diff --git "a/checkpoint-457914/trainer_state.json" "b/checkpoint-457914/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-457914/trainer_state.json"
@@ -0,0 +1,5437 @@
+{
+  "best_metric": 4.158684730529785,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/3/checkpoints/checkpoint-457914",
+  "epoch": 1.0250002784396635,
+  "eval_steps": 10,
+  "global_step": 457914,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8183,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5661,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0563,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.985,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.956,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8543,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994970069374061e-05,
+      "loss": 6.7253,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.6189,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5339,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4422,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3916,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3209,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.253,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1866,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1383,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0974,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0514,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9997,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9645,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9268,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8802,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8432,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8191,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7811,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7597,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7314,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7082,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6786,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6449,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.638,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6037,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5886,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5626,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5594,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5281,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5185,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5051,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4855,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.475,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4524,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4292,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4232,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.3952,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.408,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3798,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3763,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3602,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.358,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3509,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3221,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3108,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3022,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3086,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.2857,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.2924,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.263,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.26,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2459,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.241,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2379,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2178,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2112,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2158,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.1943,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1926,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.1823,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.1801,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9438288926028735e-05,
+      "loss": 5.1659,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9429902978518215e-05,
+      "loss": 5.1684,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1552,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.146,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404761514790384e-05,
+      "loss": 5.148,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.939637556727987e-05,
+      "loss": 5.134,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1416,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1388,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.129,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1101,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.0947,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.0881,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0927,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0953,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0905,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0644,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0589,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0672,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0535,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279021438543785e-05,
+      "loss": 5.0529,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0443,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0457,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.035,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0266,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.0134,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922872213228439e-05,
+      "loss": 5.0257,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922033618477387e-05,
+      "loss": 5.0113,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921195023726335e-05,
+      "loss": 4.9951,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920356428975283e-05,
+      "loss": 4.995,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919517834224231e-05,
+      "loss": 5.007,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918680877353552e-05,
+      "loss": 4.9967,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917842282602501e-05,
+      "loss": 4.9875,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917003687851449e-05,
+      "loss": 4.9758,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916165093100397e-05,
+      "loss": 4.9797,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9595,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914489541478666e-05,
+      "loss": 4.9742,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913650946727614e-05,
+      "loss": 4.9756,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912812351976562e-05,
+      "loss": 4.9512,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911973757225509e-05,
+      "loss": 4.9452,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.922,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910298205603779e-05,
+      "loss": 4.9337,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9349,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9289,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9306,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9136,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9153,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.917,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9185,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9047,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.8931,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.8959,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.8983,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002383443519005e-05,
+      "loss": 4.8944,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.8881,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.8778,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897724197979118e-05,
+      "loss": 4.8712,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.8642,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896047008477014e-05,
+      "loss": 4.8712,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895208413725962e-05,
+      "loss": 4.8685,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89436981897491e-05,
+      "loss": 4.8707,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893531224223858e-05,
+      "loss": 4.8451,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892692629472806e-05,
+      "loss": 4.8582,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891854034721754e-05,
+      "loss": 4.8587,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891017077851075e-05,
+      "loss": 4.8442,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890178483100023e-05,
+      "loss": 4.8594,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889339888348971e-05,
+      "loss": 4.8397,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.8539,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766433672724e-05,
+      "loss": 4.8418,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886825741976188e-05,
+      "loss": 4.8217,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859871472251365e-05,
+      "loss": 4.8207,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851485524740845e-05,
+      "loss": 4.8202,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843099577230325e-05,
+      "loss": 4.823,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834713629719805e-05,
+      "loss": 4.8275,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8246,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8817958113502494e-05,
+      "loss": 4.8196,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809572165991974e-05,
+      "loss": 4.8071,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801186218481454e-05,
