Upload folder using huggingface_hub

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9efd4d938e4d78fd90e91491806ea77e5f0cf910a0dadbf284a88f929ad73149
+oid sha256:671c007da97093393af79ca4122b30dbe0eacca718f9789ff8971d6c2e56056c
 size 1603041146

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e167f8b60bb5ab7afb5d2afe128567f76db9994b9819a4020aa8d3ae91212b30
+oid sha256:201951c3307c5287d99dba3b404524cc00f37c2c3e376adb491bf561525119f2
 size 1253695538

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e415292e0180e6651f0918c238164a9495c533d8566897a264388c7fbf4d4c32
+oid sha256:51a8927aad5b8fa5acfc364555592057ea9824d9a07853b39462940bdb3c167b
 size 14244

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d2395c83e75c2a21c7feacb0b7114db998eef16354a0a708ea05d274cbc67a3
+oid sha256:955040d0ab4a98e858e89bc7b9f85ba88a35b45c1ae13d3d89b36942f10d8f5f
 size 1064

trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.010031634190783,
+  "epoch": 1.0019513811172567,
   "eval_steps": 500,
-  "global_step": 250000,
+  "global_step": 248000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -173615,1406 +173615,6 @@
       "learning_rate": 7.797699192912955e-09,
       "loss": 98.4398,
       "step": 248000
-    },
-    {
-      "epoch": 1.0019917823826243,
-      "grad_norm": 375.5564270019531,
-      "learning_rate": 7.719922147779723e-09,
-      "loss": 62.2112,
-      "step": 248010
-    },
-    {
-      "epoch": 1.002032183647992,
-      "grad_norm": 783.7318725585938,
-      "learning_rate": 7.642534862439821e-09,
-      "loss": 60.1914,
-      "step": 248020
-    },
-    {
-      "epoch": 1.0020725849133594,
-      "grad_norm": 666.0960693359375,
-      "learning_rate": 7.56553733840093e-09,
-      "loss": 72.5856,
-      "step": 248030
-    },
-    {
-      "epoch": 1.002112986178727,
-      "grad_norm": 825.030517578125,
-      "learning_rate": 7.488929577164072e-09,
-      "loss": 60.5071,
-      "step": 248040
-    },
-    {
-      "epoch": 1.0021533874440947,
-      "grad_norm": 892.0487060546875,
-      "learning_rate": 7.412711580225829e-09,
-      "loss": 78.1599,
-      "step": 248050
-    },
-    {
-      "epoch": 1.0021937887094623,
-      "grad_norm": 610.88330078125,
-      "learning_rate": 7.336883349067236e-09,
-      "loss": 81.9374,
-      "step": 248060
-    },
-    {
-      "epoch": 1.00223418997483,
-      "grad_norm": 289.4344787597656,
-      "learning_rate": 7.261444885169333e-09,
-      "loss": 56.0395,
-      "step": 248070
-    },
-    {
-      "epoch": 1.0022745912401976,
-      "grad_norm": 751.8075561523438,
-      "learning_rate": 7.1863961900042746e-09,
-      "loss": 76.6276,
-      "step": 248080
-    },
-    {
-      "epoch": 1.0023149925055652,
-      "grad_norm": 472.09954833984375,
-      "learning_rate": 7.111737265030893e-09,
-      "loss": 47.1432,
-      "step": 248090
-    },
-    {
-      "epoch": 1.0023553937709329,
-      "grad_norm": 536.2998657226562,
-      "learning_rate": 7.037468111710244e-09,
-      "loss": 63.2109,
-      "step": 248100
-    },
-    {
-      "epoch": 1.0023957950363005,
-      "grad_norm": 521.3856201171875,
-      "learning_rate": 6.963588731485616e-09,
-      "loss": 47.3244,
-      "step": 248110
-    },
-    {
-      "epoch": 1.0024361963016681,
-      "grad_norm": 557.798583984375,
-      "learning_rate": 6.890099125798078e-09,
-      "loss": 50.9857,
-      "step": 248120
-    },
-    {
-      "epoch": 1.0024765975670358,
-      "grad_norm": 437.3095703125,
-      "learning_rate": 6.816999296082039e-09,
-      "loss": 64.5692,
-      "step": 248130
-    },
-    {
-      "epoch": 1.0025169988324034,
-      "grad_norm": 338.9017639160156,
-      "learning_rate": 6.744289243760804e-09,
-      "loss": 31.7949,
-      "step": 248140
-    },
-    {
-      "epoch": 1.002557400097771,
-      "grad_norm": 621.0401611328125,
-      "learning_rate": 6.6719689702554605e-09,
-      "loss": 75.7591,
-      "step": 248150
-    },
-    {
-      "epoch": 1.0025978013631387,
-      "grad_norm": 769.9017944335938,
-      "learning_rate": 6.600038476971548e-09,
-      "loss": 56.7265,
-      "step": 248160
-    },
-    {
-      "epoch": 1.0026382026285063,
-      "grad_norm": 756.1639404296875,
-      "learning_rate": 6.528497765312392e-09,
-      "loss": 47.5924,
-      "step": 248170
-    },
-    {
-      "epoch": 1.002678603893874,
-      "grad_norm": 583.2656860351562,
-      "learning_rate": 6.45734683667465e-09,
-      "loss": 77.5252,
-      "step": 248180
-    },
-    {
-      "epoch": 1.0027190051592416,
-      "grad_norm": 800.96337890625,
-      "learning_rate": 6.38658569244388e-09,
-      "loss": 75.3352,
-      "step": 248190
-    },
-    {
-      "epoch": 1.0027594064246093,
