Image-Text-to-Text
Transformers
Safetensors
visgym_model / mixed_freeze_language /trainer_state.json
Junyi42's picture
Upload mixed_freeze_language from Junyi42/gym
460bb8c verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0282,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002,
"grad_norm": 0.992521345615387,
"learning_rate": 1.8e-07,
"loss": 0.4669,
"step": 10
},
{
"epoch": 0.004,
"grad_norm": 2.713249444961548,
"learning_rate": 3.8e-07,
"loss": 0.46,
"step": 20
},
{
"epoch": 0.006,
"grad_norm": 6.9339165687561035,
"learning_rate": 5.800000000000001e-07,
"loss": 0.4543,
"step": 30
},
{
"epoch": 0.008,
"grad_norm": 1.508054494857788,
"learning_rate": 7.8e-07,
"loss": 0.4686,
"step": 40
},
{
"epoch": 0.01,
"grad_norm": 3.1499743461608887,
"learning_rate": 9.800000000000001e-07,
"loss": 0.4627,
"step": 50
},
{
"epoch": 0.012,
"grad_norm": 2.000009298324585,
"learning_rate": 1.1800000000000001e-06,
"loss": 0.4668,
"step": 60
},
{
"epoch": 0.014,
"grad_norm": 2.90346097946167,
"learning_rate": 1.3800000000000001e-06,
"loss": 0.4672,
"step": 70
},
{
"epoch": 0.016,
"grad_norm": 1.8733686208724976,
"learning_rate": 1.5800000000000001e-06,
"loss": 0.442,
"step": 80
},
{
"epoch": 0.018,
"grad_norm": 7.299347400665283,
"learning_rate": 1.7800000000000001e-06,
"loss": 0.4537,
"step": 90
},
{
"epoch": 0.02,
"grad_norm": 1.2114304304122925,
"learning_rate": 1.98e-06,
"loss": 0.4392,
"step": 100
},
{
"epoch": 0.022,
"grad_norm": 4.244724273681641,
"learning_rate": 2.1800000000000003e-06,
"loss": 0.4467,
"step": 110
},
{
"epoch": 0.024,
"grad_norm": 1.6745773553848267,
"learning_rate": 2.38e-06,
"loss": 0.3962,
"step": 120
},
{
"epoch": 0.026,
"grad_norm": 2.175783395767212,
"learning_rate": 2.5800000000000003e-06,
"loss": 0.3613,
"step": 130
},
{
"epoch": 0.028,
"grad_norm": 1.944590449333191,
"learning_rate": 2.7800000000000005e-06,
"loss": 0.267,
"step": 140
},
{
"epoch": 0.03,
"grad_norm": 0.9925225377082825,
"learning_rate": 2.9800000000000003e-06,
"loss": 0.2213,
"step": 150
},
{
"epoch": 0.032,
"grad_norm": 1.1854593753814697,
"learning_rate": 3.1800000000000005e-06,
"loss": 0.2109,
"step": 160
},
{
"epoch": 0.034,
"grad_norm": 1.1748063564300537,
"learning_rate": 3.3800000000000007e-06,
"loss": 0.2003,
"step": 170
},
{
"epoch": 0.036,
"grad_norm": 1.274276852607727,
"learning_rate": 3.58e-06,
"loss": 0.1888,
"step": 180
},
{
"epoch": 0.038,
"grad_norm": 1.1525161266326904,
"learning_rate": 3.7800000000000002e-06,
"loss": 0.1868,
"step": 190
},
{
"epoch": 0.04,
"grad_norm": 1.7498775720596313,
"learning_rate": 3.980000000000001e-06,
"loss": 0.1841,
"step": 200
},
{
"epoch": 0.042,
"grad_norm": 0.8628786206245422,
"learning_rate": 4.18e-06,
"loss": 0.1798,
"step": 210
},
{
"epoch": 0.044,
"grad_norm": 2.097649097442627,
"learning_rate": 4.38e-06,
"loss": 0.1746,
"step": 220
},
{
"epoch": 0.046,
"grad_norm": 3.168224811553955,
"learning_rate": 4.58e-06,
"loss": 0.1718,
"step": 230
},
{
"epoch": 0.048,
"grad_norm": 1.1456739902496338,
"learning_rate": 4.78e-06,
"loss": 0.1761,
"step": 240
},
{
"epoch": 0.05,
"grad_norm": 1.7046711444854736,
"learning_rate": 4.980000000000001e-06,
"loss": 0.1715,
"step": 250
},
{
"epoch": 0.052,
"grad_norm": 1.0530240535736084,
"learning_rate": 5.18e-06,
"loss": 0.1646,
"step": 260
},
{
"epoch": 0.054,
"grad_norm": 1.4000672101974487,
"learning_rate": 5.380000000000001e-06,
"loss": 0.169,
"step": 270
},
{
"epoch": 0.056,
"grad_norm": 0.9671304821968079,
"learning_rate": 5.580000000000001e-06,
"loss": 0.1658,
"step": 280
},
{
"epoch": 0.058,
"grad_norm": 4.116588115692139,
"learning_rate": 5.78e-06,
"loss": 0.1633,
"step": 290
},
{
"epoch": 0.06,
"grad_norm": 1.6324472427368164,
"learning_rate": 5.98e-06,
"loss": 0.1686,
"step": 300
},
{
"epoch": 0.062,
"grad_norm": 1.066884160041809,
"learning_rate": 6.18e-06,
"loss": 0.1578,
"step": 310
},
{
"epoch": 0.064,
"grad_norm": 0.9248769879341125,
"learning_rate": 6.380000000000001e-06,
"loss": 0.1648,
"step": 320
},
{
"epoch": 0.066,
"grad_norm": 2.3996987342834473,
"learning_rate": 6.5800000000000005e-06,
"loss": 0.1584,
"step": 330
},
{
"epoch": 0.068,
"grad_norm": 2.0868091583251953,
"learning_rate": 6.780000000000001e-06,
"loss": 0.1497,
"step": 340
},
{
"epoch": 0.07,
"grad_norm": 3.3375916481018066,
"learning_rate": 6.98e-06,
"loss": 0.1576,
"step": 350
},
{
"epoch": 0.072,
"grad_norm": 1.1031235456466675,
"learning_rate": 7.180000000000001e-06,
"loss": 0.1588,
"step": 360
},
{
"epoch": 0.074,
"grad_norm": 6.909502983093262,
"learning_rate": 7.3800000000000005e-06,
"loss": 0.1505,
"step": 370
},
{
"epoch": 0.076,
"grad_norm": 1.235683560371399,
"learning_rate": 7.58e-06,
"loss": 0.1465,
"step": 380
},
{
"epoch": 0.078,
"grad_norm": 1.2270852327346802,
"learning_rate": 7.78e-06,
"loss": 0.1486,
"step": 390
},
{
"epoch": 0.08,
"grad_norm": 1.0695114135742188,
"learning_rate": 7.980000000000002e-06,
"loss": 0.1525,
"step": 400
},
{
"epoch": 0.082,
"grad_norm": 2.8632586002349854,
"learning_rate": 8.18e-06,
"loss": 0.147,
"step": 410
},
{
"epoch": 0.084,
"grad_norm": 1.0436551570892334,
"learning_rate": 8.380000000000001e-06,
"loss": 0.1445,
"step": 420
},
{
"epoch": 0.086,
"grad_norm": 1.166566014289856,
"learning_rate": 8.580000000000001e-06,
"loss": 0.1542,
"step": 430
},
{
"epoch": 0.088,
"grad_norm": 1.5501545667648315,
"learning_rate": 8.78e-06,
"loss": 0.1512,
"step": 440
},
{
"epoch": 0.09,
"grad_norm": 1.8002007007598877,
"learning_rate": 8.98e-06,
"loss": 0.1416,
"step": 450
},
{
"epoch": 0.092,
"grad_norm": 0.8401817083358765,
"learning_rate": 9.180000000000002e-06,
"loss": 0.1517,
"step": 460
},
{
"epoch": 0.094,
"grad_norm": 1.2032701969146729,
"learning_rate": 9.38e-06,
"loss": 0.1451,
"step": 470
},
{
"epoch": 0.096,
"grad_norm": 0.753799319267273,
"learning_rate": 9.58e-06,
"loss": 0.1401,
"step": 480
},
{
"epoch": 0.098,
"grad_norm": 1.1318485736846924,
"learning_rate": 9.780000000000001e-06,
"loss": 0.1378,
"step": 490
},
{
"epoch": 0.1,
"grad_norm": 2.041433572769165,
"learning_rate": 9.980000000000001e-06,
"loss": 0.1459,
"step": 500
},
{
"epoch": 0.102,
"grad_norm": 2.9767746925354004,
"learning_rate": 9.999901304280686e-06,
"loss": 0.1432,
"step": 510
},
{
"epoch": 0.104,
"grad_norm": 0.9433181881904602,
"learning_rate": 9.999560138895238e-06,
"loss": 0.1441,
"step": 520
},
{
"epoch": 0.106,
"grad_norm": 2.254661798477173,
"learning_rate": 9.99897530200195e-06,
"loss": 0.1411,
"step": 530
},
{
"epoch": 0.108,
