loss": 3.8973,
+      "step": 2272768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2779463338668457e-05,
+      "loss": 3.8806,
+      "step": 2273280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2771077391157937e-05,
+      "loss": 3.8828,
+      "step": 2273792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2762691443647415e-05,
+      "loss": 3.8796,
+      "step": 2274304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2754305496136895e-05,
+      "loss": 3.8862,
+      "step": 2274816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2745919548626375e-05,
+      "loss": 3.8831,
+      "step": 2275328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2737533601115855e-05,
+      "loss": 3.8803,
+      "step": 2275840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2729147653605337e-05,
+      "loss": 3.8812,
+      "step": 2276352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2720778084898546e-05,
+      "loss": 3.8898,
+      "step": 2276864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2712392137388026e-05,
+      "loss": 3.8796,
+      "step": 2277376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2704006189877508e-05,
+      "loss": 3.8837,
+      "step": 2277888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2695620242366988e-05,
+      "loss": 3.8806,
+      "step": 2278400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2687234294856468e-05,
+      "loss": 3.8828,
+      "step": 2278912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2678848347345946e-05,
+      "loss": 3.8767,
+      "step": 2279424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2670462399835426e-05,
+      "loss": 3.8967,
+      "step": 2279936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2662092831128638e-05,
+      "loss": 3.8919,
+      "step": 2280448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2653723262421849e-05,
+      "loss": 3.8822,
+      "step": 2280960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2645337314911329e-05,
+      "loss": 3.8826,
+      "step": 2281472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2636951367400809e-05,
+      "loss": 3.8922,
+      "step": 2281984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2628565419890289e-05,
+      "loss": 3.8833,
+      "step": 2282496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.26201958511835e-05,
+      "loss": 3.885,
+      "step": 2283008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.261180990367298e-05,
+      "loss": 3.8887,
+      "step": 2283520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2603423956162461e-05,
+      "loss": 3.8979,
+      "step": 2284032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.259505438745567e-05,
+      "loss": 3.8925,
+      "step": 2284544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.258666843994515e-05,
+      "loss": 3.8906,
+      "step": 2285056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2578282492434632e-05,
+      "loss": 3.8832,
+      "step": 2285568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2569896544924112e-05,
+      "loss": 3.8949,
+      "step": 2286080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2561510597413592e-05,
+      "loss": 3.8853,
+      "step": 2286592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.255312464990307e-05,
+      "loss": 3.8774,
+      "step": 2287104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.254473870239255e-05,
+      "loss": 3.8834,
+      "step": 2287616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.253635275488203e-05,
+      "loss": 3.8714,
+      "step": 2288128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2527983186175243e-05,
+      "loss": 3.8886,
+      "step": 2288640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.2519613617468453e-05,
+      "loss": 3.8802,
+      "step": 2289152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.991286516189575,
+      "eval_runtime": 302.2399,
+      "eval_samples_per_second": 1262.544,
+      "eval_steps_per_second": 39.455,
+      "step": 2289584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2511227669957933e-05,
+      "loss": 3.8983,
+      "step": 2289664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2502841722447415e-05,
+      "loss": 3.8795,
+      "step": 2290176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2494455774936893e-05,
+      "loss": 3.8801,
+      "step": 2290688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2486069827426373e-05,
+      "loss": 3.8971,
+      "step": 2291200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2477683879915853e-05,
+      "loss": 3.8887,
+      "step": 2291712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2469297932405333e-05,
+      "loss": 3.8971,
+      "step": 2292224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2460911984894813e-05,
+      "loss": 3.8839,
+      "step": 2292736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2452526037384291e-05,
+      "loss": 3.8789,
+      "step": 2293248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2444140089873773e-05,
+      "loss": 3.8782,
+      "step": 2293760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2435754142363253e-05,
+      "loss": 3.8875,
+      "step": 2294272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2427368194852733e-05,
+      "loss": 3.8827,
+      "step": 2294784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2418982247342211e-05,
+      "loss": 3.8835,
+      "step": 2295296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2410612678635424e-05,
+      "loss": 3.8949,
+      "step": 2295808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2402226731124904e-05,
+      "loss": 3.8734,