-      "grad_norm": 373.50006103515625,
-      "learning_rate": 6.316214334001203e-09,
-      "loss": 42.8643,
-      "step": 248200
-    },
-    {
-      "epoch": 1.002799807689977,
-      "grad_norm": 362.3075256347656,
-      "learning_rate": 6.2462327627166304e-09,
-      "loss": 45.8071,
-      "step": 248210
-    },
-    {
-      "epoch": 1.0028402089553445,
-      "grad_norm": 957.6506958007812,
-      "learning_rate": 6.176640979953519e-09,
-      "loss": 70.7984,
-      "step": 248220
-    },
-    {
-      "epoch": 1.0028806102207122,
-      "grad_norm": 591.532470703125,
-      "learning_rate": 6.107438987073e-09,
-      "loss": 69.2733,
-      "step": 248230
-    },
-    {
-      "epoch": 1.0029210114860798,
-      "grad_norm": 2639.35009765625,
-      "learning_rate": 6.038626785418444e-09,
-      "loss": 87.1864,
-      "step": 248240
-    },
-    {
-      "epoch": 1.0029614127514475,
-      "grad_norm": 594.7188110351562,
-      "learning_rate": 5.970204376337663e-09,
-      "loss": 69.4996,
-      "step": 248250
-    },
-    {
-      "epoch": 1.003001814016815,
-      "grad_norm": 331.9031066894531,
-      "learning_rate": 5.902171761160702e-09,
-      "loss": 63.0195,
-      "step": 248260
-    },
-    {
-      "epoch": 1.0030422152821827,
-      "grad_norm": 732.08984375,
-      "learning_rate": 5.834528941213169e-09,
-      "loss": 62.3124,
-      "step": 248270
-    },
-    {
-      "epoch": 1.0030826165475504,
-      "grad_norm": 316.0459289550781,
-      "learning_rate": 5.767275917816229e-09,
-      "loss": 47.8008,
-      "step": 248280
-    },
-    {
-      "epoch": 1.0031230178129178,
-      "grad_norm": 180.98361206054688,
-      "learning_rate": 5.700412692279944e-09,
-      "loss": 82.8906,
-      "step": 248290
-    },
-    {
-      "epoch": 1.0031634190782854,
-      "grad_norm": 383.6042175292969,
-      "learning_rate": 5.633939265905497e-09,
-      "loss": 42.444,
-      "step": 248300
-    },
-    {
-      "epoch": 1.003203820343653,
-      "grad_norm": 698.3505859375,
-      "learning_rate": 5.5678556399940686e-09,
-      "loss": 71.1799,
-      "step": 248310
-    },
-    {
-      "epoch": 1.0032442216090207,
-      "grad_norm": 737.6282348632812,
-      "learning_rate": 5.502161815829077e-09,
-      "loss": 92.5681,
-      "step": 248320
-    },
-    {
-      "epoch": 1.0032846228743884,
-      "grad_norm": 695.4033813476562,
-      "learning_rate": 5.4368577946961596e-09,
-      "loss": 52.5139,
-      "step": 248330
-    },
-    {
-      "epoch": 1.003325024139756,
-      "grad_norm": 460.7320556640625,
-      "learning_rate": 5.371943577863192e-09,
-      "loss": 43.3595,
-      "step": 248340
-    },
-    {
-      "epoch": 1.0033654254051236,
-      "grad_norm": 563.068115234375,
-      "learning_rate": 5.3074191665980494e-09,
-      "loss": 61.4819,
-      "step": 248350
-    },
-    {
-      "epoch": 1.0034058266704913,
-      "grad_norm": 342.0809631347656,
-      "learning_rate": 5.243284562159723e-09,
-      "loss": 43.0043,
-      "step": 248360
-    },
-    {
-      "epoch": 1.003446227935859,
-      "grad_norm": 455.55938720703125,
-      "learning_rate": 5.179539765798325e-09,
-      "loss": 58.0983,
-      "step": 248370
-    },
-    {
-      "epoch": 1.0034866292012266,
-      "grad_norm": 1556.4609375,
-      "learning_rate": 5.116184778755085e-09,
-      "loss": 78.0046,
-      "step": 248380
-    },
-    {
-      "epoch": 1.0035270304665942,
-      "grad_norm": 631.3849487304688,
-      "learning_rate": 5.053219602264569e-09,
-      "loss": 53.1356,
-      "step": 248390
-    },
-    {
-      "epoch": 1.0035674317319618,
-      "grad_norm": 688.4679565429688,
-      "learning_rate": 4.990644237556908e-09,
-      "loss": 66.3212,
-      "step": 248400
-    },
-    {
-      "epoch": 1.0036078329973295,
-      "grad_norm": 384.0335388183594,
-      "learning_rate": 4.928458685851123e-09,
-      "loss": 52.2714,
-      "step": 248410
-    },
-    {
-      "epoch": 1.0036482342626971,
-      "grad_norm": 1073.441650390625,
-      "learning_rate": 4.866662948359579e-09,
-      "loss": 55.6296,
-      "step": 248420
-    },
-    {
-      "epoch": 1.0036886355280648,
-      "grad_norm": 949.4934692382812,
-      "learning_rate": 4.805257026283538e-09,
-      "loss": 71.919,
-      "step": 248430
-    },
-    {
-      "epoch": 1.0037290367934324,
-      "grad_norm": 911.975830078125,
-      "learning_rate": 4.744240920826481e-09,
-      "loss": 71.4097,
-      "step": 248440
-    },
-    {
-      "epoch": 1.0037694380588,
-      "grad_norm": 1051.722900390625,
-      "learning_rate": 4.683614633174128e-09,
-      "loss": 46.7513,
-      "step": 248450
-    },
-    {
-      "epoch": 1.0038098393241677,