"grad_norm": 1.0642566680908203,
"learning_rate": 9.998146822104943e-06,
"loss": 0.146,
"step": 540
},
{
"epoch": 0.11,
"grad_norm": 0.9729284048080444,
"learning_rate": 9.997074739583162e-06,
"loss": 0.1404,
"step": 550
},
{
"epoch": 0.112,
"grad_norm": 0.6401363015174866,
"learning_rate": 9.995759106688394e-06,
"loss": 0.1354,
"step": 560
},
{
"epoch": 0.114,
"grad_norm": 1.005709171295166,
"learning_rate": 9.99419998754273e-06,
"loss": 0.1421,
"step": 570
},
{
"epoch": 0.116,
"grad_norm": 0.7156584858894348,
"learning_rate": 9.992397458135438e-06,
"loss": 0.144,
"step": 580
},
{
"epoch": 0.118,
"grad_norm": 0.9248397946357727,
"learning_rate": 9.990351606319261e-06,
"loss": 0.1362,
"step": 590
},
{
"epoch": 0.12,
"grad_norm": 1.142520785331726,
"learning_rate": 9.988062531806127e-06,
"loss": 0.1333,
"step": 600
},
{
"epoch": 0.122,
"grad_norm": 0.9766596555709839,
"learning_rate": 9.9855303461623e-06,
"loss": 0.1329,
"step": 610
},
{
"epoch": 0.124,
"grad_norm": 0.8085676431655884,
"learning_rate": 9.982755172802933e-06,
"loss": 0.1343,
"step": 620
},
{
"epoch": 0.126,
"grad_norm": 1.0947911739349365,
"learning_rate": 9.979737146986064e-06,
"loss": 0.1312,
"step": 630
},
{
"epoch": 0.128,
"grad_norm": 0.9679737091064453,
"learning_rate": 9.976476415806013e-06,
"loss": 0.1391,
"step": 640
},
{
"epoch": 0.13,
"grad_norm": 0.8831957578659058,
"learning_rate": 9.972973138186217e-06,
"loss": 0.14,
"step": 650
},
{
"epoch": 0.132,
"grad_norm": 1.0070087909698486,
"learning_rate": 9.969227484871485e-06,
"loss": 0.1343,
"step": 660
},
{
"epoch": 0.134,
"grad_norm": 0.8003564476966858,
"learning_rate": 9.965239638419673e-06,
"loss": 0.1372,
"step": 670
},
{
"epoch": 0.136,
"grad_norm": 0.680647611618042,
"learning_rate": 9.961009793192793e-06,
"loss": 0.1395,
"step": 680
},
{
"epoch": 0.138,
"grad_norm": 1.1533617973327637,
"learning_rate": 9.956538155347534e-06,
"loss": 0.1398,
"step": 690
},
{
"epoch": 0.14,
"grad_norm": 1.7727952003479004,
"learning_rate": 9.951824942825215e-06,
"loss": 0.1331,
"step": 700
},
{
"epoch": 0.142,
"grad_norm": 0.8858327269554138,
"learning_rate": 9.946870385341167e-06,
"loss": 0.1282,
"step": 710
},
{
"epoch": 0.144,
"grad_norm": 0.917654812335968,
"learning_rate": 9.94167472437353e-06,
"loss": 0.1411,
"step": 720
},
{
"epoch": 0.146,
"grad_norm": 1.409077763557434,
"learning_rate": 9.936238213151491e-06,
"loss": 0.1263,
"step": 730
},
{
"epoch": 0.148,
"grad_norm": 0.8429045081138611,
"learning_rate": 9.930561116642936e-06,
"loss": 0.1339,
"step": 740
},
{
"epoch": 0.15,
"grad_norm": 1.2110062837600708,
"learning_rate": 9.92464371154154e-06,
"loss": 0.1302,
"step": 750
},
{
"epoch": 0.152,
"grad_norm": 1.842542052268982,
"learning_rate": 9.918486286253279e-06,
"loss": 0.1335,
"step": 760
},
{
"epoch": 0.154,
"grad_norm": 0.8756665587425232,
"learning_rate": 9.912089140882377e-06,
"loss": 0.1315,
"step": 770
},
{
"epoch": 0.156,
"grad_norm": 0.587119996547699,
"learning_rate": 9.90545258721667e-06,
"loss": 0.1294,
"step": 780
},
{
"epoch": 0.158,
"grad_norm": 1.0943330526351929,
"learning_rate": 9.898576948712427e-06,
"loss": 0.134,
"step": 790
},
{
"epoch": 0.16,
"grad_norm": 0.5449488759040833,
"learning_rate": 9.891462560478562e-06,
"loss": 0.1273,
"step": 800
},
{
"epoch": 0.162,
"grad_norm": 1.0867772102355957,
"learning_rate": 9.884109769260326e-06,
"loss": 0.136,
"step": 810
},
{
"epoch": 0.164,
"grad_norm": 0.6655702590942383,
"learning_rate": 9.876518933422385e-06,
"loss": 0.1345,
"step": 820
},
{
"epoch": 0.166,
"grad_norm": 0.9336721897125244,
"learning_rate": 9.868690422931372e-06,
"loss": 0.1377,
"step": 830
},
{
"epoch": 0.168,
"grad_norm": 0.5859615206718445,
"learning_rate": 9.860624619337844e-06,
"loss": 0.131,
"step": 840
},
{
"epoch": 0.17,
"grad_norm": 1.1051822900772095,
"learning_rate": 9.852321915757688e-06,
"loss": 0.1359,
"step": 850
},
{
"epoch": 0.172,
"grad_norm": 0.5312075614929199,
"learning_rate": 9.843782716852963e-06,
"loss": 0.1364,
"step": 860
},
{
"epoch": 0.174,
"grad_norm": 0.5661536455154419,
"learning_rate": 9.835007438812177e-06,
"loss": 0.134,
"step": 870
},
{
"epoch": 0.176,
"grad_norm": 0.4673188626766205,
"learning_rate": 9.825996509330001e-06,
"loss": 0.1311,
"step": 880
},
{
"epoch": 0.178,
"grad_norm": 0.7434860467910767,
"learning_rate": 9.816750367586424e-06,
"loss": 0.1333,
"step": 890
},
{
"epoch": 0.18,
"grad_norm": 1.071921467781067,
"learning_rate": 9.807269464225355e-06,
"loss": 0.1293,
"step": 900
},
{
"epoch": 0.182,
"grad_norm": 0.7527667284011841,
"learning_rate": 9.797554261332644e-06,
"loss": 0.1291,
"step": 910
},
{
"epoch": 0.184,
"grad_norm": 0.7582711577415466,
"learning_rate": 9.787605232413575e-06,
"loss": 0.1314,
"step": 920
},
{
"epoch": 0.186,
"grad_norm": 0.7601514458656311,
"learning_rate": 9.777422862369782e-06,
"loss": 0.1428,
"step": 930
},
{
"epoch": 0.188,
"grad_norm": 1.272291898727417,
"learning_rate": 9.767007647475618e-06,
"loss": 0.1241,
"step": 940
},
{
"epoch": 0.19,
"grad_norm": 0.4984325170516968,
"learning_rate": 9.756360095353957e-06,
"loss": 0.1293,
"step": 950
},
{
"epoch": 0.192,
"grad_norm": 0.3648125231266022,
"learning_rate": 9.745480724951473e-06,
"loss": 0.1278,
"step": 960
},
{
"epoch": 0.194,
"grad_norm": 0.5616574287414551,
"learning_rate": 9.73437006651333e-06,
"loss": 0.1327,
"step": 970
},
{
"epoch": 0.196,
"grad_norm": 0.4720121920108795,
"learning_rate": 9.723028661557345e-06,
"loss": 0.1258,
"step": 980
},
{
"epoch": 0.198,
"grad_norm": 0.4209226071834564,
"learning_rate": 9.711457062847596e-06,
"loss": 0.1272,
"step": 990
},
{
"epoch": 0.2,
"grad_norm": 18.944120407104492,
"learning_rate": 9.699655834367479e-06,
"loss": 0.1307,
"step": 1000
},
{
"epoch": 0.202,
"grad_norm": 0.4856315851211548,
"learning_rate": 9.687625551292219e-06,
"loss": 0.1322,
"step": 1010
},
{
"epoch": 0.204,
"grad_norm": 0.5057471990585327,
"learning_rate": 9.675366799960842e-06,
"loss": 0.1271,
"step": 1020
},
{
"epoch": 0.206,
"grad_norm": 0.41341617703437805,
"learning_rate": 9.662880177847595e-06,
"loss": 0.1313,
"step": 1030
},
{
"epoch": 0.208,
"grad_norm": 1.8846428394317627,
"learning_rate": 9.650166293532822e-06,
"loss": 0.1248,
"step": 1040
},
{
"epoch": 0.21,
"grad_norm": 0.40669843554496765,
"learning_rate": 9.637225766673309e-06,
"loss": 0.1219,
"step": 1050
},
{
"epoch": 0.212,
"grad_norm": 0.7833669781684875,
"learning_rate": 9.624059227972077e-06,
"loss": 0.1326,
"step": 1060
},
{