+      "loss": 4.7964,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879280027097093e-05,
+      "loss": 4.7952,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878441432346041e-05,
+      "loss": 4.805,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877604475475362e-05,
+      "loss": 4.8102,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87676588072431e-05,
+      "loss": 4.8008,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875927285973258e-05,
+      "loss": 4.7837,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875088691222206e-05,
+      "loss": 4.7832,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.748494625091553,
+      "eval_runtime": 552.6156,
+      "eval_samples_per_second": 690.518,
+      "eval_steps_per_second": 21.579,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874250096471154e-05,
+      "loss": 4.7924,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873411501720102e-05,
+      "loss": 4.7847,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87257290696905e-05,
+      "loss": 4.7762,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871734312217998e-05,
+      "loss": 4.7744,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870895717466946e-05,
+      "loss": 4.7644,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870057122715894e-05,
+      "loss": 4.7505,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869218527964842e-05,
+      "loss": 4.7583,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86837993321379e-05,
+      "loss": 4.7679,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867541338462738e-05,
+      "loss": 4.75,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866702743711686e-05,
+      "loss": 4.7578,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865864148960634e-05,
+      "loss": 4.7612,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8650255542095815e-05,
+      "loss": 4.761,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86418695945853e-05,
+      "loss": 4.7497,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863348364707478e-05,
+      "loss": 4.7469,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862509769956426e-05,
+      "loss": 4.7446,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861671175205374e-05,
+      "loss": 4.7335,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860832580454322e-05,
+      "loss": 4.7369,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85999398570327e-05,
+      "loss": 4.7371,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.859155390952218e-05,
+      "loss": 4.7361,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.858320071961912e-05,
+      "loss": 4.7337,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85748147721086e-05,
+      "loss": 4.7221,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.856642882459808e-05,
+      "loss": 4.7341,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.855804287708756e-05,
+      "loss": 4.7204,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854967330838077e-05,
+      "loss": 4.7285,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541287360870256e-05,
+      "loss": 4.7185,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532901413359736e-05,
+      "loss": 4.7171,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524515465849216e-05,
+      "loss": 4.7203,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516145897142425e-05,
+      "loss": 4.7019,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850777632843564e-05,
+      "loss": 4.7161,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8499390380925114e-05,
+      "loss": 4.6952,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8491004433414594e-05,
+      "loss": 4.7013,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8482618485904074e-05,
+      "loss": 4.698,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8474232538393554e-05,
+      "loss": 4.7058,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8465846590883034e-05,
+      "loss": 4.6932,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8457460643372514e-05,
+      "loss": 4.6863,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8449074695861993e-05,
+      "loss": 4.7029,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8440688748351473e-05,
+      "loss": 4.6903,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843230280084095e-05,
+      "loss": 4.6971,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842393323213417e-05,
+      "loss": 4.6849,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841554728462365e-05,
+      "loss": 4.6705,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840716133711313e-05,
+      "loss": 4.6734,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839877538960261e-05,
+      "loss": 4.6661,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839038944209209e-05,
+      "loss": 4.6811,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83820198733853e-05,
+      "loss": 4.6713,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837363392587478e-05,
+      "loss": 4.6767,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836524797836426e-05,
+      "loss": 4.6682,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835686203085374e-05,
+      "loss": 4.6741,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834847608334322e-05,
+      "loss": 4.6696,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83400901358327e-05,
+      "loss": 4.6476,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833170418832218e-05,
+      "loss": 4.651,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832331824081166e-05,
+      "loss": 4.6494,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8314948672104874e-05,
+      "loss": 4.6627,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8306562724594354e-05,
+      "loss": 4.6403,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829819315588756e-05,
+      "loss": 4.6644,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828980720837704e-05,
+      "loss": 4.6381,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828142126086652e-05,
+      "loss": 4.6386,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273035313356e-05,
+      "loss": 4.6402,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8264649365845476e-05,
+      "loss": 4.6369,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8256263418334956e-05,
+      "loss": 4.639,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8247877470824436e-05,
+      "loss": 4.6302,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8239491523313916e-05,
+      "loss": 4.6264,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231121954607125e-05,
+      "loss": 4.6386,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.822273600709661e-05,
+      "loss": 4.624,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.821435005958609e-05,
+      "loss": 4.6222,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.820596411207557e-05,
+      "loss": 4.6226,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819759454336878e-05,
+      "loss": 4.6268,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818920859585826e-05,
+      "loss": 4.6113,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818082264834774e-05,
+      "loss": 4.6301,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817243670083722e-05,
+      "loss": 4.6165,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816406713213043e-05,
+      "loss": 4.6144,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815568118461991e-05,
+      "loss": 4.6209,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814729523710939e-05,
+      "loss": 4.6083,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813890928959887e-05,
+      "loss": 4.6265,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813053972089208e-05,
+      "loss": 4.6297,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8122153773381565e-05,
+      "loss": 4.623,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8113767825871045e-05,
+      "loss": 4.604,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8105381878360525e-05,
+      "loss": 4.5988,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8096995930850005e-05,
+      "loss": 4.5934,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8088609983339485e-05,
+      "loss": 4.6041,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8080240414632694e-05,
+      "loss": 4.6161,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8071854467122174e-05,
+      "loss": 4.6076,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8063468519611654e-05,
+      "loss": 4.5948,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8055082572101134e-05,
+      "loss": 4.5829,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8046696624590614e-05,
+      "loss": 4.6014,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8038310677080094e-05,
+      "loss": 4.5865,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8029924729569574e-05,
+      "loss": 4.5891,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8021538782059054e-05,
+      "loss": 4.5909,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801316921335226e-05,
+      "loss": 4.5991,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800479964464548e-05,
+      "loss": 4.587,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799641369713496e-05,
+      "loss": 4.5844,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798802774962444e-05,
+      "loss": 4.5752,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797964180211392e-05,
+      "loss": 4.5873,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797127223340713e-05,
+      "loss": 4.5863,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796288628589661e-05,
+      "loss": 4.5621,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795450033838609e-05,
+      "loss": 4.5711,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794611439087557e-05,
+      "loss": 4.5864,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793772844336505e-05,
+      "loss": 4.5867,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792934249585453e-05,