-      "grad_norm": 786.8934326171875,
-      "learning_rate": 4.623378164507752e-09,
-      "loss": 65.1381,
-      "step": 248460
-    },
-    {
-      "epoch": 1.0038502405895353,
-      "grad_norm": 1190.6405029296875,
-      "learning_rate": 4.563531516004194e-09,
-      "loss": 49.0076,
-      "step": 248470
-    },
-    {
-      "epoch": 1.003890641854903,
-      "grad_norm": 609.81103515625,
-      "learning_rate": 4.5040746888269646e-09,
-      "loss": 94.319,
-      "step": 248480
-    },
-    {
-      "epoch": 1.0039310431202706,
-      "grad_norm": 1157.087158203125,
-      "learning_rate": 4.44500768413958e-09,
-      "loss": 103.0867,
-      "step": 248490
-    },
-    {
-      "epoch": 1.0039714443856382,
-      "grad_norm": 843.581298828125,
-      "learning_rate": 4.386330503090008e-09,
-      "loss": 72.9247,
-      "step": 248500
-    },
-    {
-      "epoch": 1.0040118456510059,
-      "grad_norm": 507.43182373046875,
-      "learning_rate": 4.328043146824002e-09,
-      "loss": 54.9502,
-      "step": 248510
-    },
-    {
-      "epoch": 1.0040522469163735,
-      "grad_norm": 915.3667602539062,
-      "learning_rate": 4.2701456164762066e-09,
-      "loss": 79.8888,
-      "step": 248520
-    },
-    {
-      "epoch": 1.0040926481817412,
-      "grad_norm": 612.6775512695312,
-      "learning_rate": 4.2126379131768314e-09,
-      "loss": 51.1185,
-      "step": 248530
-    },
-    {
-      "epoch": 1.0041330494471086,
-      "grad_norm": 921.3275146484375,
-      "learning_rate": 4.155520038047201e-09,
-      "loss": 61.4335,
-      "step": 248540
-    },
-    {
-      "epoch": 1.0041734507124762,
-      "grad_norm": 454.0970458984375,
-      "learning_rate": 4.098791992199758e-09,
-      "loss": 66.9522,
-      "step": 248550
-    },
-    {
-      "epoch": 1.0042138519778439,
-      "grad_norm": 594.523681640625,
-      "learning_rate": 4.042453776740285e-09,
-      "loss": 40.608,
-      "step": 248560
-    },
-    {
-      "epoch": 1.0042542532432115,
-      "grad_norm": 807.0231323242188,
-      "learning_rate": 3.986505392770124e-09,
-      "loss": 93.9332,
-      "step": 248570
-    },
-    {
-      "epoch": 1.0042946545085791,
-      "grad_norm": 307.3790283203125,
-      "learning_rate": 3.9309468413750715e-09,
-      "loss": 74.7959,
-      "step": 248580
-    },
-    {
-      "epoch": 1.0043350557739468,
-      "grad_norm": 427.72808837890625,
-      "learning_rate": 3.875778123643148e-09,
-      "loss": 65.4532,
-      "step": 248590
-    },
-    {
-      "epoch": 1.0043754570393144,
-      "grad_norm": 1061.6275634765625,
-      "learning_rate": 3.820999240644608e-09,
-      "loss": 71.2552,
-      "step": 248600
-    },
-    {
-      "epoch": 1.004415858304682,
-      "grad_norm": 352.9907531738281,
-      "learning_rate": 3.766610193454146e-09,
-      "loss": 54.195,
-      "step": 248610
-    },
-    {
-      "epoch": 1.0044562595700497,
-      "grad_norm": 668.2301635742188,
-      "learning_rate": 3.7126109831264746e-09,
-      "loss": 60.256,
-      "step": 248620
-    },
-    {
-      "epoch": 1.0044966608354173,
-      "grad_norm": 228.10552978515625,
-      "learning_rate": 3.6590016107163064e-09,
-      "loss": 45.8084,
-      "step": 248630
-    },
-    {
-      "epoch": 1.004537062100785,
-      "grad_norm": 237.22738647460938,
-      "learning_rate": 3.60578207726725e-09,
-      "loss": 40.9312,
-      "step": 248640
-    },
-    {
-      "epoch": 1.0045774633661526,
-      "grad_norm": 550.7319946289062,
-      "learning_rate": 3.552952383820696e-09,
-      "loss": 66.7888,
-      "step": 248650
-    },
-    {
-      "epoch": 1.0046178646315203,
-      "grad_norm": 951.9227294921875,
-      "learning_rate": 3.500512531402489e-09,
-      "loss": 81.9458,
-      "step": 248660
-    },
-    {
-      "epoch": 1.004658265896888,
-      "grad_norm": 483.5680236816406,
-      "learning_rate": 3.4484625210362554e-09,
-      "loss": 55.9166,
-      "step": 248670
-    },
-    {
-      "epoch": 1.0046986671622555,
-      "grad_norm": 1149.939697265625,
-      "learning_rate": 3.3968023537389594e-09,
-      "loss": 68.5474,
-      "step": 248680
-    },
-    {
-      "epoch": 1.0047390684276232,
-      "grad_norm": 615.0205078125,
-      "learning_rate": 3.3455320305142423e-09,
-      "loss": 76.0233,
-      "step": 248690
-    },
-    {
-      "epoch": 1.0047794696929908,
-      "grad_norm": 355.2969055175781,
-      "learning_rate": 3.294651552365746e-09,
-      "loss": 64.7722,
-      "step": 248700
-    },
-    {
-      "epoch": 1.0048198709583585,
-      "grad_norm": 1062.1531982421875,
-      "learning_rate": 3.2441609202815694e-09,
-      "loss": 80.2818,
-      "step": 248710
-    },
-    {
-      "epoch": 1.004860272223726,