"epoch": 0.214,
"grad_norm": 1.194365382194519,
"learning_rate": 9.610667319147648e-06,
"loss": 0.1298,
"step": 1070
},
{
"epoch": 0.216,
"grad_norm": 0.47751522064208984,
"learning_rate": 9.597050692902765e-06,
"loss": 0.1254,
"step": 1080
},
{
"epoch": 0.218,
"grad_norm": 0.7141849398612976,
"learning_rate": 9.583210012892582e-06,
"loss": 0.1332,
"step": 1090
},
{
"epoch": 0.22,
"grad_norm": 1.166374921798706,
"learning_rate": 9.569145953692316e-06,
"loss": 0.1205,
"step": 1100
},
{
"epoch": 0.222,
"grad_norm": 0.33305805921554565,
"learning_rate": 9.554859200764371e-06,
"loss": 0.1242,
"step": 1110
},
{
"epoch": 0.224,
"grad_norm": 0.3989585340023041,
"learning_rate": 9.540350450424927e-06,
"loss": 0.1253,
"step": 1120
},
{
"epoch": 0.226,
"grad_norm": 0.503139853477478,
"learning_rate": 9.525620409810009e-06,
"loss": 0.1226,
"step": 1130
},
{
"epoch": 0.228,
"grad_norm": 0.4812481105327606,
"learning_rate": 9.510669796841014e-06,
"loss": 0.1222,
"step": 1140
},
{
"epoch": 0.23,
"grad_norm": 0.9657276272773743,
"learning_rate": 9.495499340189729e-06,
"loss": 0.1256,
"step": 1150
},
{
"epoch": 0.232,
"grad_norm": 0.4245375096797943,
"learning_rate": 9.480109779242805e-06,
"loss": 0.1333,
"step": 1160
},
{
"epoch": 0.234,
"grad_norm": 0.694359302520752,
"learning_rate": 9.464501864065735e-06,
"loss": 0.123,
"step": 1170
},
{
"epoch": 0.236,
"grad_norm": 0.5880414843559265,
"learning_rate": 9.448676355366282e-06,
"loss": 0.1263,
"step": 1180
},
{
"epoch": 0.238,
"grad_norm": 0.6939809322357178,
"learning_rate": 9.432634024457414e-06,
"loss": 0.1211,
"step": 1190
},
{
"epoch": 0.24,
"grad_norm": 0.87459796667099,
"learning_rate": 9.41637565321971e-06,
"loss": 0.1176,
"step": 1200
},
{
"epoch": 0.242,
"grad_norm": 1.157516598701477,
"learning_rate": 9.399902034063244e-06,
"loss": 0.1196,
"step": 1210
},
{
"epoch": 0.244,
"grad_norm": 0.602496862411499,
"learning_rate": 9.383213969888972e-06,
"loss": 0.1291,
"step": 1220
},
{
"epoch": 0.246,
"grad_norm": 0.8280712366104126,
"learning_rate": 9.366312274049602e-06,
"loss": 0.1154,
"step": 1230
},
{
"epoch": 0.248,
"grad_norm": 0.6725120544433594,
"learning_rate": 9.349197770309942e-06,
"loss": 0.1214,
"step": 1240
},
{
"epoch": 0.25,
"grad_norm": 0.46815815567970276,
"learning_rate": 9.33187129280676e-06,
"loss": 0.123,
"step": 1250
},
{
"epoch": 0.252,
"grad_norm": 0.4778998792171478,
"learning_rate": 9.314333686008125e-06,
"loss": 0.1268,
"step": 1260
},
{
"epoch": 0.254,
"grad_norm": 0.6865217685699463,
"learning_rate": 9.296585804672253e-06,
"loss": 0.1226,
"step": 1270
},
{
"epoch": 0.256,
"grad_norm": 0.408330500125885,
"learning_rate": 9.278628513805838e-06,
"loss": 0.1173,
"step": 1280
},
{
"epoch": 0.258,
"grad_norm": 0.4890899956226349,
"learning_rate": 9.260462688621906e-06,
"loss": 0.1242,
"step": 1290
},
{
"epoch": 0.26,
"grad_norm": 0.4487694501876831,
"learning_rate": 9.242089214497146e-06,
"loss": 0.1199,
"step": 1300
},
{
"epoch": 0.262,
"grad_norm": 0.5977551937103271,
"learning_rate": 9.223508986928766e-06,
"loss": 0.119,
"step": 1310
},
{
"epoch": 0.264,
"grad_norm": 0.5191870331764221,
"learning_rate": 9.204722911490847e-06,
"loss": 0.1209,
"step": 1320
},
{
"epoch": 0.266,
"grad_norm": 0.427013635635376,
"learning_rate": 9.1857319037902e-06,
"loss": 0.1208,
"step": 1330
},
{
"epoch": 0.268,
"grad_norm": 0.7320643663406372,
"learning_rate": 9.16653688942175e-06,
"loss": 0.13,
"step": 1340
},
{
"epoch": 0.27,
"grad_norm": 0.5645836591720581,
"learning_rate": 9.147138803923417e-06,
"loss": 0.126,
"step": 1350
},
{
"epoch": 0.272,
"grad_norm": 0.6136955618858337,
"learning_rate": 9.12753859273052e-06,
"loss": 0.1193,
"step": 1360
},
{
"epoch": 0.274,
"grad_norm": 0.884530782699585,
"learning_rate": 9.107737211129702e-06,
"loss": 0.119,
"step": 1370
},
{
"epoch": 0.276,
"grad_norm": 0.5332031846046448,
"learning_rate": 9.087735624212365e-06,
"loss": 0.1256,
"step": 1380
},
{
"epoch": 0.278,
"grad_norm": 0.4672992527484894,
"learning_rate": 9.06753480682764e-06,
"loss": 0.1284,
"step": 1390
},
{
"epoch": 0.28,
"grad_norm": 0.3032495379447937,
"learning_rate": 9.047135743534866e-06,
"loss": 0.1237,
"step": 1400
},
{
"epoch": 0.282,
"grad_norm": 0.2789173126220703,
"learning_rate": 9.026539428555609e-06,
"loss": 0.1228,
"step": 1410
},
{
"epoch": 0.284,
"grad_norm": 0.4365665316581726,
"learning_rate": 9.005746865725206e-06,
"loss": 0.1195,
"step": 1420
},
{
"epoch": 0.286,
"grad_norm": 0.4817768633365631,
"learning_rate": 8.984759068443832e-06,
"loss": 0.1268,
"step": 1430
},
{
"epoch": 0.288,
"grad_norm": 0.5241479873657227,
"learning_rate": 8.963577059627117e-06,
"loss": 0.1227,
"step": 1440
},
{
"epoch": 0.29,
"grad_norm": 0.40581586956977844,
"learning_rate": 8.942201871656292e-06,
"loss": 0.1251,
"step": 1450
},
{
"epoch": 0.292,
"grad_norm": 0.3088854253292084,
"learning_rate": 8.920634546327857e-06,
"loss": 0.1194,
"step": 1460
},
{
"epoch": 0.294,
"grad_norm": 0.40653863549232483,
"learning_rate": 8.898876134802827e-06,
"loss": 0.1209,
"step": 1470
},
{
"epoch": 0.296,
"grad_norm": 0.3182103633880615,
"learning_rate": 8.87692769755548e-06,
"loss": 0.1222,
"step": 1480
},
{
"epoch": 0.298,
"grad_norm": 1.059717059135437,
"learning_rate": 8.854790304321682e-06,
"loss": 0.1229,
"step": 1490
},
{
"epoch": 0.3,
"grad_norm": 0.4621742069721222,
"learning_rate": 8.83246503404675e-06,
"loss": 0.1279,
"step": 1500
},
{
"epoch": 0.302,
"grad_norm": 0.3916736841201782,
"learning_rate": 8.80995297483286e-06,
"loss": 0.1233,
"step": 1510
},
{
"epoch": 0.304,
"grad_norm": 0.32610222697257996,
"learning_rate": 8.78725522388602e-06,
"loss": 0.1193,
"step": 1520
},
{
"epoch": 0.306,
"grad_norm": 0.5615723133087158,
"learning_rate": 8.764372887462587e-06,
"loss": 0.1235,
"step": 1530
},
{
"epoch": 0.308,
"grad_norm": 0.5475795865058899,
"learning_rate": 8.741307080815357e-06,
"loss": 0.1207,
"step": 1540
},
{
"epoch": 0.31,
"grad_norm": 0.30787885189056396,
"learning_rate": 8.718058928139205e-06,
"loss": 0.1206,
"step": 1550
},
{
"epoch": 0.312,
"grad_norm": 0.8395270705223083,
"learning_rate": 8.694629562516295e-06,
"loss": 0.1262,
"step": 1560
},
{
"epoch": 0.314,
"grad_norm": 0.4411715567111969,
"learning_rate": 8.671020125860851e-06,
"loss": 0.1126,
"step": 1570
},
{
"epoch": 0.316,
"grad_norm": 0.3818731904029846,
"learning_rate": 8.647231768863513e-06,
"loss": 0.1184,
"step": 1580
},
{
"epoch": 0.318,
"grad_norm": 1.530585527420044,