+      "step": 2296320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2393840783614384e-05,
+      "loss": 3.8779,
+      "step": 2296832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2385454836103864e-05,
+      "loss": 3.8705,
+      "step": 2297344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2377068888593343e-05,
+      "loss": 3.8837,
+      "step": 2297856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2368682941082822e-05,
+      "loss": 3.8848,
+      "step": 2298368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2360296993572303e-05,
+      "loss": 3.878,
+      "step": 2298880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2351911046061783e-05,
+      "loss": 3.8883,
+      "step": 2299392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2343525098551263e-05,
+      "loss": 3.8918,
+      "step": 2299904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2335139151040742e-05,
+      "loss": 3.886,
+      "step": 2300416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2326753203530223e-05,
+      "loss": 3.8861,
+      "step": 2300928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2318383634823434e-05,
+      "loss": 3.884,
+      "step": 2301440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2309997687312912e-05,
+      "loss": 3.8966,
+      "step": 2301952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2301611739802394e-05,
+      "loss": 3.8844,
+      "step": 2302464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2293225792291874e-05,
+      "loss": 3.8798,
+      "step": 2302976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2284839844781354e-05,
+      "loss": 3.8858,
+      "step": 2303488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2276453897270832e-05,
+      "loss": 3.8743,
+      "step": 2304000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2268067949760314e-05,
+      "loss": 3.8827,
+      "step": 2304512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2259682002249794e-05,
+      "loss": 3.8821,
+      "step": 2305024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2251328812346737e-05,
+      "loss": 3.8859,
+      "step": 2305536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2242942864836215e-05,
+      "loss": 3.8887,
+      "step": 2306048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2234556917325695e-05,
+      "loss": 3.8939,
+      "step": 2306560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2226170969815177e-05,
+      "loss": 3.8851,
+      "step": 2307072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2217801401108386e-05,
+      "loss": 3.89,
+      "step": 2307584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2209415453597866e-05,
+      "loss": 3.8902,
+      "step": 2308096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2201029506087348e-05,
+      "loss": 3.8744,
+      "step": 2308608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2192643558576828e-05,
+      "loss": 3.8801,
+      "step": 2309120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2184257611066306e-05,
+      "loss": 3.8705,
+      "step": 2309632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2175871663555786e-05,
+      "loss": 3.8787,
+      "step": 2310144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2167485716045267e-05,
+      "loss": 3.8798,
+      "step": 2310656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2159099768534746e-05,
+      "loss": 3.8969,
+      "step": 2311168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2150713821024226e-05,
+      "loss": 3.8731,
+      "step": 2311680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2142344252317438e-05,
+      "loss": 3.8806,
+      "step": 2312192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2133958304806918e-05,
+      "loss": 3.8838,
+      "step": 2312704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2125572357296396e-05,
+      "loss": 3.8821,
+      "step": 2313216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2117186409785878e-05,
+      "loss": 3.8792,
+      "step": 2313728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2108800462275358e-05,
+      "loss": 3.8623,
+      "step": 2314240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2100430893568567e-05,
+      "loss": 3.8836,
+      "step": 2314752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2092044946058049e-05,
+      "loss": 3.8958,
+      "step": 2315264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2083658998547529e-05,
+      "loss": 3.8902,
+      "step": 2315776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2075273051037009e-05,
+      "loss": 3.8782,
+      "step": 2316288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2066887103526489e-05,
+      "loss": 3.8769,
+      "step": 2316800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2058501156015968e-05,
+      "loss": 3.8906,
+      "step": 2317312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2050115208505448e-05,
+      "loss": 3.8674,
+      "step": 2317824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2041729260994927e-05,
+      "loss": 3.8954,
+      "step": 2318336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.203335969228814e-05,
+      "loss": 3.8692,
+      "step": 2318848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2024973744777619e-05,
+      "loss": 3.902,
+      "step": 2319360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2016587797267099e-05,
+      "loss": 3.8852,
+      "step": 2319872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.200821822856031e-05,
+      "loss": 3.8648,
+      "step": 2320384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.199983228104979e-05,
+      "loss": 3.8812,