-      "grad_norm": 651.049560546875,
-      "learning_rate": 3.1940601352475895e-09,
-      "loss": 52.2253,
-      "step": 248720
-    },
-    {
-      "epoch": 1.0049006734890937,
-      "grad_norm": 588.716064453125,
-      "learning_rate": 3.1443491982408035e-09,
-      "loss": 68.1338,
-      "step": 248730
-    },
-    {
-      "epoch": 1.0049410747544614,
-      "grad_norm": 757.9308471679688,
-      "learning_rate": 3.0950281102293258e-09,
-      "loss": 59.9994,
-      "step": 248740
-    },
-    {
-      "epoch": 1.004981476019829,
-      "grad_norm": 682.371337890625,
-      "learning_rate": 3.0460968721746086e-09,
-      "loss": 41.9455,
-      "step": 248750
-    },
-    {
-      "epoch": 1.0050218772851967,
-      "grad_norm": 850.54833984375,
-      "learning_rate": 2.997555485033665e-09,
-      "loss": 48.9406,
-      "step": 248760
-    },
-    {
-      "epoch": 1.0050622785505643,
-      "grad_norm": 585.2266845703125,
-      "learning_rate": 2.9494039497479645e-09,
-      "loss": 70.8767,
-      "step": 248770
-    },
-    {
-      "epoch": 1.005102679815932,
-      "grad_norm": 923.8726806640625,
-      "learning_rate": 2.9016422672611954e-09,
-      "loss": 71.7036,
-      "step": 248780
-    },
-    {
-      "epoch": 1.0051430810812996,
-      "grad_norm": 811.24658203125,
-      "learning_rate": 2.854270438501505e-09,
-      "loss": 67.8668,
-      "step": 248790
-    },
-    {
-      "epoch": 1.005183482346667,
-      "grad_norm": 653.4196166992188,
-      "learning_rate": 2.807288464392599e-09,
-      "loss": 57.9755,
-      "step": 248800
-    },
-    {
-      "epoch": 1.0052238836120346,
-      "grad_norm": 1551.59375,
-      "learning_rate": 2.7606963458493006e-09,
-      "loss": 83.9572,
-      "step": 248810
-    },
-    {
-      "epoch": 1.0052642848774023,
-      "grad_norm": 435.47625732421875,
-      "learning_rate": 2.7144940837842138e-09,
-      "loss": 54.7893,
-      "step": 248820
-    },
-    {
-      "epoch": 1.00530468614277,
-      "grad_norm": 726.771484375,
-      "learning_rate": 2.668681679094398e-09,
-      "loss": 63.5434,
-      "step": 248830
-    },
-    {
-      "epoch": 1.0053450874081375,
-      "grad_norm": 648.3975219726562,
-      "learning_rate": 2.6232591326724733e-09,
-      "loss": 60.4749,
-      "step": 248840
-    },
-    {
-      "epoch": 1.0053854886735052,
-      "grad_norm": 644.8402709960938,
-      "learning_rate": 2.5782264454066175e-09,
-      "loss": 73.8123,
-      "step": 248850
-    },
-    {
-      "epoch": 1.0054258899388728,
-      "grad_norm": 513.4389038085938,
-      "learning_rate": 2.5335836181739072e-09,
-      "loss": 50.9707,
-      "step": 248860
-    },
-    {
-      "epoch": 1.0054662912042405,
-      "grad_norm": 479.2508239746094,
-      "learning_rate": 2.489330651844757e-09,
-      "loss": 45.7774,
-      "step": 248870
-    },
-    {
-      "epoch": 1.005506692469608,
-      "grad_norm": 489.208984375,
-      "learning_rate": 2.4454675472807e-09,
-      "loss": 52.9505,
-      "step": 248880
-    },
-    {
-      "epoch": 1.0055470937349757,
-      "grad_norm": 634.3941650390625,
-      "learning_rate": 2.401994305336608e-09,
-      "loss": 67.2182,
-      "step": 248890
-    },
-    {
-      "epoch": 1.0055874950003434,
-      "grad_norm": 834.0480346679688,
-      "learning_rate": 2.3589109268629116e-09,
-      "loss": 61.1101,
-      "step": 248900
-    },
-    {
-      "epoch": 1.005627896265711,
-      "grad_norm": 1048.023193359375,
-      "learning_rate": 2.3162174126967196e-09,
-      "loss": 73.4154,
-      "step": 248910
-    },
-    {
-      "epoch": 1.0056682975310787,
-      "grad_norm": 400.6410827636719,
-      "learning_rate": 2.273913763672919e-09,
-      "loss": 68.9832,
-      "step": 248920
-    },
-    {
-      "epoch": 1.0057086987964463,
-      "grad_norm": 622.4219970703125,
-      "learning_rate": 2.2319999806130753e-09,
-      "loss": 46.3725,
-      "step": 248930
-    },
-    {
-      "epoch": 1.005749100061814,
-      "grad_norm": 1237.3565673828125,
-      "learning_rate": 2.190476064336533e-09,
-      "loss": 79.3537,
-      "step": 248940
-    },
-    {
-      "epoch": 1.0057895013271816,
-      "grad_norm": 1718.1204833984375,
-      "learning_rate": 2.1493420156515345e-09,
-      "loss": 98.4554,
-      "step": 248950
-    },
-    {
-      "epoch": 1.0058299025925492,
-      "grad_norm": 549.4330444335938,
-      "learning_rate": 2.1085978353618807e-09,
-      "loss": 53.079,
-      "step": 248960
-    },
-    {
-      "epoch": 1.0058703038579169,
-      "grad_norm": 1439.4869384765625,
-      "learning_rate": 2.0682435242602716e-09,
-      "loss": 59.7202,
-      "step": 248970
-    },
-    {
-      "epoch": 1.0059107051232845,