"learning_rate": 8.623265650935233e-06,
"loss": 0.1197,
"step": 1590
},
{
"epoch": 0.32,
"grad_norm": 0.40408292412757874,
"learning_rate": 8.599122940150795e-06,
"loss": 0.1208,
"step": 1600
},
{
"epoch": 0.322,
"grad_norm": 0.6378718614578247,
"learning_rate": 8.574804813191859e-06,
"loss": 0.1164,
"step": 1610
},
{
"epoch": 0.324,
"grad_norm": 0.4105075001716614,
"learning_rate": 8.550312455289624e-06,
"loss": 0.1213,
"step": 1620
},
{
"epoch": 0.326,
"grad_norm": 0.4447656273841858,
"learning_rate": 8.525647060167063e-06,
"loss": 0.125,
"step": 1630
},
{
"epoch": 0.328,
"grad_norm": 0.4272136986255646,
"learning_rate": 8.500809829980734e-06,
"loss": 0.1152,
"step": 1640
},
{
"epoch": 0.33,
"grad_norm": 0.5936419367790222,
"learning_rate": 8.4758019752622e-06,
"loss": 0.1122,
"step": 1650
},
{
"epoch": 0.332,
"grad_norm": 0.5025408864021301,
"learning_rate": 8.450624714859016e-06,
"loss": 0.1212,
"step": 1660
},
{
"epoch": 0.334,
"grad_norm": 1.1935460567474365,
"learning_rate": 8.425279275875336e-06,
"loss": 0.1183,
"step": 1670
},
{
"epoch": 0.336,
"grad_norm": 0.33655646443367004,
"learning_rate": 8.399766893612096e-06,
"loss": 0.1172,
"step": 1680
},
{
"epoch": 0.338,
"grad_norm": 0.44501084089279175,
"learning_rate": 8.374088811506819e-06,
"loss": 0.1222,
"step": 1690
},
{
"epoch": 0.34,
"grad_norm": 0.42191627621650696,
"learning_rate": 8.348246281072998e-06,
"loss": 0.1175,
"step": 1700
},
{
"epoch": 0.342,
"grad_norm": 0.43164587020874023,
"learning_rate": 8.32224056183911e-06,
"loss": 0.1157,
"step": 1710
},
{
"epoch": 0.344,
"grad_norm": 0.5635483860969543,
"learning_rate": 8.296072921287217e-06,
"loss": 0.1199,
"step": 1720
},
{
"epoch": 0.346,
"grad_norm": 0.5875423550605774,
"learning_rate": 8.269744634791207e-06,
"loss": 0.1189,
"step": 1730
},
{
"epoch": 0.348,
"grad_norm": 0.6257967948913574,
"learning_rate": 8.243256985554622e-06,
"loss": 0.12,
"step": 1740
},
{
"epoch": 0.35,
"grad_norm": 0.5064549446105957,
"learning_rate": 8.21661126454811e-06,
"loss": 0.1205,
"step": 1750
},
{
"epoch": 0.352,
"grad_norm": 0.4206721782684326,
"learning_rate": 8.189808770446528e-06,
"loss": 0.1195,
"step": 1760
},
{
"epoch": 0.354,
"grad_norm": 1.636749029159546,
"learning_rate": 8.162850809565623e-06,
"loss": 0.1135,
"step": 1770
},
{
"epoch": 0.356,
"grad_norm": 0.6703956723213196,
"learning_rate": 8.135738695798377e-06,
"loss": 0.1216,
"step": 1780
},
{
"epoch": 0.358,
"grad_norm": 0.7827121615409851,
"learning_rate": 8.108473750550965e-06,
"loss": 0.123,
"step": 1790
},
{
"epoch": 0.36,
"grad_norm": 0.3569027781486511,
"learning_rate": 8.081057302678352e-06,
"loss": 0.1101,
"step": 1800
},
{
"epoch": 0.362,
"grad_norm": 0.36223459243774414,
"learning_rate": 8.053490688419532e-06,
"loss": 0.1222,
"step": 1810
},
{
"epoch": 0.364,
"grad_norm": 0.40967607498168945,
"learning_rate": 8.02577525133239e-06,
"loss": 0.123,
"step": 1820
},
{
"epoch": 0.366,
"grad_norm": 1.797508955001831,
"learning_rate": 7.997912342228232e-06,
"loss": 0.1146,
"step": 1830
},
{
"epoch": 0.368,
"grad_norm": 0.3894302546977997,
"learning_rate": 7.969903319105935e-06,
"loss": 0.117,
"step": 1840
},
{
"epoch": 0.37,
"grad_norm": 0.47636857628822327,
"learning_rate": 7.941749547085778e-06,
"loss": 0.1175,
"step": 1850
},
{
"epoch": 0.372,
"grad_norm": 5.96767520904541,
"learning_rate": 7.913452398342882e-06,
"loss": 0.1151,
"step": 1860
},
{
"epoch": 0.374,
"grad_norm": 0.8992507457733154,
"learning_rate": 7.88501325204036e-06,
"loss": 0.1208,
"step": 1870
},
{
"epoch": 0.376,
"grad_norm": 0.37870603799819946,
"learning_rate": 7.856433494262078e-06,
"loss": 0.119,
"step": 1880
},
{
"epoch": 0.378,
"grad_norm": 0.3793867826461792,
"learning_rate": 7.827714517945116e-06,
"loss": 0.1176,
"step": 1890
},
{
"epoch": 0.38,
"grad_norm": 0.5373750925064087,
"learning_rate": 7.798857722811857e-06,
"loss": 0.1228,
"step": 1900
},
{
"epoch": 0.382,
"grad_norm": 0.5349807143211365,
"learning_rate": 7.769864515301787e-06,
"loss": 0.1245,
"step": 1910
},
{
"epoch": 0.384,
"grad_norm": 0.5244778394699097,
"learning_rate": 7.740736308502939e-06,
"loss": 0.1206,
"step": 1920
},
{
"epoch": 0.386,
"grad_norm": 0.9785746932029724,
"learning_rate": 7.711474522083015e-06,
"loss": 0.1189,
"step": 1930
},
{
"epoch": 0.388,
"grad_norm": 0.7063882946968079,
"learning_rate": 7.682080582220206e-06,
"loss": 0.1182,
"step": 1940
},
{
"epoch": 0.39,
"grad_norm": 3.1905932426452637,
"learning_rate": 7.652555921533671e-06,
"loss": 0.114,
"step": 1950
},
{
"epoch": 0.392,
"grad_norm": 0.5251575112342834,
"learning_rate": 7.622901979013717e-06,
"loss": 0.1167,
"step": 1960
},
{
"epoch": 0.394,
"grad_norm": 0.6657569408416748,
"learning_rate": 7.5931201999516715e-06,
"loss": 0.1182,
"step": 1970
},
{
"epoch": 0.396,
"grad_norm": 0.3866606056690216,
"learning_rate": 7.563212035869426e-06,
"loss": 0.1242,
"step": 1980
},
{
"epoch": 0.398,
"grad_norm": 0.4084101915359497,
"learning_rate": 7.533178944448705e-06,
"loss": 0.1165,
"step": 1990
},
{
"epoch": 0.4,
"grad_norm": 0.4045194983482361,
"learning_rate": 7.503022389460014e-06,
"loss": 0.1224,
"step": 2000
},
{
"epoch": 0.402,
"grad_norm": 0.6457034945487976,
"learning_rate": 7.4727438406912986e-06,
"loss": 0.1187,
"step": 2010
},
{
"epoch": 0.404,
"grad_norm": 0.5980866551399231,
"learning_rate": 7.44234477387631e-06,
"loss": 0.1179,
"step": 2020
},
{
"epoch": 0.406,
"grad_norm": 0.3560180962085724,
"learning_rate": 7.411826670622676e-06,
"loss": 0.1134,
"step": 2030
},
{
"epoch": 0.408,
"grad_norm": 0.4002629816532135,
"learning_rate": 7.381191018339697e-06,
"loss": 0.1175,
"step": 2040
},
{
"epoch": 0.41,
"grad_norm": 0.36518630385398865,
"learning_rate": 7.350439310165842e-06,
"loss": 0.1102,
"step": 2050
},
{
"epoch": 0.412,
"grad_norm": 1.9818588495254517,
"learning_rate": 7.319573044895986e-06,
"loss": 0.1151,
"step": 2060
},
{
"epoch": 0.414,
"grad_norm": 0.4517176151275635,
"learning_rate": 7.288593726908351e-06,
"loss": 0.1207,
"step": 2070
},
{
"epoch": 0.416,
"grad_norm": 0.40875038504600525,
"learning_rate": 7.257502866091192e-06,
"loss": 0.113,
"step": 2080
},
{
"epoch": 0.418,
"grad_norm": 0.6046447157859802,
"learning_rate": 7.226301977769199e-06,
"loss": 0.1169,
"step": 2090
},
{
"epoch": 0.42,
"grad_norm": 0.6228557825088501,
"learning_rate": 7.194992582629654e-06,
"loss": 0.1151,
"step": 2100
},
{
"epoch": 0.422,
"grad_norm": 0.4920182526111603,
"learning_rate": 7.1635762066483035e-06,
"loss": 0.1202,
"step": 2110