+      "loss": 4.5735,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792095654834401e-05,
+      "loss": 4.5675,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791257060083349e-05,
+      "loss": 4.5785,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7904201032126697e-05,
+      "loss": 4.5568,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789581508461618e-05,
+      "loss": 4.5744,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788744551590939e-05,
+      "loss": 4.5775,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787905956839887e-05,
+      "loss": 4.5649,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787067362088835e-05,
+      "loss": 4.5643,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786228767337783e-05,
+      "loss": 4.5412,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.785390172586731e-05,
+      "loss": 4.5464,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845515778356785e-05,
+      "loss": 4.5531,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837129830846265e-05,
+      "loss": 4.5553,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782876026213948e-05,
+      "loss": 4.5584,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782037431462896e-05,
+      "loss": 4.5463,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7811988367118434e-05,
+      "loss": 4.5451,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780360241960792e-05,
+      "loss": 4.5519,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77952164720974e-05,
+      "loss": 4.558,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778683052458688e-05,
+      "loss": 4.546,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777844457707636e-05,
+      "loss": 4.5458,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777005862956584e-05,
+      "loss": 4.5433,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776168906085905e-05,
+      "loss": 4.5465,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775330311334853e-05,
+      "loss": 4.547,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774491716583801e-05,
+      "loss": 4.5395,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773654759713122e-05,
+      "loss": 4.5382,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77281616496207e-05,
+      "loss": 4.5322,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771977570211018e-05,
+      "loss": 4.5274,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771138975459966e-05,
+      "loss": 4.5335,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770300380708914e-05,
+      "loss": 4.5363,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769461785957862e-05,
+      "loss": 4.5385,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686231912068105e-05,
+      "loss": 4.5224,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7677862343361315e-05,
+      "loss": 4.5285,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669476395850795e-05,
+      "loss": 4.5403,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661090448340274e-05,
+      "loss": 4.5251,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7652704500829754e-05,
+      "loss": 4.5384,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7644318553319234e-05,
+      "loss": 4.523,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7635932605808714e-05,
+      "loss": 4.541,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7627546658298194e-05,
+      "loss": 4.5275,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7619160710787674e-05,
+      "loss": 4.5116,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761079114208088e-05,
+      "loss": 4.5152,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760242157337409e-05,
+      "loss": 4.5125,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759403562586357e-05,
+      "loss": 4.5231,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758566605715679e-05,
+      "loss": 4.5229,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757728010964627e-05,
+      "loss": 4.5252,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756889416213575e-05,
+      "loss": 4.52,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756050821462523e-05,
+      "loss": 4.5111,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755212226711471e-05,
+      "loss": 4.506,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754373631960419e-05,
+      "loss": 4.5029,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.753535037209367e-05,
+      "loss": 4.5134,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752696442458315e-05,
+      "loss": 4.5265,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751857847707262e-05,
+      "loss": 4.5096,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751020890836584e-05,
+      "loss": 4.5027,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.750182296085532e-05,
+      "loss": 4.5027,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.47078800201416,
+      "eval_runtime": 551.3077,
+      "eval_samples_per_second": 692.156,
+      "eval_steps_per_second": 21.63,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74934370133448e-05,
+      "loss": 4.5138,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748505106583428e-05,
+      "loss": 4.5086,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747666511832376e-05,
+      "loss": 4.4956,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746827917081324e-05,
+      "loss": 4.4975,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745989322330272e-05,
+      "loss": 4.4939,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74515072757922e-05,
+      "loss": 4.479,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744312132828168e-05,
+      "loss": 4.4885,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743473538077116e-05,
+      "loss": 4.5019,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426349433260637e-05,
+      "loss": 4.4809,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417963485750117e-05,
+      "loss": 4.4904,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409577538239597e-05,
+      "loss": 4.5005,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401191590729076e-05,
+      "loss": 4.4998,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392805643218556e-05,
+      "loss": 4.4867,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384419695708036e-05,
+      "loss": 4.4868,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737603374819751e-05,
+      "loss": 4.4857,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736764780068699e-05,
+      "loss": 4.4789,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359261853176476e-05,
+      "loss": 4.4781,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350892284469685e-05,
+      "loss": 4.4845,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342506336959165e-05,
+      "loss": 4.4875,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334120389448645e-05,
+      "loss": 4.4839,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325734441938125e-05,
+      "loss": 4.4755,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317348494427605e-05,
+      "loss": 4.4866,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7308978925720814e-05,
+      "loss": 4.4764,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300592978210294e-05,
+      "loss": 4.4821,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292207030699774e-05,
+      "loss": 4.4766,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283821083189254e-05,
+      "loss": 4.4747,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275435135678734e-05,
+      "loss": 4.4804,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267049188168214e-05,
+      "loss": 4.4637,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258663240657694e-05,
+      "loss": 4.4781,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250277293147174e-05,
+      "loss": 4.4627,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724190772444039e-05,
+      "loss": 4.4659,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72335381557336e-05,
+      "loss": 4.4669,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722515220822308e-05,
+      "loss": 4.4697,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721676626071256e-05,
+      "loss": 4.4634,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720838031320204e-05,
+      "loss": 4.4557,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719999436569152e-05,
+      "loss": 4.4744,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191608418181e-05,
+      "loss": 4.4625,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718322247067048e-05,
+      "loss": 4.4759,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717483652315996e-05,
+      "loss": 4.4619,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716646695445317e-05,
+      "loss": 4.4455,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715808100694265e-05,
+      "loss": 4.4461,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714969505943213e-05,
+      "loss": 4.4495,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141309111921614e-05,
+      "loss": 4.4664,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7132939543214823e-05,
+      "loss": 4.4457,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124553595704303e-05,
+      "loss": 4.459,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711616764819378e-05,
+      "loss": 4.4541,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710778170068326e-05,
+      "loss": 4.4534,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709941213197647e-05,
+      "loss": 4.4595,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709104256326968e-05,
+      "loss": 4.4342,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708265661575916e-05,
+      "loss": 4.4396,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707427066824864e-05,
+      "loss": 4.4334,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706588472073812e-05,
+      "loss": 4.4548,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70574987732276e-05,
+      "loss": 4.431,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704911282571708e-05,