+      "step": 2320896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.199144633353927e-05,
+      "loss": 3.8841,
+      "step": 2321408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.198306038602875e-05,
+      "loss": 3.8685,
+      "step": 2321920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.197469081732196e-05,
+      "loss": 3.8732,
+      "step": 2322432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1966321248615173e-05,
+      "loss": 3.8903,
+      "step": 2322944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1957935301104653e-05,
+      "loss": 3.8808,
+      "step": 2323456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1949549353594131e-05,
+      "loss": 3.878,
+      "step": 2323968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1941163406083613e-05,
+      "loss": 3.8628,
+      "step": 2324480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1932777458573093e-05,
+      "loss": 3.878,
+      "step": 2324992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1924391511062573e-05,
+      "loss": 3.8699,
+      "step": 2325504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1916005563552051e-05,
+      "loss": 3.8882,
+      "step": 2326016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1907619616041533e-05,
+      "loss": 3.8729,
+      "step": 2326528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1899233668531013e-05,
+      "loss": 3.8849,
+      "step": 2327040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1890864099824223e-05,
+      "loss": 3.891,
+      "step": 2327552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1882478152313703e-05,
+      "loss": 3.8841,
+      "step": 2328064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1874092204803183e-05,
+      "loss": 3.8794,
+      "step": 2328576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1865722636096394e-05,
+      "loss": 3.8809,
+      "step": 2329088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1857336688585874e-05,
+      "loss": 3.8782,
+      "step": 2329600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1848950741075354e-05,
+      "loss": 3.8857,
+      "step": 2330112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1840564793564834e-05,
+      "loss": 3.8774,
+      "step": 2330624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1832178846054314e-05,
+      "loss": 3.8805,
+      "step": 2331136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1823792898543794e-05,
+      "loss": 3.8877,
+      "step": 2331648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1815406951033274e-05,
+      "loss": 3.8816,
+      "step": 2332160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1807021003522754e-05,
+      "loss": 3.8635,
+      "step": 2332672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1798635056012234e-05,
+      "loss": 3.883,
+      "step": 2333184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1790249108501714e-05,
+      "loss": 3.8761,
+      "step": 2333696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1781863160991194e-05,
+      "loss": 3.8809,
+      "step": 2334208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1773493592284405e-05,
+      "loss": 3.8783,
+      "step": 2334720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1765124023577615e-05,
+      "loss": 3.8831,
+      "step": 2335232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1756738076067097e-05,
+      "loss": 3.8691,
+      "step": 2335744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1748352128556577e-05,
+      "loss": 3.8734,
+      "step": 2336256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1739966181046055e-05,
+      "loss": 3.8758,
+      "step": 2336768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1731580233535535e-05,
+      "loss": 3.8901,
+      "step": 2337280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1723194286025017e-05,
+      "loss": 3.8815,
+      "step": 2337792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1714808338514495e-05,
+      "loss": 3.8877,
+      "step": 2338304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1706422391003975e-05,
+      "loss": 3.8755,
+      "step": 2338816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1698052822297187e-05,
+      "loss": 3.8762,
+      "step": 2339328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1689666874786667e-05,
+      "loss": 3.8759,
+      "step": 2339840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1681280927276146e-05,
+      "loss": 3.886,
+      "step": 2340352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1672894979765626e-05,
+      "loss": 3.8818,
+      "step": 2340864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1664509032255107e-05,
+      "loss": 3.8839,
+      "step": 2341376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1656123084744586e-05,
+      "loss": 3.8796,
+      "step": 2341888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1647737137234066e-05,
+      "loss": 3.8751,
+      "step": 2342400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1639351189723547e-05,
+      "loss": 3.887,
+      "step": 2342912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1630981621016758e-05,
+      "loss": 3.8726,
+      "step": 2343424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1622612052309969e-05,
+      "loss": 3.8742,
+      "step": 2343936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1614226104799449e-05,
+      "loss": 3.8794,
+      "step": 2344448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1605840157288929e-05,
+      "loss": 3.8724,
+      "step": 2344960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1597454209778409e-05,
+      "loss": 3.8778,