},
{
"epoch": 0.424,
"grad_norm": 0.3749644458293915,
"learning_rate": 7.1320543810149945e-06,
"loss": 0.1082,
"step": 2120
},
{
"epoch": 0.426,
"grad_norm": 0.5646130442619324,
"learning_rate": 7.100428642059033e-06,
"loss": 0.1173,
"step": 2130
},
{
"epoch": 0.428,
"grad_norm": 0.6624546051025391,
"learning_rate": 7.0687005311743195e-06,
"loss": 0.1159,
"step": 2140
},
{
"epoch": 0.43,
"grad_norm": 1.4650940895080566,
"learning_rate": 7.036871594744218e-06,
"loss": 0.1237,
"step": 2150
},
{
"epoch": 0.432,
"grad_norm": 0.35690367221832275,
"learning_rate": 7.0049433840661875e-06,
"loss": 0.1158,
"step": 2160
},
{
"epoch": 0.434,
"grad_norm": 0.4230184555053711,
"learning_rate": 6.97291745527617e-06,
"loss": 0.1148,
"step": 2170
},
{
"epoch": 0.436,
"grad_norm": 0.44549325108528137,
"learning_rate": 6.940795369272754e-06,
"loss": 0.1237,
"step": 2180
},
{
"epoch": 0.438,
"grad_norm": 0.29287219047546387,
"learning_rate": 6.908578691641092e-06,
"loss": 0.1134,
"step": 2190
},
{
"epoch": 0.44,
"grad_norm": 0.41562554240226746,
"learning_rate": 6.876268992576605e-06,
"loss": 0.1128,
"step": 2200
},
{
"epoch": 0.442,
"grad_norm": 0.35246309638023376,
"learning_rate": 6.843867846808438e-06,
"loss": 0.1141,
"step": 2210
},
{
"epoch": 0.444,
"grad_norm": 2.075042247772217,
"learning_rate": 6.811376833522729e-06,
"loss": 0.1147,
"step": 2220
},
{
"epoch": 0.446,
"grad_norm": 1.0448739528656006,
"learning_rate": 6.778797536285625e-06,
"loss": 0.1135,
"step": 2230
},
{
"epoch": 0.448,
"grad_norm": 0.6575298309326172,
"learning_rate": 6.746131542966112e-06,
"loss": 0.1169,
"step": 2240
},
{
"epoch": 0.45,
"grad_norm": 1.2195367813110352,
"learning_rate": 6.713380445658618e-06,
"loss": 0.1133,
"step": 2250
},
{
"epoch": 0.452,
"grad_norm": 1.0849454402923584,
"learning_rate": 6.680545840605423e-06,
"loss": 0.1159,
"step": 2260
},
{
"epoch": 0.454,
"grad_norm": 0.557558000087738,
"learning_rate": 6.647629328118852e-06,
"loss": 0.1189,
"step": 2270
},
{
"epoch": 0.456,
"grad_norm": 0.4906410276889801,
"learning_rate": 6.614632512503289e-06,
"loss": 0.1118,
"step": 2280
},
{
"epoch": 0.458,
"grad_norm": 0.5431510210037231,
"learning_rate": 6.58155700197697e-06,
"loss": 0.1154,
"step": 2290
},
{
"epoch": 0.46,
"grad_norm": 0.4932880401611328,
"learning_rate": 6.548404408593622e-06,
"loss": 0.1154,
"step": 2300
},
{
"epoch": 0.462,
"grad_norm": 0.9299081563949585,
"learning_rate": 6.5151763481638705e-06,
"loss": 0.1149,
"step": 2310
},
{
"epoch": 0.464,
"grad_norm": 0.39536532759666443,
"learning_rate": 6.481874440176506e-06,
"loss": 0.1208,
"step": 2320
},
{
"epoch": 0.466,
"grad_norm": 0.3463858366012573,
"learning_rate": 6.448500307719537e-06,
"loss": 0.1139,
"step": 2330
},
{
"epoch": 0.468,
"grad_norm": 0.3795383870601654,
"learning_rate": 6.415055577401101e-06,
"loss": 0.1119,
"step": 2340
},
{
"epoch": 0.47,
"grad_norm": 0.30274516344070435,
"learning_rate": 6.3815418792701686e-06,
"loss": 0.1103,
"step": 2350
},
{
"epoch": 0.472,
"grad_norm": 0.7076600193977356,
"learning_rate": 6.3479608467371055e-06,
"loss": 0.1137,
"step": 2360
},
{
"epoch": 0.474,
"grad_norm": 0.6139280796051025,
"learning_rate": 6.314314116494061e-06,
"loss": 0.1139,
"step": 2370
},
{
"epoch": 0.476,
"grad_norm": 1.3072528839111328,
"learning_rate": 6.280603328435199e-06,
"loss": 0.1074,
"step": 2380
},
{
"epoch": 0.478,
"grad_norm": 0.7616708874702454,
"learning_rate": 6.24683012557677e-06,
"loss": 0.1164,
"step": 2390
},
{
"epoch": 0.48,
"grad_norm": 0.5742841958999634,
"learning_rate": 6.212996153977038e-06,
"loss": 0.12,
"step": 2400
},
{
"epoch": 0.482,
"grad_norm": 0.3245879113674164,
"learning_rate": 6.179103062656042e-06,
"loss": 0.1081,
"step": 2410
},
{
"epoch": 0.484,
"grad_norm": 0.46111881732940674,
"learning_rate": 6.145152503515239e-06,
"loss": 0.1158,
"step": 2420
},
{
"epoch": 0.486,
"grad_norm": 0.39089033007621765,
"learning_rate": 6.111146131256983e-06,
"loss": 0.1238,
"step": 2430
},
{
"epoch": 0.488,
"grad_norm": 0.4809490144252777,
"learning_rate": 6.077085603303883e-06,
"loss": 0.1101,
"step": 2440
},
{
"epoch": 0.49,
"grad_norm": 0.3588216006755829,
"learning_rate": 6.04297257971802e-06,
"loss": 0.1118,
"step": 2450
},
{
"epoch": 0.492,
"grad_norm": 0.2877720892429352,
"learning_rate": 6.008808723120035e-06,
"loss": 0.1158,
"step": 2460
},
{
"epoch": 0.494,
"grad_norm": 0.7675597667694092,
"learning_rate": 5.974595698608103e-06,
"loss": 0.1119,
"step": 2470
},
{
"epoch": 0.496,
"grad_norm": 0.9885912537574768,
"learning_rate": 5.94033517367677e-06,
"loss": 0.1119,
"step": 2480
},
{
"epoch": 0.498,
"grad_norm": 0.6791333556175232,
"learning_rate": 5.906028818135687e-06,
"loss": 0.1165,
"step": 2490
},
{
"epoch": 0.5,
"grad_norm": 0.42410117387771606,
"learning_rate": 5.871678304028224e-06,
"loss": 0.1027,
"step": 2500
},
{
"epoch": 0.502,
"grad_norm": 0.4329153001308441,
"learning_rate": 5.837285305549978e-06,
"loss": 0.1091,
"step": 2510
},
{
"epoch": 0.504,
"grad_norm": 0.6316914558410645,
"learning_rate": 5.802851498967173e-06,
"loss": 0.1144,
"step": 2520
},
{
"epoch": 0.506,
"grad_norm": 0.29134896397590637,
"learning_rate": 5.768378562534962e-06,
"loss": 0.1179,
"step": 2530
},
{
"epoch": 0.508,
"grad_norm": 0.4128161668777466,
"learning_rate": 5.733868176415633e-06,
"loss": 0.1125,
"step": 2540
},
{
"epoch": 0.51,
"grad_norm": 0.44271254539489746,
"learning_rate": 5.6993220225967214e-06,
"loss": 0.1081,
"step": 2550
},
{
"epoch": 0.512,
"grad_norm": 0.5136324763298035,
"learning_rate": 5.6647417848090225e-06,
"loss": 0.1115,
"step": 2560
},
{
"epoch": 0.514,
"grad_norm": 0.5882424712181091,
"learning_rate": 5.630129148444543e-06,
"loss": 0.1086,
"step": 2570
},
{
"epoch": 0.516,
"grad_norm": 0.4773464798927307,
"learning_rate": 5.59548580047435e-06,
"loss": 0.1077,
"step": 2580
},
{
"epoch": 0.518,
"grad_norm": 0.48995503783226013,
"learning_rate": 5.560813429366345e-06,
"loss": 0.1098,
"step": 2590
},
{
"epoch": 0.52,
"grad_norm": 0.9368364810943604,
"learning_rate": 5.526113725002984e-06,
"loss": 0.1173,
"step": 2600
},
{
"epoch": 0.522,
"grad_norm": 0.4998178780078888,
"learning_rate": 5.491388378598899e-06,
"loss": 0.1115,
"step": 2610
},
{
"epoch": 0.524,
"grad_norm": 0.3727714717388153,
"learning_rate": 5.456639082618489e-06,
"loss": 0.1047,
"step": 2620
},
{
"epoch": 0.526,
"grad_norm": 0.5195613503456116,
"learning_rate": 5.4218675306934145e-06,
"loss": 0.1087,
"step": 2630
},
{
"epoch": 0.528,