+      "loss": 4.4573,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704072687820657e-05,
+      "loss": 4.4324,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703234093069605e-05,
+      "loss": 4.4326,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702397136198926e-05,
+      "loss": 4.4335,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701558541447874e-05,
+      "loss": 4.4301,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700719946696822e-05,
+      "loss": 4.4377,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69988135194577e-05,
+      "loss": 4.4316,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990443950750906e-05,
+      "loss": 4.4208,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982058003240386e-05,
+      "loss": 4.4367,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973672055729866e-05,
+      "loss": 4.4247,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696528610821934e-05,
+      "loss": 4.4277,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956916539512555e-05,
+      "loss": 4.4245,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948530592002035e-05,
+      "loss": 4.432,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940144644491515e-05,
+      "loss": 4.411,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931758696980995e-05,
+      "loss": 4.437,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923372749470475e-05,
+      "loss": 4.4218,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6914986801959955e-05,
+      "loss": 4.4249,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906600854449435e-05,
+      "loss": 4.4279,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898231285742644e-05,
+      "loss": 4.4174,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6889845338232124e-05,
+      "loss": 4.4357,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881459390721604e-05,
+      "loss": 4.438,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873073443211084e-05,
+      "loss": 4.4371,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864687495700564e-05,
+      "loss": 4.4216,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856301548190044e-05,
+      "loss": 4.4115,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684793197948325e-05,
+      "loss": 4.4046,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683954603197274e-05,
+      "loss": 4.4228,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683116008446222e-05,
+      "loss": 4.43,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68227741369517e-05,
+      "loss": 4.4261,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681438818944118e-05,
+      "loss": 4.4129,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680600224193066e-05,
+      "loss": 4.4032,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679761629442014e-05,
+      "loss": 4.4162,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678923034690962e-05,
+      "loss": 4.4031,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678086077820283e-05,
+      "loss": 4.4083,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677247483069231e-05,
+      "loss": 4.4168,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676408888318179e-05,
+      "loss": 4.4165,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675570293567127e-05,
+      "loss": 4.4094,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674731698816075e-05,
+      "loss": 4.4124,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673893104065023e-05,
+      "loss": 4.3982,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673054509313971e-05,
+      "loss": 4.4121,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672215914562919e-05,
+      "loss": 4.4082,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671380595572613e-05,
+      "loss": 4.3928,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670542000821561e-05,
+      "loss": 4.3989,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669703406070509e-05,
+      "loss": 4.4133,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668864811319457e-05,
+      "loss": 4.4144,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668026216568405e-05,
+      "loss": 4.4001,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667187621817353e-05,
+      "loss": 4.4019,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663490270663006e-05,
+      "loss": 4.4075,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655104323152486e-05,
+      "loss": 4.3892,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66467347544457e-05,
+      "loss": 4.4063,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638348806935175e-05,
+      "loss": 4.4113,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662996285942466e-05,
+      "loss": 4.3958,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662157691191414e-05,
+      "loss": 4.4024,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661320734320736e-05,
+      "loss": 4.3777,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660482139569683e-05,
+      "loss": 4.3787,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659643544818631e-05,
+      "loss": 4.3912,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658804950067579e-05,
+      "loss": 4.3924,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579679931969006e-05,
+      "loss": 4.3987,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657129398445848e-05,
+      "loss": 4.3834,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562924415751695e-05,
+      "loss": 4.3815,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554538468241175e-05,
+      "loss": 4.3877,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654615252073065e-05,
+      "loss": 4.3984,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653776657322013e-05,
+      "loss": 4.3905,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529380625709615e-05,
+      "loss": 4.3934,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6520994678199095e-05,
+      "loss": 4.3828,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512608730688575e-05,
+      "loss": 4.384,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504222783178055e-05,
+      "loss": 4.3954,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495853214471264e-05,
+      "loss": 4.3812,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487467266960744e-05,
+      "loss": 4.382,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479081319450224e-05,
+      "loss": 4.3794,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647071175074343e-05,
+      "loss": 4.3716,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646232580323291e-05,
+      "loss": 4.3803,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645393985572239e-05,
+      "loss": 4.38,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644555390821187e-05,
+      "loss": 4.3893,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643718433950508e-05,
+      "loss": 4.37,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642879839199457e-05,
+      "loss": 4.3743,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642041244448405e-05,
+      "loss": 4.3919,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641202649697353e-05,
+      "loss": 4.374,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640365692826674e-05,
+      "loss": 4.3887,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639527098075622e-05,
+      "loss": 4.3764,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63868850332457e-05,
+      "loss": 4.3921,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637849908573518e-05,
+      "loss": 4.3781,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637011313822466e-05,
+      "loss": 4.3683,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636174356951787e-05,
+      "loss": 4.3674,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635335762200735e-05,
+      "loss": 4.3647,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634497167449683e-05,
+      "loss": 4.376,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633658572698631e-05,
+      "loss": 4.3844,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632819977947579e-05,
+      "loss": 4.3763,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631981383196527e-05,
+      "loss": 4.3735,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631144426325848e-05,
+      "loss": 4.3668,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630305831574796e-05,
+      "loss": 4.3644,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629467236823744e-05,
+      "loss": 4.3598,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628628642072692e-05,
+      "loss": 4.3698,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62779004732164e-05,
+      "loss": 4.3816,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626951452570588e-05,
+      "loss": 4.3672,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626114495699909e-05,
+      "loss": 4.3635,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625275900948857e-05,
+      "loss": 4.3608,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.33704948425293,
+      "eval_runtime": 541.6671,
+      "eval_samples_per_second": 704.475,
+      "eval_steps_per_second": 22.015,
+      "step": 228957
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624437306197805e-05,
+      "loss": 4.3708,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623598711446753e-05,
+      "loss": 4.3725,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622760116695701e-05,
+      "loss": 4.3573,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219215219446484e-05,
+      "loss": 4.3573,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621082927193597e-05,
+      "loss": 4.3587,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620245970322919e-05,
+      "loss": 4.3399,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619407375571867e-05,
+      "loss": 4.3501,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618568780820814e-05,
+      "loss": 4.3619,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617730186069762e-05,
+      "loss": 4.3494,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168932291990836e-05,