+      "step": 2345472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1589068262267889e-05,
+      "loss": 3.8758,
+      "step": 2345984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.15806986935611e-05,
+      "loss": 3.8683,
+      "step": 2346496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.157231274605058e-05,
+      "loss": 3.8827,
+      "step": 2347008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.156392679854006e-05,
+      "loss": 3.8804,
+      "step": 2347520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.155554085102954e-05,
+      "loss": 3.8761,
+      "step": 2348032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.154715490351902e-05,
+      "loss": 3.8904,
+      "step": 2348544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1538768956008499e-05,
+      "loss": 3.8955,
+      "step": 2349056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1530383008497979e-05,
+      "loss": 3.8812,
+      "step": 2349568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.152201343979119e-05,
+      "loss": 3.8792,
+      "step": 2350080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1513627492280672e-05,
+      "loss": 3.8729,
+      "step": 2350592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.150524154477015e-05,
+      "loss": 3.8818,
+      "step": 2351104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.149685559725963e-05,
+      "loss": 3.8749,
+      "step": 2351616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.148846964974911e-05,
+      "loss": 3.8815,
+      "step": 2352128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1480100081042322e-05,
+      "loss": 3.8763,
+      "step": 2352640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.14717141335318e-05,
+      "loss": 3.8903,
+      "step": 2353152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1463344564825013e-05,
+      "loss": 3.8781,
+      "step": 2353664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1454958617314493e-05,
+      "loss": 3.8771,
+      "step": 2354176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1446572669803971e-05,
+      "loss": 3.8791,
+      "step": 2354688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1438186722293453e-05,
+      "loss": 3.8791,
+      "step": 2355200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1429800774782933e-05,
+      "loss": 3.8715,
+      "step": 2355712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1421414827272413e-05,
+      "loss": 3.8908,
+      "step": 2356224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1413028879761891e-05,
+      "loss": 3.8911,
+      "step": 2356736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1404642932251373e-05,
+      "loss": 3.8813,
+      "step": 2357248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1396256984740853e-05,
+      "loss": 3.8771,
+      "step": 2357760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.138787103723033e-05,
+      "loss": 3.8851,
+      "step": 2358272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1379501468523543e-05,
+      "loss": 3.8812,
+      "step": 2358784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1371115521013023e-05,
+      "loss": 3.8835,
+      "step": 2359296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1362745952306234e-05,
+      "loss": 3.8836,
+      "step": 2359808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1354360004795714e-05,
+      "loss": 3.8975,
+      "step": 2360320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1345974057285194e-05,
+      "loss": 3.8849,
+      "step": 2360832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1337588109774674e-05,
+      "loss": 3.8898,
+      "step": 2361344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1329202162264154e-05,
+      "loss": 3.8819,
+      "step": 2361856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1320816214753634e-05,
+      "loss": 3.8883,
+      "step": 2362368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1312446646046845e-05,
+      "loss": 3.8783,
+      "step": 2362880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1304060698536326e-05,
+      "loss": 3.8753,
+      "step": 2363392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1295674751025805e-05,
+      "loss": 3.8815,
+      "step": 2363904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1287288803515285e-05,
+      "loss": 3.8684,
+      "step": 2364416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1278902856004764e-05,
+      "loss": 3.8839,
+      "step": 2364928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1270533287297977e-05,
+      "loss": 3.8797,
+      "step": 2365440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.990168809890747,
+      "eval_runtime": 294.8755,
+      "eval_samples_per_second": 1294.075,
+      "eval_steps_per_second": 40.441,
+      "step": 2365904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1262147339787455e-05,
+      "loss": 3.8825,
+      "step": 2365952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1253761392276937e-05,
+      "loss": 3.8817,
+      "step": 2366464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1245375444766417e-05,
+      "loss": 3.8753,
+      "step": 2366976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1236989497255895e-05,
+      "loss": 3.8906,
+      "step": 2367488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1228603549745375e-05,
+      "loss": 3.8868,
+      "step": 2368000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1220217602234857e-05,
+      "loss": 3.8939,
+      "step": 2368512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1211831654724335e-05,
+      "loss": 3.8806,
+      "step": 2369024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1203462086017546e-05,