"grad_norm": 1.1314338445663452,
"learning_rate": 5.3870754175400595e-06,
"loss": 0.1127,
"step": 2640
},
{
"epoch": 0.53,
"grad_norm": 0.6690823435783386,
"learning_rate": 5.352264438876935e-06,
"loss": 0.1098,
"step": 2650
},
{
"epoch": 0.532,
"grad_norm": 0.3548797369003296,
"learning_rate": 5.317436291342031e-06,
"loss": 0.1099,
"step": 2660
},
{
"epoch": 0.534,
"grad_norm": 0.34155887365341187,
"learning_rate": 5.282592672410124e-06,
"loss": 0.1185,
"step": 2670
},
{
"epoch": 0.536,
"grad_norm": 0.3449115455150604,
"learning_rate": 5.247735280310041e-06,
"loss": 0.1075,
"step": 2680
},
{
"epoch": 0.538,
"grad_norm": 0.5513066053390503,
"learning_rate": 5.212865813941899e-06,
"loss": 0.1152,
"step": 2690
},
{
"epoch": 0.54,
"grad_norm": 0.496774286031723,
"learning_rate": 5.177985972794293e-06,
"loss": 0.1053,
"step": 2700
},
{
"epoch": 0.542,
"grad_norm": 0.47940221428871155,
"learning_rate": 5.143097456861474e-06,
"loss": 0.1096,
"step": 2710
},
{
"epoch": 0.544,
"grad_norm": 0.26659730076789856,
"learning_rate": 5.1082019665604895e-06,
"loss": 0.1097,
"step": 2720
},
{
"epoch": 0.546,
"grad_norm": 0.359784871339798,
"learning_rate": 5.073301202648304e-06,
"loss": 0.1004,
"step": 2730
},
{
"epoch": 0.548,
"grad_norm": 1.3557486534118652,
"learning_rate": 5.038396866138915e-06,
"loss": 0.118,
"step": 2740
},
{
"epoch": 0.55,
"grad_norm": 0.3325219750404358,
"learning_rate": 5.003490658220438e-06,
"loss": 0.1116,
"step": 2750
},
{
"epoch": 0.552,
"grad_norm": 0.5154880881309509,
"learning_rate": 4.968584280172206e-06,
"loss": 0.1022,
"step": 2760
},
{
"epoch": 0.554,
"grad_norm": 1.5963865518569946,
"learning_rate": 4.933679433281837e-06,
"loss": 0.1166,
"step": 2770
},
{
"epoch": 0.556,
"grad_norm": 1.1274834871292114,
"learning_rate": 4.898777818762325e-06,
"loss": 0.1136,
"step": 2780
},
{
"epoch": 0.558,
"grad_norm": 0.4707079529762268,
"learning_rate": 4.863881137669123e-06,
"loss": 0.1094,
"step": 2790
},
{
"epoch": 0.56,
"grad_norm": 0.31004512310028076,
"learning_rate": 4.828991090817238e-06,
"loss": 0.1086,
"step": 2800
},
{
"epoch": 0.562,
"grad_norm": 0.5906158685684204,
"learning_rate": 4.794109378698327e-06,
"loss": 0.121,
"step": 2810
},
{
"epoch": 0.564,
"grad_norm": 0.44631025195121765,
"learning_rate": 4.759237701397831e-06,
"loss": 0.1148,
"step": 2820
},
{
"epoch": 0.566,
"grad_norm": 1.058272361755371,
"learning_rate": 4.7243777585121034e-06,
"loss": 0.1113,
"step": 2830
},
{
"epoch": 0.568,
"grad_norm": 0.7068793773651123,
"learning_rate": 4.689531249065581e-06,
"loss": 0.1066,
"step": 2840
},
{
"epoch": 0.57,
"grad_norm": 0.47186627984046936,
"learning_rate": 4.654699871427972e-06,
"loss": 0.1177,
"step": 2850
},
{
"epoch": 0.572,
"grad_norm": 0.45073261857032776,
"learning_rate": 4.619885323231484e-06,
"loss": 0.1134,
"step": 2860
},
{
"epoch": 0.574,
"grad_norm": 0.5080240964889526,
"learning_rate": 4.5850893012880806e-06,
"loss": 0.1183,
"step": 2870
},
{
"epoch": 0.576,
"grad_norm": 0.48024675250053406,
"learning_rate": 4.5503135015067815e-06,
"loss": 0.1132,
"step": 2880
},
{
"epoch": 0.578,
"grad_norm": 0.4118949770927429,
"learning_rate": 4.5155596188110055e-06,
"loss": 0.1094,
"step": 2890
},
{
"epoch": 0.58,
"grad_norm": 0.4416310489177704,
"learning_rate": 4.4808293470559645e-06,
"loss": 0.1178,
"step": 2900
},
{
"epoch": 0.582,
"grad_norm": 0.3114508092403412,
"learning_rate": 4.446124378946108e-06,
"loss": 0.1216,
"step": 2910
},
{
"epoch": 0.584,
"grad_norm": 0.48362258076667786,
"learning_rate": 4.4114464059526185e-06,
"loss": 0.0995,
"step": 2920
},
{
"epoch": 0.586,
"grad_norm": 0.34811100363731384,
"learning_rate": 4.376797118230978e-06,
"loss": 0.1089,
"step": 2930
},
{
"epoch": 0.588,
"grad_norm": 0.3432498276233673,
"learning_rate": 4.342178204538588e-06,
"loss": 0.1141,
"step": 2940
},
{
"epoch": 0.59,
"grad_norm": 0.7286999225616455,
"learning_rate": 4.307591352152459e-06,
"loss": 0.1142,
"step": 2950
},
{
"epoch": 0.592,
"grad_norm": 0.3824271261692047,
"learning_rate": 4.273038246786986e-06,
"loss": 0.1146,
"step": 2960
},
{
"epoch": 0.594,
"grad_norm": 0.5760470032691956,
"learning_rate": 4.238520572511773e-06,
"loss": 0.1125,
"step": 2970
},
{
"epoch": 0.596,
"grad_norm": 0.29192543029785156,
"learning_rate": 4.204040011669567e-06,
"loss": 0.1143,
"step": 2980
},
{
"epoch": 0.598,
"grad_norm": 0.4159230887889862,
"learning_rate": 4.169598244794261e-06,
"loss": 0.109,
"step": 2990
},
{
"epoch": 0.6,
"grad_norm": 0.27462849020957947,
"learning_rate": 4.135196950528982e-06,
"loss": 0.1157,
"step": 3000
},
{
"epoch": 0.602,
"grad_norm": 0.4015858471393585,
"learning_rate": 4.100837805544279e-06,
"loss": 0.1019,
"step": 3010
},
{
"epoch": 0.604,
"grad_norm": 0.6288099884986877,
"learning_rate": 4.066522484456406e-06,
"loss": 0.1119,
"step": 3020
},
{
"epoch": 0.606,
"grad_norm": 0.7088296413421631,
"learning_rate": 4.032252659745699e-06,
"loss": 0.1119,
"step": 3030
},
{
"epoch": 0.608,
"grad_norm": 2.2175350189208984,
"learning_rate": 3.9980300016750696e-06,
"loss": 0.1135,
"step": 3040
},
{
"epoch": 0.61,
"grad_norm": 0.2995910346508026,
"learning_rate": 3.963856178208588e-06,
"loss": 0.1068,
"step": 3050
},
{
"epoch": 0.612,
"grad_norm": 0.4773070514202118,
"learning_rate": 3.9297328549302e-06,
"loss": 0.1118,
"step": 3060
},
{
"epoch": 0.614,
"grad_norm": 0.3807465732097626,
"learning_rate": 3.895661694962542e-06,
"loss": 0.1115,
"step": 3070
},
{
"epoch": 0.616,
"grad_norm": 0.5447002649307251,
"learning_rate": 3.86164435888588e-06,
"loss": 0.1148,
"step": 3080
},
{
"epoch": 0.618,
"grad_norm": 0.6512142419815063,
"learning_rate": 3.827682504657187e-06,
"loss": 0.1143,
"step": 3090
},
{
"epoch": 0.62,
"grad_norm": 0.4874299168586731,
"learning_rate": 3.793777787529325e-06,
"loss": 0.1151,
"step": 3100
},
{
"epoch": 0.622,
"grad_norm": 0.4157492220401764,
"learning_rate": 3.759931859970374e-06,
"loss": 0.1108,
"step": 3110
},
{
"epoch": 0.624,
"grad_norm": 0.4890842139720917,
"learning_rate": 3.7261463715830902e-06,
"loss": 0.1114,
"step": 3120
},
{
"epoch": 0.626,
"grad_norm": 0.3871977627277374,
"learning_rate": 3.6924229690245163e-06,
"loss": 0.1112,
"step": 3130
},
{
"epoch": 0.628,
"grad_norm": 0.3650118410587311,
"learning_rate": 3.6587632959257168e-06,
"loss": 0.1051,
"step": 3140
},
{
"epoch": 0.63,
"grad_norm": 0.2773863971233368,
"learning_rate": 3.625168992811671e-06,
"loss": 0.1125,
"step": 3150
},
{
"epoch": 0.632,
"grad_norm": 0.3802359402179718,