+      "loss": 4.3574,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616054634448031e-05,
+      "loss": 4.3637,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615216039696979e-05,
+      "loss": 4.3618,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143790828263005e-05,
+      "loss": 4.3534,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135404880752485e-05,
+      "loss": 4.3565,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612701893324196e-05,
+      "loss": 4.3497,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611863298573144e-05,
+      "loss": 4.3485,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110247038220925e-05,
+      "loss": 4.3438,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6101861090710405e-05,
+      "loss": 4.3517,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093475143199885e-05,
+      "loss": 4.3533,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085089195689365e-05,
+      "loss": 4.3535,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076703248178845e-05,
+      "loss": 4.3463,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068317300668325e-05,
+      "loss": 4.3576,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059931353157805e-05,
+      "loss": 4.3474,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051561784451014e-05,
+      "loss": 4.3479,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604319221574422e-05,
+      "loss": 4.3496,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60348062682337e-05,
+      "loss": 4.3446,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602642032072318e-05,
+      "loss": 4.3515,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601803437321266e-05,
+      "loss": 4.3369,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600964842570214e-05,
+      "loss": 4.3521,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600126247819162e-05,
+      "loss": 4.336,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599289290948484e-05,
+      "loss": 4.337,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3422,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3479,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.3352,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.3385,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3432,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594259360322545e-05,
+      "loss": 4.3386,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593420765571493e-05,
+      "loss": 4.3554,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592582170820441e-05,
+      "loss": 4.3385,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917452139497616e-05,
+      "loss": 4.3244,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909066191987096e-05,
+      "loss": 4.3271,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900680244476576e-05,
+      "loss": 4.3235,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589229429696606e-05,
+      "loss": 4.3433,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588392472825927e-05,
+      "loss": 4.3276,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587553878074875e-05,
+      "loss": 4.3371,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586715283323823e-05,
+      "loss": 4.3328,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585876688572771e-05,
+      "loss": 4.3343,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585038093821719e-05,
+      "loss": 4.3395,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584199499070667e-05,
+      "loss": 4.3162,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583362542199988e-05,
+      "loss": 4.317,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582523947448936e-05,
+      "loss": 4.323,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581686990578257e-05,
+      "loss": 4.3355,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580848395827205e-05,
+      "loss": 4.3152,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580009801076153e-05,
+      "loss": 4.3373,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791712063251017e-05,
+      "loss": 4.3165,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783326115740497e-05,
+      "loss": 4.3129,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577494016822997e-05,
+      "loss": 4.3199,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576655422071945e-05,
+      "loss": 4.3089,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575816827320893e-05,
+      "loss": 4.3271,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574978232569841e-05,
+      "loss": 4.3188,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574139637818789e-05,
+      "loss": 4.3067,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573301043067737e-05,
+      "loss": 4.3216,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572462448316685e-05,
+      "loss": 4.3102,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571623853565633e-05,
+      "loss": 4.3164,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570786896694954e-05,
+      "loss": 4.3086,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569948301943902e-05,
+      "loss": 4.3209,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56910970719285e-05,
+      "loss": 4.3007,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568271112441798e-05,
+      "loss": 4.3223,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674341555711194e-05,
+      "loss": 4.3083,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56659719870044e-05,
+      "loss": 4.3101,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565758603949388e-05,
+      "loss": 4.3174,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564920009198336e-05,
+      "loss": 4.3099,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564081414447284e-05,
+      "loss": 4.3186,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563242819696232e-05,
+      "loss": 4.3304,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562405862825553e-05,
+      "loss": 4.325,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561567268074501e-05,
+      "loss": 4.3113,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560728673323449e-05,
+      "loss": 4.3054,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559890078572397e-05,
+      "loss": 4.2969,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559051483821345e-05,
+      "loss": 4.3133,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3215,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557375932199615e-05,
+      "loss": 4.3182,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556537337448563e-05,
+      "loss": 4.3029,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555698742697511e-05,
+      "loss": 4.2951,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.3139,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55402319107578e-05,
+      "loss": 4.2908,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553184596324728e-05,
+      "loss": 4.3045,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552346001573676e-05,
+      "loss": 4.3069,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551507406822624e-05,
+      "loss": 4.3111,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506704499519446e-05,
+      "loss": 4.3042,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498318552008926e-05,
+      "loss": 4.3097,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489932604498406e-05,
+      "loss": 4.2914,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481546656987886e-05,
+      "loss": 4.3032,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547316070947737e-05,
+      "loss": 4.302,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546477476196685e-05,
+      "loss": 4.2921,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.294,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3107,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.3074,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.3038,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.2965,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.306,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540610588700067e-05,
+      "loss": 4.2881,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539771993949015e-05,
+      "loss": 4.303,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538933399197963e-05,
+      "loss": 4.3092,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538094804446911e-05,
+      "loss": 4.2914,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537256209695859e-05,
+      "loss": 4.3076,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536417614944807e-05,
+      "loss": 4.2723,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535580658074128e-05,
+      "loss": 4.28,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534742063323076e-05,
+      "loss": 4.293,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533903468572024e-05,
+      "loss": 4.2929,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533064873820972e-05,
+      "loss": 4.3006,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53222627906992e-05,
+      "loss": 4.2832,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531387684318868e-05,
+      "loss": 4.2822,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530549089567816e-05,
+      "loss": 4.2893,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.299,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528873537946085e-05,
+      "loss": 4.2952,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528034943195033e-05,
+      "loss": 4.2951,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527196348443981e-05,
+      "loss": 4.2831,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263577536929295e-05,
+      "loss": 4.2857,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255207968222504e-05,
+      "loss": 4.3017,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246822020711984e-05,
+      "loss": 4.2842,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238436073201464e-05,
+      "loss": 4.2826,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230050125690944e-05,