-      "grad_norm": 554.4739990234375,
-      "learning_rate": 2.0282790831349653e-09,
-      "loss": 81.9638,
-      "step": 248980
-    },
-    {
-      "epoch": 1.0059511063886521,
-      "grad_norm": 586.0916748046875,
-      "learning_rate": 1.988704512763118e-09,
-      "loss": 53.7479,
-      "step": 248990
-    },
-    {
-      "epoch": 1.0059915076540198,
-      "grad_norm": 230.25294494628906,
-      "learning_rate": 1.949519813915224e-09,
-      "loss": 53.467,
-      "step": 249000
-    },
-    {
-      "epoch": 1.0060319089193874,
-      "grad_norm": 651.7587890625,
-      "learning_rate": 1.910724987359558e-09,
-      "loss": 75.1255,
-      "step": 249010
-    },
-    {
-      "epoch": 1.006072310184755,
-      "grad_norm": 253.41368103027344,
-      "learning_rate": 1.872320033848851e-09,
-      "loss": 70.5722,
-      "step": 249020
-    },
-    {
-      "epoch": 1.0061127114501227,
-      "grad_norm": 165.32437133789062,
-      "learning_rate": 1.8343049541336143e-09,
-      "loss": 37.0831,
-      "step": 249030
-    },
-    {
-      "epoch": 1.0061531127154903,
-      "grad_norm": 707.3920288085938,
-      "learning_rate": 1.796679748953256e-09,
-      "loss": 56.9702,
-      "step": 249040
-    },
-    {
-      "epoch": 1.006193513980858,
-      "grad_norm": 692.0291748046875,
-      "learning_rate": 1.7594444190449645e-09,
-      "loss": 56.6415,
-      "step": 249050
-    },
-    {
-      "epoch": 1.0062339152462254,
-      "grad_norm": 469.0581359863281,
-      "learning_rate": 1.7225989651303842e-09,
-      "loss": 49.4065,
-      "step": 249060
-    },
-    {
-      "epoch": 1.006274316511593,
-      "grad_norm": 331.9457092285156,
-      "learning_rate": 1.68614338792894e-09,
-      "loss": 54.5548,
-      "step": 249070
-    },
-    {
-      "epoch": 1.0063147177769607,
-      "grad_norm": 404.8883056640625,
-      "learning_rate": 1.6500776881511748e-09,
-      "loss": 87.9609,
-      "step": 249080
-    },
-    {
-      "epoch": 1.0063551190423283,
-      "grad_norm": 789.7308959960938,
-      "learning_rate": 1.6144018665031901e-09,
-      "loss": 39.0294,
-      "step": 249090
-    },
-    {
-      "epoch": 1.006395520307696,
-      "grad_norm": 590.8009643554688,
-      "learning_rate": 1.5791159236777654e-09,
-      "loss": 64.4315,
-      "step": 249100
-    },
-    {
-      "epoch": 1.0064359215730636,
-      "grad_norm": 393.2922058105469,
-      "learning_rate": 1.5442198603632386e-09,
-      "loss": 82.9317,
-      "step": 249110
-    },
-    {
-      "epoch": 1.0064763228384312,
-      "grad_norm": 449.51104736328125,
-      "learning_rate": 1.5097136772390663e-09,
-      "loss": 63.2408,
-      "step": 249120
-    },
-    {
-      "epoch": 1.0065167241037989,
-      "grad_norm": 662.6757202148438,
-      "learning_rate": 1.4755973749802642e-09,
-      "loss": 64.8274,
-      "step": 249130
-    },
-    {
-      "epoch": 1.0065571253691665,
-      "grad_norm": 575.337646484375,
-      "learning_rate": 1.441870954250746e-09,
-      "loss": 76.7478,
-      "step": 249140
-    },
-    {
-      "epoch": 1.0065975266345342,
-      "grad_norm": 450.4961853027344,
-      "learning_rate": 1.4085344157055425e-09,
-      "loss": 42.8611,
-      "step": 249150
-    },
-    {
-      "epoch": 1.0066379278999018,
-      "grad_norm": 794.6490478515625,
-      "learning_rate": 1.3755877599996858e-09,
-      "loss": 64.4882,
-      "step": 249160
-    },
-    {
-      "epoch": 1.0066783291652694,
-      "grad_norm": 371.93194580078125,
-      "learning_rate": 1.3430309877726643e-09,
-      "loss": 82.8797,
-      "step": 249170
-    },
-    {
-      "epoch": 1.006718730430637,
-      "grad_norm": 459.426025390625,
-      "learning_rate": 1.3108640996573053e-09,
-      "loss": 80.5951,
-      "step": 249180
-    },
-    {
-      "epoch": 1.0067591316960047,
-      "grad_norm": 829.705322265625,
-      "learning_rate": 1.279087096284215e-09,
-      "loss": 51.7935,
-      "step": 249190
-    },
-    {
-      "epoch": 1.0067995329613724,
-      "grad_norm": 981.8072509765625,
-      "learning_rate": 1.2476999782706779e-09,
-      "loss": 48.9967,
-      "step": 249200
-    },
-    {
-      "epoch": 1.00683993422674,
-      "grad_norm": 662.7955322265625,
-      "learning_rate": 1.2167027462295366e-09,
-      "loss": 48.9169,
-      "step": 249210
-    },
-    {
-      "epoch": 1.0068803354921076,
-      "grad_norm": 498.3249816894531,
-      "learning_rate": 1.186095400764753e-09,
-      "loss": 65.781,
-      "step": 249220
-    },
-    {
-      "epoch": 1.0069207367574753,
-      "grad_norm": 1017.97900390625,
-      "learning_rate": 1.1558779424714063e-09,
-      "loss": 74.366,
-      "step": 249230
-    },
-    {