"learning_rate": 3.5916416970213173e-06,
"loss": 0.1073,
"step": 3160
},
{
"epoch": 0.634,
"grad_norm": 0.5529680252075195,
"learning_rate": 3.5581830426277554e-06,
"loss": 0.1132,
"step": 3170
},
{
"epoch": 0.636,
"grad_norm": 0.33904126286506653,
"learning_rate": 3.524794660358593e-06,
"loss": 0.1107,
"step": 3180
},
{
"epoch": 0.638,
"grad_norm": 0.7427698373794556,
"learning_rate": 3.491478177516484e-06,
"loss": 0.1086,
"step": 3190
},
{
"epoch": 0.64,
"grad_norm": 0.3335733711719513,
"learning_rate": 3.4582352178997937e-06,
"loss": 0.1104,
"step": 3200
},
{
"epoch": 0.642,
"grad_norm": 0.4681756794452667,
"learning_rate": 3.4250674017234774e-06,
"loss": 0.117,
"step": 3210
},
{
"epoch": 0.644,
"grad_norm": 0.7870750427246094,
"learning_rate": 3.3919763455401016e-06,
"loss": 0.1141,
"step": 3220
},
{
"epoch": 0.646,
"grad_norm": 0.4855883717536926,
"learning_rate": 3.358963662161062e-06,
"loss": 0.1049,
"step": 3230
},
{
"epoch": 0.648,
"grad_norm": 1.0050535202026367,
"learning_rate": 3.3260309605779717e-06,
"loss": 0.1097,
"step": 3240
},
{
"epoch": 0.65,
"grad_norm": 0.3933185935020447,
"learning_rate": 3.293179845884245e-06,
"loss": 0.1152,
"step": 3250
},
{
"epoch": 0.652,
"grad_norm": 0.38565793633461,
"learning_rate": 3.260411919196866e-06,
"loss": 0.1094,
"step": 3260
},
{
"epoch": 0.654,
"grad_norm": 0.4234326481819153,
"learning_rate": 3.227728777578353e-06,
"loss": 0.1168,
"step": 3270
},
{
"epoch": 0.656,
"grad_norm": 0.5142901539802551,
"learning_rate": 3.195132013958918e-06,
"loss": 0.1114,
"step": 3280
},
{
"epoch": 0.658,
"grad_norm": 0.3599749505519867,
"learning_rate": 3.1626232170588343e-06,
"loss": 0.1078,
"step": 3290
},
{
"epoch": 0.66,
"grad_norm": 0.3898942172527313,
"learning_rate": 3.130203971310999e-06,
"loss": 0.113,
"step": 3300
},
{
"epoch": 0.662,
"grad_norm": 1.5486271381378174,
"learning_rate": 3.097875856783713e-06,
"loss": 0.109,
"step": 3310
},
{
"epoch": 0.664,
"grad_norm": 1.6251258850097656,
"learning_rate": 3.0656404491036696e-06,
"loss": 0.116,
"step": 3320
},
{
"epoch": 0.666,
"grad_norm": 3.3787240982055664,
"learning_rate": 3.033499319379163e-06,
"loss": 0.1083,
"step": 3330
},
{
"epoch": 0.668,
"grad_norm": 0.4697399437427521,
"learning_rate": 3.001454034123512e-06,
"loss": 0.117,
"step": 3340
},
{
"epoch": 0.67,
"grad_norm": 0.35250768065452576,
"learning_rate": 2.969506155178711e-06,
"loss": 0.1091,
"step": 3350
},
{
"epoch": 0.672,
"grad_norm": 0.7035725116729736,
"learning_rate": 2.9376572396393047e-06,
"loss": 0.111,
"step": 3360
},
{
"epoch": 0.674,
"grad_norm": 0.4109930992126465,
"learning_rate": 2.905908839776509e-06,
"loss": 0.1106,
"step": 3370
},
{
"epoch": 0.676,
"grad_norm": 0.43214330077171326,
"learning_rate": 2.874262502962537e-06,
"loss": 0.1072,
"step": 3380
},
{
"epoch": 0.678,
"grad_norm": 0.3473849296569824,
"learning_rate": 2.8427197715952047e-06,
"loss": 0.1158,
"step": 3390
},
{
"epoch": 0.68,
"grad_norm": 0.40624290704727173,
"learning_rate": 2.811282183022736e-06,
"loss": 0.1139,
"step": 3400
},
{
"epoch": 0.682,
"grad_norm": 3.3285329341888428,
"learning_rate": 2.779951269468847e-06,
"loss": 0.1162,
"step": 3410
},
{
"epoch": 0.684,
"grad_norm": 0.624461829662323,
"learning_rate": 2.7487285579580635e-06,
"loss": 0.1084,
"step": 3420
},
{
"epoch": 0.686,
"grad_norm": 1.3213962316513062,
"learning_rate": 2.717615570241294e-06,
"loss": 0.1098,
"step": 3430
},
{
"epoch": 0.688,
"grad_norm": 0.3557455539703369,
"learning_rate": 2.686613822721666e-06,
"loss": 0.113,
"step": 3440
},
{
"epoch": 0.69,
"grad_norm": 0.4707227051258087,
"learning_rate": 2.6557248263806175e-06,
"loss": 0.1135,
"step": 3450
},
{
"epoch": 0.692,
"grad_norm": 0.5622274875640869,
"learning_rate": 2.6249500867042523e-06,
"loss": 0.1224,
"step": 3460
},
{
"epoch": 0.694,
"grad_norm": 0.43661999702453613,
"learning_rate": 2.5942911036099657e-06,
"loss": 0.1178,
"step": 3470
},
{
"epoch": 0.696,
"grad_norm": 0.8577882051467896,
"learning_rate": 2.5637493713733376e-06,
"loss": 0.1101,
"step": 3480
},
{
"epoch": 0.698,
"grad_norm": 0.5041825771331787,
"learning_rate": 2.533326378555314e-06,
"loss": 0.1115,
"step": 3490
},
{
"epoch": 0.7,
"grad_norm": 0.3203236758708954,
"learning_rate": 2.5030236079296443e-06,
"loss": 0.1093,
"step": 3500
},
{
"epoch": 0.702,
"grad_norm": 0.37466859817504883,
"learning_rate": 2.4728425364106136e-06,
"loss": 0.1063,
"step": 3510
},
{
"epoch": 0.704,
"grad_norm": 0.7311847805976868,
"learning_rate": 2.442784634981071e-06,
"loss": 0.1092,
"step": 3520
},
{
"epoch": 0.706,
"grad_norm": 0.7368522882461548,
"learning_rate": 2.412851368620726e-06,
"loss": 0.1094,
"step": 3530
},
{
"epoch": 0.708,
"grad_norm": 1.2340730428695679,
"learning_rate": 2.3830441962347528e-06,
"loss": 0.1141,
"step": 3540
},
{
"epoch": 0.71,
"grad_norm": 0.4553486406803131,
"learning_rate": 2.353364570582681e-06,
"loss": 0.1092,
"step": 3550
},
{
"epoch": 0.712,
"grad_norm": 0.49003028869628906,
"learning_rate": 2.323813938207593e-06,
"loss": 0.1071,
"step": 3560
},
{
"epoch": 0.714,
"grad_norm": 0.8033204674720764,
"learning_rate": 2.294393739365621e-06,
"loss": 0.1159,
"step": 3570
},
{
"epoch": 0.716,
"grad_norm": 0.3968663513660431,
"learning_rate": 2.265105407955752e-06,
"loss": 0.1162,
"step": 3580
},
{
"epoch": 0.718,
"grad_norm": 0.567513108253479,
"learning_rate": 2.235950371449938e-06,
"loss": 0.1125,
"step": 3590
},
{
"epoch": 0.72,
"grad_norm": 0.4921203851699829,
"learning_rate": 2.2069300508235273e-06,
"loss": 0.1131,
"step": 3600
},
{
"epoch": 0.722,
"grad_norm": 2.202014684677124,
"learning_rate": 2.1780458604860056e-06,
"loss": 0.1093,
"step": 3610
},
{
"epoch": 0.724,
"grad_norm": 0.5519041419029236,
"learning_rate": 2.14929920821206e-06,
"loss": 0.1158,
"step": 3620
},
{
"epoch": 0.726,
"grad_norm": 0.41222497820854187,
"learning_rate": 2.1206914950729673e-06,
"loss": 0.1095,
"step": 3630
},
{
"epoch": 0.728,
"grad_norm": 1.8621835708618164,
"learning_rate": 2.0922241153683064e-06,
"loss": 0.1148,
"step": 3640
},
{
"epoch": 0.73,
"grad_norm": 0.47333869338035583,
"learning_rate": 2.063898456558002e-06,
"loss": 0.1089,
"step": 3650
},
{
"epoch": 0.732,
"grad_norm": 0.37637799978256226,
"learning_rate": 2.035715899194704e-06,
"loss": 0.106,
"step": 3660
},
{
"epoch": 0.734,
"grad_norm": 1.6549164056777954,
"learning_rate": 2.007677816856498e-06,
"loss": 0.104,
"step": 3670
},
{
"epoch": 0.736,
"grad_norm": 0.3913106918334961,
"learning_rate": 1.979785576079961e-06,