+      "loss": 4.2856,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2725,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.2871,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520490866196311e-05,
+      "loss": 4.284,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519652271445259e-05,
+      "loss": 4.2948,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.2747,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.2773,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517138125072476e-05,
+      "loss": 4.2998,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516299530321425e-05,
+      "loss": 4.2789,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515460935570373e-05,
+      "loss": 4.2894,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514622340819321e-05,
+      "loss": 4.2862,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513783746068269e-05,
+      "loss": 4.2955,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512945151317217e-05,
+      "loss": 4.2871,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512108194446538e-05,
+      "loss": 4.2773,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511269599695486e-05,
+      "loss": 4.2719,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510431004944434e-05,
+      "loss": 4.2721,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509592410193382e-05,
+      "loss": 4.2824,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087554533227026e-05,
+      "loss": 4.2928,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079168585716506e-05,
+      "loss": 4.2843,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070782638205986e-05,
+      "loss": 4.2851,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062413069499195e-05,
+      "loss": 4.2736,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.2742,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.2667,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503725522696764e-05,
+      "loss": 4.282,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028869279457115e-05,
+      "loss": 4.2848,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020483331946595e-05,
+      "loss": 4.2852,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012097384436075e-05,
+      "loss": 4.2705,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003711436925555e-05,
+      "loss": 4.2693,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.255746364593506,
+      "eval_runtime": 538.6451,
+      "eval_samples_per_second": 708.427,
+      "eval_steps_per_second": 22.139,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995325489415035e-05,
+      "loss": 4.283,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.279,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2683,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2678,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2693,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2563,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.2615,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2674,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2666,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.2718,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491148239311357e-05,
+      "loss": 4.2751,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490309644560305e-05,
+      "loss": 4.2703,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489471049809253e-05,
+      "loss": 4.268,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886324550582003e-05,
+      "loss": 4.2718,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877938603071483e-05,
+      "loss": 4.2652,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486955265556096e-05,
+      "loss": 4.2591,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486116670805044e-05,
+      "loss": 4.2572,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485278076053992e-05,
+      "loss": 4.2695,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48443948130294e-05,
+      "loss": 4.2645,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483600886551888e-05,
+      "loss": 4.2718,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482762291800836e-05,
+      "loss": 4.2607,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481923697049785e-05,
+      "loss": 4.2708,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481085102298733e-05,
+      "loss": 4.2647,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480248145428054e-05,
+      "loss": 4.2652,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479409550677002e-05,
+      "loss": 4.2622,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2599,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.267,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.2516,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476055171672794e-05,
+      "loss": 4.2673,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2533,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2498,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.2611,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.263,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471863835797907e-05,
+      "loss": 4.2565,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471025241046855e-05,
+      "loss": 4.2537,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701882841761763e-05,
+      "loss": 4.2588,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693496894251243e-05,
+      "loss": 4.2619,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685110946740717e-05,
+      "loss": 4.2737,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676724999230197e-05,
+      "loss": 4.2544,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668339051719676e-05,
+      "loss": 4.2459,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465996948301289e-05,
+      "loss": 4.244,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651583535502366e-05,
+      "loss": 4.2456,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643197587991846e-05,
+      "loss": 4.2607,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634811640481325e-05,
+      "loss": 4.2479,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626425692970805e-05,
+      "loss": 4.2593,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618039745460285e-05,
+      "loss": 4.2521,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609653797949765e-05,
+      "loss": 4.2501,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460126785043925e-05,
+      "loss": 4.2614,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459289828173246e-05,
+      "loss": 4.2392,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458451233422194e-05,
+      "loss": 4.2341,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457614276551515e-05,
+      "loss": 4.2462,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456775681800463e-05,
+      "loss": 4.2525,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455937087049411e-05,
+      "loss": 4.2341,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455098492298359e-05,
+      "loss": 4.2575,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45426153542768e-05,
+      "loss": 4.238,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2333,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452584345925576e-05,
+      "loss": 4.2412,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451745751174524e-05,
+      "loss": 4.2308,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509087943038455e-05,
+      "loss": 4.2519,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500701995527935e-05,
+      "loss": 4.2431,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492316048017415e-05,
+      "loss": 4.2274,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483930100506895e-05,
+      "loss": 4.2403,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475544152996375e-05,
+      "loss": 4.234,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467158205485855e-05,
+      "loss": 4.2453,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458772257975335e-05,
+      "loss": 4.2298,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450402689268544e-05,
+      "loss": 4.2425,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442016741758024e-05,
+      "loss": 4.2228,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433630794247504e-05,
+      "loss": 4.2429,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425244846736984e-05,
+      "loss": 4.2331,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416858899226464e-05,
+      "loss": 4.2316,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408472951715944e-05,
+      "loss": 4.2441,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400087004205423e-05,
+      "loss": 4.2323,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391701056694903e-05,
+      "loss": 4.2429,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438333148798812e-05,
+      "loss": 4.2509,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43749455404776e-05,
+      "loss": 4.2514,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436655959296708e-05,
+      "loss": 4.2359,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435817364545655e-05,
+      "loss": 4.2322,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434978769794603e-05,
+      "loss": 4.2222,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434143450804298e-05,
+      "loss": 4.2376,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433304856053246e-05,
+      "loss": 4.2518,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432466261302194e-05,
+      "loss": 4.2415,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431627666551142e-05,
+      "loss": 4.2316,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43078907180009e-05,
+      "loss": 4.2181,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429950477049038e-05,
+      "loss": 4.2443,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429111882297986e-05,
+      "loss": 4.219,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428274925427307e-05,
+      "loss": 4.2282,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427436330676255e-05,