-      "epoch": 1.006961138022843,
-      "grad_norm": 566.4378662109375,
-      "learning_rate": 1.1260503719423554e-09,
-      "loss": 59.101,
-      "step": 249240
-    },
-    {
-      "epoch": 1.0070015392882106,
-      "grad_norm": 690.8829345703125,
-      "learning_rate": 1.0966126897571373e-09,
-      "loss": 51.221,
-      "step": 249250
-    },
-    {
-      "epoch": 1.0070419405535782,
-      "grad_norm": 530.546875,
-      "learning_rate": 1.0675648964886265e-09,
-      "loss": 71.1756,
-      "step": 249260
-    },
-    {
-      "epoch": 1.0070823418189458,
-      "grad_norm": 482.6839599609375,
-      "learning_rate": 1.0389069927052575e-09,
-      "loss": 62.3818,
-      "step": 249270
-    },
-    {
-      "epoch": 1.0071227430843135,
-      "grad_norm": 567.1624755859375,
-      "learning_rate": 1.0106389789643622e-09,
-      "loss": 61.6863,
-      "step": 249280
-    },
-    {
-      "epoch": 1.0071631443496811,
-      "grad_norm": 506.1968688964844,
-      "learning_rate": 9.827608558166113e-10,
-      "loss": 54.4761,
-      "step": 249290
-    },
-    {
-      "epoch": 1.0072035456150488,
-      "grad_norm": 541.2426147460938,
-      "learning_rate": 9.55272623806014e-10,
-      "loss": 37.5779,
-      "step": 249300
-    },
-    {
-      "epoch": 1.0072439468804162,
-      "grad_norm": 716.0906982421875,
-      "learning_rate": 9.281742834699181e-10,
-      "loss": 70.1744,
-      "step": 249310
-    },
-    {
-      "epoch": 1.0072843481457838,
-      "grad_norm": 703.7621459960938,
-      "learning_rate": 9.014658353323491e-10,
-      "loss": 60.6156,
-      "step": 249320
-    },
-    {
-      "epoch": 1.0073247494111515,
-      "grad_norm": 586.1680297851562,
-      "learning_rate": 8.751472799195527e-10,
-      "loss": 70.0694,
-      "step": 249330
-    },
-    {
-      "epoch": 1.007365150676519,
-      "grad_norm": 1170.073486328125,
-      "learning_rate": 8.492186177422312e-10,
-      "loss": 66.7577,
-      "step": 249340
-    },
-    {
-      "epoch": 1.0074055519418867,
-      "grad_norm": 482.56781005859375,
-      "learning_rate": 8.236798493044262e-10,
-      "loss": 61.0401,
-      "step": 249350
-    },
-    {
-      "epoch": 1.0074459532072544,
-      "grad_norm": 611.2447509765625,
-      "learning_rate": 7.985309751057379e-10,
-      "loss": 48.5732,
-      "step": 249360
-    },
-    {
-      "epoch": 1.007486354472622,
-      "grad_norm": 923.9035034179688,
-      "learning_rate": 7.737719956346646e-10,
-      "loss": 68.4029,
-      "step": 249370
-    },
-    {
-      "epoch": 1.0075267557379897,
-      "grad_norm": 511.7546081542969,
-      "learning_rate": 7.494029113752632e-10,
-      "loss": 48.3581,
-      "step": 249380
-    },
-    {
-      "epoch": 1.0075671570033573,
-      "grad_norm": 711.5328369140625,
-      "learning_rate": 7.254237228049298e-10,
-      "loss": 86.4123,
-      "step": 249390
-    },
-    {
-      "epoch": 1.007607558268725,
-      "grad_norm": 560.9522094726562,
-      "learning_rate": 7.018344303877378e-10,
-      "loss": 41.3287,
-      "step": 249400
-    },
-    {
-      "epoch": 1.0076479595340926,
-      "grad_norm": 568.5785522460938,
-      "learning_rate": 6.786350345833193e-10,
-      "loss": 48.3969,
-      "step": 249410
-    },
-    {
-      "epoch": 1.0076883607994602,
-      "grad_norm": 543.312744140625,
-      "learning_rate": 6.558255358468657e-10,
-      "loss": 57.2176,
-      "step": 249420
-    },
-    {
-      "epoch": 1.0077287620648279,
-      "grad_norm": 801.2297973632812,
-      "learning_rate": 6.334059346202459e-10,
-      "loss": 77.0439,
-      "step": 249430
-    },
-    {
-      "epoch": 1.0077691633301955,
-      "grad_norm": 127.95111846923828,
-      "learning_rate": 6.113762313431082e-10,
-      "loss": 52.8138,
-      "step": 249440
-    },
-    {
-      "epoch": 1.0078095645955631,
-      "grad_norm": 485.751953125,
-      "learning_rate": 5.897364264417782e-10,
-      "loss": 65.7876,
-      "step": 249450
-    },
-    {
-      "epoch": 1.0078499658609308,
-      "grad_norm": 1014.9998168945312,
-      "learning_rate": 5.684865203403611e-10,
-      "loss": 66.254,
-      "step": 249460
-    },
-    {
-      "epoch": 1.0078903671262984,
-      "grad_norm": 660.0430297851562,
-      "learning_rate": 5.476265134540804e-10,
-      "loss": 58.4983,
-      "step": 249470
-    },
-    {
-      "epoch": 1.007930768391666,
-      "grad_norm": 381.69024658203125,
-      "learning_rate": 5.271564061870571e-10,
-      "loss": 68.6676,
-      "step": 249480
-    },
-    {
-      "epoch": 1.0079711696570337,
-      "grad_norm": 328.0639343261719,
-      "learning_rate": 5.070761989411921e-10,
-      "loss": 81.1313,
-      "step": 249490
-    },