+      "loss": 3.8788,
+      "step": 2369536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1195076138507027e-05,
+      "loss": 3.873,
+      "step": 2370048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1186690190996507e-05,
+      "loss": 3.8839,
+      "step": 2370560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1178304243485986e-05,
+      "loss": 3.8797,
+      "step": 2371072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1169918295975467e-05,
+      "loss": 3.8835,
+      "step": 2371584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1161548727268678e-05,
+      "loss": 3.8881,
+      "step": 2372096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.115317915856189e-05,
+      "loss": 3.8747,
+      "step": 2372608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1144793211051369e-05,
+      "loss": 3.8723,
+      "step": 2373120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1136407263540849e-05,
+      "loss": 3.8705,
+      "step": 2373632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1128021316030329e-05,
+      "loss": 3.8773,
+      "step": 2374144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1119635368519809e-05,
+      "loss": 3.8812,
+      "step": 2374656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1111249421009289e-05,
+      "loss": 3.8786,
+      "step": 2375168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1102863473498769e-05,
+      "loss": 3.879,
+      "step": 2375680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1094477525988249e-05,
+      "loss": 3.8916,
+      "step": 2376192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1086091578477728e-05,
+      "loss": 3.8823,
+      "step": 2376704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1077705630967208e-05,
+      "loss": 3.8847,
+      "step": 2377216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1069319683456688e-05,
+      "loss": 3.8804,
+      "step": 2377728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.10609501147499e-05,
+      "loss": 3.8942,
+      "step": 2378240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1052564167239379e-05,
+      "loss": 3.8764,
+      "step": 2378752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1044178219728859e-05,
+      "loss": 3.8796,
+      "step": 2379264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1035808651022072e-05,
+      "loss": 3.8761,
+      "step": 2379776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.102742270351155e-05,
+      "loss": 3.8784,
+      "step": 2380288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.101903675600103e-05,
+      "loss": 3.8775,
+      "step": 2380800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.1010650808490511e-05,
+      "loss": 3.8795,
+      "step": 2381312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.100226486097999e-05,
+      "loss": 3.8835,
+      "step": 2381824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.099387891346947e-05,
+      "loss": 3.8848,
+      "step": 2382336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.098549296595895e-05,
+      "loss": 3.8894,
+      "step": 2382848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0977139776055893e-05,
+      "loss": 3.8801,
+      "step": 2383360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0968770207349104e-05,
+      "loss": 3.8866,
+      "step": 2383872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0960384259838584e-05,
+      "loss": 3.8866,
+      "step": 2384384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0951998312328064e-05,
+      "loss": 3.872,
+      "step": 2384896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0943612364817545e-05,
+      "loss": 3.8781,
+      "step": 2385408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0935226417307024e-05,
+      "loss": 3.8658,
+      "step": 2385920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0926840469796503e-05,
+      "loss": 3.8772,
+      "step": 2386432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0918454522285983e-05,
+      "loss": 3.875,
+      "step": 2386944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0910068574775463e-05,
+      "loss": 3.8954,
+      "step": 2387456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0901682627264943e-05,
+      "loss": 3.8732,
+      "step": 2387968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0893313058558154e-05,
+      "loss": 3.8746,
+      "step": 2388480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0884927111047636e-05,
+      "loss": 3.8776,
+      "step": 2388992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0876541163537114e-05,
+      "loss": 3.883,
+      "step": 2389504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0868155216026594e-05,
+      "loss": 3.8782,
+      "step": 2390016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0859769268516076e-05,
+      "loss": 3.8593,
+      "step": 2390528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0851383321005554e-05,
+      "loss": 3.8775,
+      "step": 2391040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0842997373495034e-05,
+      "loss": 3.8961,
+      "step": 2391552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0834627804788246e-05,
+      "loss": 3.8832,
+      "step": 2392064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0826241857277726e-05,
+      "loss": 3.8766,
+      "step": 2392576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0817855909767205e-05,
+      "loss": 3.8709,
+      "step": 2393088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0809469962256685e-05,
+      "loss": 3.886,
+      "step": 2393600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0801084014746166e-05,