+      "loss": 4.2303,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4265977359252026e-05,
+      "loss": 4.2356,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257591411741506e-05,
+      "loss": 4.2354,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424922184303472e-05,
+      "loss": 4.2372,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42408358955242e-05,
+      "loss": 4.2187,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232449948013675e-05,
+      "loss": 4.2311,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224064000503155e-05,
+      "loss": 4.2296,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215678052992635e-05,
+      "loss": 4.2212,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420730848428585e-05,
+      "loss": 4.22,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419892253677533e-05,
+      "loss": 4.2401,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419053658926481e-05,
+      "loss": 4.2338,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418215064175429e-05,
+      "loss": 4.2326,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417376469424377e-05,
+      "loss": 4.2232,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416539512553698e-05,
+      "loss": 4.2312,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415700917802646e-05,
+      "loss": 4.2188,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414862323051594e-05,
+      "loss": 4.2333,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414023728300542e-05,
+      "loss": 4.2348,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413186771429863e-05,
+      "loss": 4.2217,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412348176678811e-05,
+      "loss": 4.2369,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411509581927759e-05,
+      "loss": 4.2046,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410670987176707e-05,
+      "loss": 4.2112,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409832392425655e-05,
+      "loss": 4.2196,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408993797674603e-05,
+      "loss": 4.221,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081568408039244e-05,
+      "loss": 4.233,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073182460528724e-05,
+      "loss": 4.2182,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064796513018204e-05,
+      "loss": 4.2097,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056410565507684e-05,
+      "loss": 4.2171,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048024617997164e-05,
+      "loss": 4.2303,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039638670486644e-05,
+      "loss": 4.2273,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031252722976124e-05,
+      "loss": 4.2286,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402288315426933e-05,
+      "loss": 4.2148,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401449720675881e-05,
+      "loss": 4.2162,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400611125924829e-05,
+      "loss": 4.2313,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399772531173777e-05,
+      "loss": 4.2188,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398935574303098e-05,
+      "loss": 4.2144,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398096979552047e-05,
+      "loss": 4.2179,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397258384800995e-05,
+      "loss": 4.2059,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396419790049943e-05,
+      "loss": 4.2195,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395581195298891e-05,
+      "loss": 4.2192,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394744238428212e-05,
+      "loss": 4.2258,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39390564367716e-05,
+      "loss": 4.2087,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393067048926108e-05,
+      "loss": 4.2113,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392228454175056e-05,
+      "loss": 4.2281,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391391497304377e-05,
+      "loss": 4.2163,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390552902553325e-05,
+      "loss": 4.2238,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389714307802273e-05,
+      "loss": 4.2222,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388875713051221e-05,
+      "loss": 4.2252,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388037118300169e-05,
+      "loss": 4.2241,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387198523549117e-05,
+      "loss": 4.2081,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863599287980647e-05,
+      "loss": 4.2067,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855213340470127e-05,
+      "loss": 4.209,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846827392959606e-05,
+      "loss": 4.2157,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838441445449086e-05,
+      "loss": 4.2226,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830071876742296e-05,
+      "loss": 4.2206,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821702308035505e-05,
+      "loss": 4.2198,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813316360524985e-05,
+      "loss": 4.2111,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804930413014465e-05,
+      "loss": 4.2111,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796544465503944e-05,
+      "loss": 4.2003,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788158517993424e-05,
+      "loss": 4.2183,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779772570482904e-05,
+      "loss": 4.2208,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377138662297239e-05,
+      "loss": 4.2219,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376300067546187e-05,
+      "loss": 4.2039,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375463110675508e-05,
+      "loss": 4.2052,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.200170993804932,
+      "eval_runtime": 541.6701,
+      "eval_samples_per_second": 704.471,
+      "eval_steps_per_second": 22.015,
+      "step": 381595
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374626153804829e-05,
+      "loss": 4.2216,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373787559053777e-05,
+      "loss": 4.213,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372948964302725e-05,
+      "loss": 4.2051,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372110369551673e-05,
+      "loss": 4.2004,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371271774800621e-05,
+      "loss": 4.2047,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370433180049569e-05,
+      "loss": 4.1946,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369594585298517e-05,
+      "loss": 4.1998,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368755990547465e-05,
+      "loss": 4.1979,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367919033676786e-05,
+      "loss": 4.204,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670820768061074e-05,
+      "loss": 4.2137,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662434820550554e-05,
+      "loss": 4.2068,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654048873040034e-05,
+      "loss": 4.2054,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645662925529514e-05,
+      "loss": 4.2093,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637276978018994e-05,
+      "loss": 4.2074,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36289074093122e-05,
+      "loss": 4.2009,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362052146180168e-05,
+      "loss": 4.1961,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361215189309489e-05,
+      "loss": 4.1956,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360376594558437e-05,
+      "loss": 4.2041,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359537999807385e-05,
+      "loss": 4.2044,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358699405056333e-05,
+      "loss": 4.2128,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357860810305281e-05,
+      "loss": 4.1991,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35702221555423e-05,
+      "loss": 4.2097,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356183620803178e-05,
+      "loss": 4.2066,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355345026052126e-05,
+      "loss": 4.2026,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354506431301074e-05,
+      "loss": 4.2003,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353667836550022e-05,
+      "loss": 4.197,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352830879679343e-05,
+      "loss": 4.2077,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351992284928291e-05,
+      "loss": 4.1882,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351153690177239e-05,
+      "loss": 4.2097,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350315095426187e-05,
+      "loss": 4.192,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349476500675134e-05,
+      "loss": 4.187,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348637905924082e-05,
+      "loss": 4.2042,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34779931117303e-05,
+      "loss": 4.2004,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469623543023516e-05,
+      "loss": 4.2019,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461237595512996e-05,
+      "loss": 4.1937,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3452851648002476e-05,
+      "loss": 4.1936,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444465700491956e-05,
+      "loss": 4.1992,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436079752981436e-05,
+      "loss": 4.2142,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427693805470916e-05,
+      "loss": 4.1959,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419307857960396e-05,
+      "loss": 4.1863,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3410921910449876e-05,
+      "loss": 4.1866,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402552341743085e-05,