-    {
-      "epoch": 1.0080115709224013,
-      "grad_norm": 635.9850463867188,
-      "learning_rate": 4.873858921050634e-10,
-      "loss": 52.8049,
-      "step": 249500
-    },
-    {
-      "epoch": 1.008051972187769,
-      "grad_norm": 556.7789306640625,
-      "learning_rate": 4.680854860628081e-10,
-      "loss": 54.1077,
-      "step": 249510
-    },
-    {
-      "epoch": 1.0080923734531366,
-      "grad_norm": 1849.8094482421875,
-      "learning_rate": 4.491749811919022e-10,
-      "loss": 68.67,
-      "step": 249520
-    },
-    {
-      "epoch": 1.0081327747185043,
-      "grad_norm": 424.52734375,
-      "learning_rate": 4.3065437786316e-10,
-      "loss": 57.378,
-      "step": 249530
-    },
-    {
-      "epoch": 1.008173175983872,
-      "grad_norm": 415.8068542480469,
-      "learning_rate": 4.1252367643185297e-10,
-      "loss": 45.5723,
-      "step": 249540
-    },
-    {
-      "epoch": 1.0082135772492395,
-      "grad_norm": 650.1827392578125,
-      "learning_rate": 3.9478287725769337e-10,
-      "loss": 53.6675,
-      "step": 249550
-    },
-    {
-      "epoch": 1.0082539785146072,
-      "grad_norm": 313.34033203125,
-      "learning_rate": 3.7743198068262986e-10,
-      "loss": 51.5655,
-      "step": 249560
-    },
-    {
-      "epoch": 1.0082943797799746,
-      "grad_norm": 754.9653930664062,
-      "learning_rate": 3.604709870463907e-10,
-      "loss": 60.8922,
-      "step": 249570
-    },
-    {
-      "epoch": 1.0083347810453422,
-      "grad_norm": 707.226806640625,
-      "learning_rate": 3.438998966776019e-10,
-      "loss": 64.934,
-      "step": 249580
-    },
-    {
-      "epoch": 1.0083751823107099,
-      "grad_norm": 1383.1534423828125,
-      "learning_rate": 3.277187099026691e-10,
-      "loss": 90.5521,
-      "step": 249590
-    },
-    {
-      "epoch": 1.0084155835760775,
-      "grad_norm": 527.865966796875,
-      "learning_rate": 3.11927427034675e-10,
-      "loss": 70.7257,
-      "step": 249600
-    },
-    {
-      "epoch": 1.0084559848414452,
-      "grad_norm": 454.8075256347656,
-      "learning_rate": 2.965260483844823e-10,
-      "loss": 71.8497,
-      "step": 249610
-    },
-    {
-      "epoch": 1.0084963861068128,
-      "grad_norm": 461.6095275878906,
-      "learning_rate": 2.8151457424741014e-10,
-      "loss": 55.6032,
-      "step": 249620
-    },
-    {
-      "epoch": 1.0085367873721804,
-      "grad_norm": 906.763916015625,
-      "learning_rate": 2.668930049187779e-10,
-      "loss": 84.5327,
-      "step": 249630
-    },
-    {
-      "epoch": 1.008577188637548,
-      "grad_norm": 281.6214904785156,
-      "learning_rate": 2.526613406850231e-10,
-      "loss": 47.5958,
-      "step": 249640
-    },
-    {
-      "epoch": 1.0086175899029157,
-      "grad_norm": 621.9278564453125,
-      "learning_rate": 2.388195818214811e-10,
-      "loss": 52.0255,
-      "step": 249650
-    },
-    {
-      "epoch": 1.0086579911682834,
-      "grad_norm": 814.9580688476562,
-      "learning_rate": 2.2536772859904632e-10,
-      "loss": 50.8963,
-      "step": 249660
-    },
-    {
-      "epoch": 1.008698392433651,
-      "grad_norm": 566.6548461914062,
-      "learning_rate": 2.1230578127973134e-10,
-      "loss": 70.7275,
-      "step": 249670
-    },
-    {
-      "epoch": 1.0087387936990186,
-      "grad_norm": 1687.9730224609375,
-      "learning_rate": 1.9963374011666703e-10,
-      "loss": 78.5721,
-      "step": 249680
-    },
-    {
-      "epoch": 1.0087791949643863,
-      "grad_norm": 486.9939270019531,
-      "learning_rate": 1.873516053607638e-10,
-      "loss": 57.9145,
-      "step": 249690
-    },
-    {
-      "epoch": 1.008819596229754,
-      "grad_norm": 481.72943115234375,
-      "learning_rate": 1.7545937724738894e-10,
-      "loss": 57.3564,
-      "step": 249700
-    },
-    {
-      "epoch": 1.0088599974951216,
-      "grad_norm": 582.765869140625,
-      "learning_rate": 1.6395705600968926e-10,
-      "loss": 66.8067,
-      "step": 249710
-    },
-    {
-      "epoch": 1.0089003987604892,
-      "grad_norm": 722.7216186523438,
-      "learning_rate": 1.5284464187192981e-10,
-      "loss": 86.6796,
-      "step": 249720
-    },
-    {
-      "epoch": 1.0089408000258568,
-      "grad_norm": 423.3017578125,
-      "learning_rate": 1.421221350539348e-10,
-      "loss": 36.3374,
-      "step": 249730
-    },
-    {
-      "epoch": 1.0089812012912245,
-      "grad_norm": 765.08056640625,
-      "learning_rate": 1.317895357599852e-10,
-      "loss": 55.4276,
-      "step": 249740
-    },
-    {
-      "epoch": 1.0090216025565921,
-      "grad_norm": 458.633544921875,
-      "learning_rate": 1.2184684419214166e-10,
-      "loss": 62.0634,
-      "step": 249750