+      "loss": 3.8645,
+      "step": 2394112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0792698067235644e-05,
+      "loss": 3.8923,
+      "step": 2394624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0784328498528857e-05,
+      "loss": 3.8649,
+      "step": 2395136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0775942551018337e-05,
+      "loss": 3.8965,
+      "step": 2395648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0767556603507817e-05,
+      "loss": 3.8841,
+      "step": 2396160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0759170655997295e-05,
+      "loss": 3.8633,
+      "step": 2396672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0750784708486777e-05,
+      "loss": 3.8752,
+      "step": 2397184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0742398760976257e-05,
+      "loss": 3.8867,
+      "step": 2397696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0734012813465735e-05,
+      "loss": 3.8576,
+      "step": 2398208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0725626865955215e-05,
+      "loss": 3.8702,
+      "step": 2398720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0717257297248427e-05,
+      "loss": 3.8899,
+      "step": 2399232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0708871349737907e-05,
+      "loss": 3.8773,
+      "step": 2399744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0700485402227387e-05,
+      "loss": 3.8732,
+      "step": 2400256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0692099454716867e-05,
+      "loss": 3.8607,
+      "step": 2400768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0683729886010078e-05,
+      "loss": 3.873,
+      "step": 2401280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0675343938499558e-05,
+      "loss": 3.8681,
+      "step": 2401792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0666957990989038e-05,
+      "loss": 3.8853,
+      "step": 2402304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0658572043478518e-05,
+      "loss": 3.8695,
+      "step": 2402816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0650186095967998e-05,
+      "loss": 3.8829,
+      "step": 2403328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0641800148457478e-05,
+      "loss": 3.8835,
+      "step": 2403840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0633414200946958e-05,
+      "loss": 3.8819,
+      "step": 2404352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0625044632240169e-05,
+      "loss": 3.8771,
+      "step": 2404864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0616658684729649e-05,
+      "loss": 3.8775,
+      "step": 2405376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.060828911602286e-05,
+      "loss": 3.8812,
+      "step": 2405888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.059990316851234e-05,
+      "loss": 3.8809,
+      "step": 2406400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0591517221001821e-05,
+      "loss": 3.8725,
+      "step": 2406912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.05831312734913e-05,
+      "loss": 3.8776,
+      "step": 2407424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.057474532598078e-05,
+      "loss": 3.8803,
+      "step": 2407936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0566359378470259e-05,
+      "loss": 3.8812,
+      "step": 2408448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0557973430959739e-05,
+      "loss": 3.8591,
+      "step": 2408960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0549587483449219e-05,
+      "loss": 3.8774,
+      "step": 2409472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0541201535938699e-05,
+      "loss": 3.8755,
+      "step": 2409984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0532815588428179e-05,
+      "loss": 3.8793,
+      "step": 2410496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.0524429640917659e-05,
+      "loss": 3.8764,
+      "step": 2411008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.051606007221087e-05,
+      "loss": 3.8745,
+      "step": 2411520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0507674124700351e-05,
+      "loss": 3.8692,
+      "step": 2412032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.049928817718983e-05,
+      "loss": 3.8667,
+      "step": 2412544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.049090222967931e-05,
+      "loss": 3.8701,
+      "step": 2413056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.048251628216879e-05,
+      "loss": 3.8872,
+      "step": 2413568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0474146713462002e-05,
+      "loss": 3.8776,
+      "step": 2414080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.046576076595148e-05,
+      "loss": 3.8858,
+      "step": 2414592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0457374818440962e-05,
+      "loss": 3.8724,
+      "step": 2415104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0448988870930442e-05,
+      "loss": 3.8724,
+      "step": 2415616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.044060292341992e-05,
+      "loss": 3.8715,
+      "step": 2416128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.04322169759094e-05,
+      "loss": 3.8855,
+      "step": 2416640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0423847407202613e-05,
+      "loss": 3.8776,
+      "step": 2417152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0415461459692093e-05,
+      "loss": 3.88,
+      "step": 2417664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.040707551218157e-05,
+      "loss": 3.879,
+      "step": 2418176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0398689564671052e-05,