+      "loss": 4.1835,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394166394232565e-05,
+      "loss": 4.198,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385780446722045e-05,
+      "loss": 4.1932,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377410878015254e-05,
+      "loss": 4.1956,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3369024930504734e-05,
+      "loss": 4.1968,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360638982994214e-05,
+      "loss": 4.1906,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33522530354837e-05,
+      "loss": 4.202,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334386708797318e-05,
+      "loss": 4.1815,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333548114046266e-05,
+      "loss": 4.1742,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332709519295214e-05,
+      "loss": 4.1916,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331870924544162e-05,
+      "loss": 4.1871,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.1797,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330195372922431e-05,
+      "loss": 4.1973,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329358416051752e-05,
+      "loss": 4.1787,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285198213007e-05,
+      "loss": 4.1784,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327681226549648e-05,
+      "loss": 4.1798,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326842631798596e-05,
+      "loss": 4.1744,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326004037047544e-05,
+      "loss": 4.1932,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325165442296492e-05,
+      "loss": 4.1829,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32432684754544e-05,
+      "loss": 4.1794,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3234882527943885e-05,
+      "loss": 4.1782,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3226512959237094e-05,
+      "loss": 4.1714,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218127011726574e-05,
+      "loss": 4.1922,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209741064216054e-05,
+      "loss": 4.1742,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320137149550926e-05,
+      "loss": 4.1907,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319298554799874e-05,
+      "loss": 4.1648,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.1837,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31762136529777e-05,
+      "loss": 4.1783,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.1731,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159441757956656e-05,
+      "loss": 4.1865,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151055810446136e-05,
+      "loss": 4.1789,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314266986293562e-05,
+      "loss": 4.184,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313430029422883e-05,
+      "loss": 4.1931,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312593072552205e-05,
+      "loss": 4.1943,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311754477801153e-05,
+      "loss": 4.1846,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109158830501e-05,
+      "loss": 4.177,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310077288299048e-05,
+      "loss": 4.1689,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30924033142837e-05,
+      "loss": 4.18,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308401736677318e-05,
+      "loss": 4.1922,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307563141926265e-05,
+      "loss": 4.1859,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306724547175213e-05,
+      "loss": 4.1785,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058875903045346e-05,
+      "loss": 4.161,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050489955534826e-05,
+      "loss": 4.1916,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042104008024306e-05,
+      "loss": 4.1595,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033718060513786e-05,
+      "loss": 4.1765,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025332113003266e-05,
+      "loss": 4.1761,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3016946165492746e-05,
+      "loss": 4.181,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.1773,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000174270471705e-05,
+      "loss": 4.1871,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991788322961185e-05,
+      "loss": 4.1618,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983402375450665e-05,
+      "loss": 4.1742,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975016427940145e-05,
+      "loss": 4.1784,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966630480429625e-05,
+      "loss": 4.1686,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.166,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.1849,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294150539550553e-05,
+      "loss": 4.1785,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293311944799501e-05,
+      "loss": 4.1857,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292473350048449e-05,
+      "loss": 4.1702,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291634755297397e-05,
+      "loss": 4.1773,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290796160546345e-05,
+      "loss": 4.1637,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289957565795293e-05,
+      "loss": 4.179,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289118971044241e-05,
+      "loss": 4.1845,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288280376293189e-05,
+      "loss": 4.167,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287441781542136e-05,
+      "loss": 4.1833,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286604824671458e-05,
+      "loss": 4.1502,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285766229920406e-05,
+      "loss": 4.1622,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284929273049727e-05,
+      "loss": 4.1626,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284090678298675e-05,
+      "loss": 4.1689,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283252083547623e-05,
+      "loss": 4.1801,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.1709,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2815765319258924e-05,
+      "loss": 4.1481,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807379371748404e-05,
+      "loss": 4.1699,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2798993424237884e-05,
+      "loss": 4.1746,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790607476727364e-05,
+      "loss": 4.1763,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278222152921684e-05,
+      "loss": 4.1738,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277383558170632e-05,
+      "loss": 4.1639,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27654496341958e-05,
+      "loss": 4.163,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275708006548901e-05,
+      "loss": 4.1829,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274871049678222e-05,
+      "loss": 4.1647,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27403245492717e-05,
+      "loss": 4.1648,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.1642,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1559,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271516670674014e-05,
+      "loss": 4.1661,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270678075922962e-05,
+      "loss": 4.1671,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26983948117191e-05,
+      "loss": 4.1742,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269000886420858e-05,
+      "loss": 4.1603,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268162291669806e-05,
+      "loss": 4.159,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267323696918754e-05,
+      "loss": 4.1761,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266485102167702e-05,
+      "loss": 4.1622,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26564650741665e-05,
+      "loss": 4.1696,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264807912665598e-05,
+      "loss": 4.1723,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263970955794919e-05,
+      "loss": 4.1783,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263132361043867e-05,
+      "loss": 4.1719,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262293766292815e-05,
+      "loss": 4.1595,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261455171541763e-05,
+      "loss": 4.1511,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606182146710846e-05,
+      "loss": 4.1603,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597796199200326e-05,
+      "loss": 4.1668,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589410251689806e-05,
+      "loss": 4.1748,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581024304179286e-05,
+      "loss": 4.1674,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2572638356668766e-05,
+      "loss": 4.1647,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564268787961975e-05,
+      "loss": 4.166,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2555882840451455e-05,
+      "loss": 4.1629,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547496892940935e-05,
+      "loss": 4.1522,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539110945430415e-05,
+      "loss": 4.1609,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530741376723624e-05,
+      "loss": 4.1731,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522355429213104e-05,
+      "loss": 4.1733,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2513969481702584e-05,
+      "loss": 4.1536,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250558353419207e-05,
+      "loss": 4.1539,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.158684730529785,
+      "eval_runtime": 538.9879,
+      "eval_samples_per_second": 707.977,
+      "eval_steps_per_second": 22.125,
+      "step": 457914
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 1.889853492561904e+17,
+  "trial_name": null,
+  "trial_params": null
+}