-    },
-    {
-      "epoch": 1.0090620038219598,
-      "grad_norm": 512.1096801757812,
-      "learning_rate": 1.1229406054802383e-10,
-      "loss": 72.1461,
-      "step": 249760
-    },
-    {
-      "epoch": 1.0091024050873274,
-      "grad_norm": 575.4948120117188,
-      "learning_rate": 1.0313118500970831e-10,
-      "loss": 57.1718,
-      "step": 249770
-    },
-    {
-      "epoch": 1.009142806352695,
-      "grad_norm": 362.7676086425781,
-      "learning_rate": 9.435821775705123e-11,
-      "loss": 44.035,
-      "step": 249780
-    },
-    {
-      "epoch": 1.0091832076180627,
-      "grad_norm": 818.1105346679688,
-      "learning_rate": 8.597515896324737e-11,
-      "loss": 71.8585,
-      "step": 249790
-    },
-    {
-      "epoch": 1.0092236088834303,
-      "grad_norm": 736.1803588867188,
-      "learning_rate": 7.798200878816886e-11,
-      "loss": 87.0728,
-      "step": 249800
-    },
-    {
-      "epoch": 1.009264010148798,
-      "grad_norm": 216.5111541748047,
-      "learning_rate": 7.03787673916878e-11,
-      "loss": 63.7444,
-      "step": 249810
-    },
-    {
-      "epoch": 1.0093044114141656,
-      "grad_norm": 384.854248046875,
-      "learning_rate": 6.31654349181332e-11,
-      "loss": 53.48,
-      "step": 249820
-    },
-    {
-      "epoch": 1.009344812679533,
-      "grad_norm": 495.5076904296875,
-      "learning_rate": 5.63420115096136e-11,
-      "loss": 63.2487,
-      "step": 249830
-    },
-    {
-      "epoch": 1.0093852139449007,
-      "grad_norm": 579.0645751953125,
-      "learning_rate": 4.9908497301576206e-11,
-      "loss": 50.9092,
-      "step": 249840
-    },
-    {
-      "epoch": 1.0094256152102683,
-      "grad_norm": 675.76953125,
-      "learning_rate": 4.386489241392511e-11,
-      "loss": 63.6097,
-      "step": 249850
-    },
-    {
-      "epoch": 1.009466016475636,
-      "grad_norm": 380.5902404785156,
-      "learning_rate": 3.821119697100528e-11,
-      "loss": 63.8182,
-      "step": 249860
-    },
-    {
-      "epoch": 1.0095064177410036,
-      "grad_norm": 355.2414245605469,
-      "learning_rate": 3.2947411077177694e-11,
-      "loss": 55.9546,
-      "step": 249870
-    },
-    {
-      "epoch": 1.0095468190063712,
-      "grad_norm": 267.0750427246094,
-      "learning_rate": 2.8073534839023753e-11,
-      "loss": 82.4405,
-      "step": 249880
-    },
-    {
-      "epoch": 1.0095872202717389,
-      "grad_norm": 302.30224609375,
-      "learning_rate": 2.358956835202264e-11,
-      "loss": 58.0008,
-      "step": 249890
-    },
-    {
-      "epoch": 1.0096276215371065,
-      "grad_norm": 669.5781860351562,
-      "learning_rate": 1.949551169833086e-11,
-      "loss": 48.5882,
-      "step": 249900
-    },
-    {
-      "epoch": 1.0096680228024741,
-      "grad_norm": 618.9566040039062,
-      "learning_rate": 1.579136496454581e-11,
-      "loss": 39.3103,
-      "step": 249910
-    },
-    {
-      "epoch": 1.0097084240678418,
-      "grad_norm": 2055.600830078125,
-      "learning_rate": 1.2477128217280864e-11,
-      "loss": 63.2549,
-      "step": 249920
-    },
-    {
-      "epoch": 1.0097488253332094,
-      "grad_norm": 954.0150756835938,
-      "learning_rate": 9.552801523149412e-12,
-      "loss": 67.9715,
-      "step": 249930
-    },
-    {
-      "epoch": 1.009789226598577,
-      "grad_norm": 901.2742309570312,
-      "learning_rate": 7.018384942103496e-12,
-      "loss": 66.1644,
-      "step": 249940
-    },
-    {
-      "epoch": 1.0098296278639447,
-      "grad_norm": 663.9369506835938,
-      "learning_rate": 4.873878518552033e-12,
-      "loss": 60.1538,
-      "step": 249950
-    },
-    {
-      "epoch": 1.0098700291293123,
-      "grad_norm": 809.4927978515625,
-      "learning_rate": 3.119282296903947e-12,
-      "loss": 78.8019,
-      "step": 249960
-    },
-    {
-      "epoch": 1.00991043039468,
-      "grad_norm": 347.0174865722656,
-      "learning_rate": 1.754596312686374e-12,
-      "loss": 54.3162,
-      "step": 249970
-    },
-    {
-      "epoch": 1.0099508316600476,
-      "grad_norm": 259.03460693359375,
-      "learning_rate": 7.798205903242206e-13,
-      "loss": 81.9569,
-      "step": 249980
-    },
-    {
-      "epoch": 1.0099912329254153,
-      "grad_norm": 841.800537109375,
-      "learning_rate": 1.9495514758105516e-13,
-      "loss": 61.0793,
-      "step": 249990
-    },
-    {
-      "epoch": 1.010031634190783,
-      "grad_norm": 511.8416442871094,
-      "learning_rate": 0.0,
-      "loss": 39.5704,
-      "step": 250000
     }
   ],
   "logging_steps": 10,
@@ -175029,7 +173629,7 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": true
+        "should_training_stop": false
       },
       "attributes": {}
     }