+      "loss": 3.8674,
+      "step": 2418688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0390303617160532e-05,
+      "loss": 3.889,
+      "step": 2419200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0381934048453743e-05,
+      "loss": 3.8666,
+      "step": 2419712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0373548100943223e-05,
+      "loss": 3.8709,
+      "step": 2420224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0365162153432703e-05,
+      "loss": 3.8796,
+      "step": 2420736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0356776205922183e-05,
+      "loss": 3.8716,
+      "step": 2421248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0348406637215394e-05,
+      "loss": 3.8714,
+      "step": 2421760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0340020689704874e-05,
+      "loss": 3.8745,
+      "step": 2422272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0331634742194354e-05,
+      "loss": 3.867,
+      "step": 2422784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0323248794683834e-05,
+      "loss": 3.8772,
+      "step": 2423296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0314862847173314e-05,
+      "loss": 3.874,
+      "step": 2423808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0306476899662794e-05,
+      "loss": 3.8779,
+      "step": 2424320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0298107330956006e-05,
+      "loss": 3.8826,
+      "step": 2424832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0289721383445484e-05,
+      "loss": 3.8931,
+      "step": 2425344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0281335435934964e-05,
+      "loss": 3.88,
+      "step": 2425856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0272949488424444e-05,
+      "loss": 3.8744,
+      "step": 2426368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0264563540913924e-05,
+      "loss": 3.8722,
+      "step": 2426880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0256177593403404e-05,
+      "loss": 3.8763,
+      "step": 2427392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0247791645892884e-05,
+      "loss": 3.8733,
+      "step": 2427904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0239422077186097e-05,
+      "loss": 3.8738,
+      "step": 2428416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0231036129675575e-05,
+      "loss": 3.8725,
+      "step": 2428928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0222650182165055e-05,
+      "loss": 3.8843,
+      "step": 2429440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0214264234654536e-05,
+      "loss": 3.8737,
+      "step": 2429952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0205878287144015e-05,
+      "loss": 3.8793,
+      "step": 2430464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0197492339633495e-05,
+      "loss": 3.8723,
+      "step": 2430976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0189106392122975e-05,
+      "loss": 3.8726,
+      "step": 2431488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0180720444612456e-05,
+      "loss": 3.8724,
+      "step": 2432000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0172367254709398e-05,
+      "loss": 3.8867,
+      "step": 2432512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0163981307198878e-05,
+      "loss": 3.8899,
+      "step": 2433024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0155595359688358e-05,
+      "loss": 3.882,
+      "step": 2433536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0147209412177838e-05,
+      "loss": 3.8695,
+      "step": 2434048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0138823464667318e-05,
+      "loss": 3.8781,
+      "step": 2434560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0130437517156798e-05,
+      "loss": 3.8811,
+      "step": 2435072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0122051569646278e-05,
+      "loss": 3.8797,
+      "step": 2435584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0113682000939488e-05,
+      "loss": 3.8837,
+      "step": 2436096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0105296053428968e-05,
+      "loss": 3.888,
+      "step": 2436608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0096910105918448e-05,
+      "loss": 3.8825,
+      "step": 2437120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0088524158407928e-05,
+      "loss": 3.889,
+      "step": 2437632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0080138210897408e-05,
+      "loss": 3.8744,
+      "step": 2438144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0071752263386888e-05,
+      "loss": 3.8843,
+      "step": 2438656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0063366315876368e-05,
+      "loss": 3.8778,
+      "step": 2439168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0054980368365848e-05,
+      "loss": 3.8705,
+      "step": 2439680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0046610799659059e-05,
+      "loss": 3.8823,
+      "step": 2440192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.0038224852148539e-05,
+      "loss": 3.8655,
+      "step": 2440704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.002983890463802e-05,
+      "loss": 3.8777,
+      "step": 2441216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.002146933593123e-05,
+      "loss": 3.8769,
+      "step": 2441728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.9891669750213623,
+      "eval_runtime": 288.2875,
+      "eval_samples_per_second": 1323.647,
+      "eval_steps_per_second": 41.365,
+      "step": 2442224
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 9.894877250222725e+17,
+  "trial_name": null,
+  "trial_params": null
+}