"loss": 0.1041,
"step": 3680
},
{
"epoch": 0.738,
"grad_norm": 0.498811811208725,
"learning_rate": 1.95204053629356e-06,
"loss": 0.1137,
"step": 3690
},
{
"epoch": 0.74,
"grad_norm": 0.3852579593658447,
"learning_rate": 1.9244440497513895e-06,
"loss": 0.1018,
"step": 3700
},
{
"epoch": 0.742,
"grad_norm": 0.5074270963668823,
"learning_rate": 1.896997461467272e-06,
"loss": 0.1101,
"step": 3710
},
{
"epoch": 0.744,
"grad_norm": 0.6303516626358032,
"learning_rate": 1.8697021091491991e-06,
"loss": 0.1088,
"step": 3720
},
{
"epoch": 0.746,
"grad_norm": 0.8138649463653564,
"learning_rate": 1.842559323134136e-06,
"loss": 0.1072,
"step": 3730
},
{
"epoch": 0.748,
"grad_norm": 0.4117524027824402,
"learning_rate": 1.8155704263231777e-06,
"loss": 0.1024,
"step": 3740
},
{
"epoch": 0.75,
"grad_norm": 1.3261663913726807,
"learning_rate": 1.7887367341170781e-06,
"loss": 0.1093,
"step": 3750
},
{
"epoch": 0.752,
"grad_norm": 0.4887453317642212,
"learning_rate": 1.762059554352143e-06,
"loss": 0.1145,
"step": 3760
},
{
"epoch": 0.754,
"grad_norm": 0.7797528505325317,
"learning_rate": 1.7355401872364759e-06,
"loss": 0.1124,
"step": 3770
},
{
"epoch": 0.756,
"grad_norm": 0.5533345937728882,
"learning_rate": 1.709179925286617e-06,
"loss": 0.1132,
"step": 3780
},
{
"epoch": 0.758,
"grad_norm": 0.8638681173324585,
"learning_rate": 1.6829800532645447e-06,
"loss": 0.1168,
"step": 3790
},
{
"epoch": 0.76,
"grad_norm": 0.38189423084259033,
"learning_rate": 1.6569418481150596e-06,
"loss": 0.1083,
"step": 3800
},
{
"epoch": 0.762,
"grad_norm": 0.5026066303253174,
"learning_rate": 1.6310665789035468e-06,
"loss": 0.109,
"step": 3810
},
{
"epoch": 0.764,
"grad_norm": 0.5332008004188538,
"learning_rate": 1.605355506754121e-06,
"loss": 0.1096,
"step": 3820
},
{
"epoch": 0.766,
"grad_norm": 0.5544666051864624,
"learning_rate": 1.5798098847881664e-06,
"loss": 0.1038,
"step": 3830
},
{
"epoch": 0.768,
"grad_norm": 0.5581374764442444,
"learning_rate": 1.554430958063259e-06,
"loss": 0.1069,
"step": 3840
},
{
"epoch": 0.77,
"grad_norm": 0.4732300341129303,
"learning_rate": 1.529219963512481e-06,
"loss": 0.1097,
"step": 3850
},
{
"epoch": 0.772,
"grad_norm": 0.46628376841545105,
"learning_rate": 1.5041781298841424e-06,
"loss": 0.1193,
"step": 3860
},
{
"epoch": 0.774,
"grad_norm": 0.4595516622066498,
"learning_rate": 1.4793066776818843e-06,
"loss": 0.1128,
"step": 3870
},
{
"epoch": 0.776,
"grad_norm": 0.9645623564720154,
"learning_rate": 1.4546068191051988e-06,
"loss": 0.1119,
"step": 3880
},
{
"epoch": 0.778,
"grad_norm": 0.5310599207878113,
"learning_rate": 1.4300797579903476e-06,
"loss": 0.1086,
"step": 3890
},
{
"epoch": 0.78,
"grad_norm": 0.710594892501831,
"learning_rate": 1.4057266897516842e-06,
"loss": 0.1183,
"step": 3900
},
{
"epoch": 0.782,
"grad_norm": 0.37046968936920166,
"learning_rate": 1.3815488013233986e-06,
"loss": 0.1056,
"step": 3910
},
{
"epoch": 0.784,
"grad_norm": 0.3346898555755615,
"learning_rate": 1.3575472711016634e-06,
"loss": 0.106,
"step": 3920
},
{
"epoch": 0.786,
"grad_norm": 0.34566453099250793,
"learning_rate": 1.333723268887201e-06,
"loss": 0.104,
"step": 3930
},
{
"epoch": 0.788,
"grad_norm": 0.696998655796051,
"learning_rate": 1.3100779558282673e-06,
"loss": 0.1046,
"step": 3940
},
{
"epoch": 0.79,
"grad_norm": 0.35373494029045105,
"learning_rate": 1.2866124843640614e-06,
"loss": 0.1091,
"step": 3950
},
{
"epoch": 0.792,
"grad_norm": 0.5243257284164429,
"learning_rate": 1.2633279981685608e-06,
"loss": 0.1152,
"step": 3960
},
{
"epoch": 0.794,
"grad_norm": 0.4453822672367096,
"learning_rate": 1.240225632094773e-06,
"loss": 0.1098,
"step": 3970
},
{
"epoch": 0.796,
"grad_norm": 0.566088080406189,
"learning_rate": 1.217306512119425e-06,
"loss": 0.1168,
"step": 3980
},
{
"epoch": 0.798,
"grad_norm": 0.7452012896537781,
"learning_rate": 1.1945717552880919e-06,
"loss": 0.1085,
"step": 3990
},
{
"epoch": 0.8,
"grad_norm": 0.4251628816127777,
"learning_rate": 1.1720224696607474e-06,
"loss": 0.1076,
"step": 4000
},
{
"epoch": 0.802,
"grad_norm": 0.6231107711791992,
"learning_rate": 1.1496597542577603e-06,
"loss": 0.1077,
"step": 4010
},
{
"epoch": 0.804,
"grad_norm": 0.6129289865493774,
"learning_rate": 1.1274846990063314e-06,
"loss": 0.1075,
"step": 4020
},
{
"epoch": 0.806,
"grad_norm": 0.35106995701789856,
"learning_rate": 1.1054983846873684e-06,
"loss": 0.105,
"step": 4030
},
{
"epoch": 0.808,
"grad_norm": 0.4739697575569153,
"learning_rate": 1.0837018828828133e-06,
"loss": 0.1058,
"step": 4040
},
{
"epoch": 0.81,
"grad_norm": 0.47443220019340515,
"learning_rate": 1.0620962559234144e-06,
"loss": 0.1067,
"step": 4050
},
{
"epoch": 0.812,
"grad_norm": 0.47327733039855957,
"learning_rate": 1.0406825568369478e-06,
"loss": 0.1097,
"step": 4060
},
{
"epoch": 0.814,
"grad_norm": 0.5550324320793152,
"learning_rate": 1.0194618292968972e-06,
"loss": 0.1033,
"step": 4070
},
{
"epoch": 0.816,
"grad_norm": 0.8000510334968567,
"learning_rate": 9.984351075715848e-07,
"loss": 0.1049,
"step": 4080
},
{
"epoch": 0.818,
"grad_norm": 0.4578244686126709,
"learning_rate": 9.77603416473763e-07,
"loss": 0.1107,
"step": 4090
},
{
"epoch": 0.82,
"grad_norm": 0.5600775480270386,
"learning_rate": 9.569677713106673e-07,
"loss": 0.1089,
"step": 4100
},
{
"epoch": 0.822,
"grad_norm": 0.46126943826675415,
"learning_rate": 9.365291778345303e-07,
"loss": 0.099,
"step": 4110
},
{
"epoch": 0.824,
"grad_norm": 0.555090069770813,
"learning_rate": 9.162886321935632e-07,
"loss": 0.1035,
"step": 4120
},
{
"epoch": 0.826,
"grad_norm": 0.4217548668384552,
"learning_rate": 8.962471208834056e-07,
"loss": 0.1059,
"step": 4130
},
{
"epoch": 0.828,
"grad_norm": 0.7031282186508179,
"learning_rate": 8.764056206990446e-07,
"loss": 0.1117,
"step": 4140
},
{
"epoch": 0.83,
"grad_norm": 0.6522052884101868,
"learning_rate": 8.567650986872061e-07,
"loss": 0.1049,
"step": 4150
},
{
"epoch": 0.832,
"grad_norm": 0.5371230840682983,
"learning_rate": 8.373265120992252e-07,
"loss": 0.1157,
"step": 4160
},
{
"epoch": 0.834,
"grad_norm": 0.4255172312259674,
"learning_rate": 8.180908083443884e-07,
"loss": 0.1052,
"step": 4170
},
{
"epoch": 0.836,
"grad_norm": 0.4094915986061096,
"learning_rate": 7.990589249437591e-07,
"loss": 0.1083,
"step": 4180
},
{
"epoch": 0.838,
"grad_norm": 0.32630613446235657,
"learning_rate": 7.802317894844835e-07,
"loss": 0.1048,
"step": 4190
},
{
"epoch": 0.84,
"grad_norm": 0.35342568159103394,
"learning_rate": 7.61610319574585e-07,
"loss": 0.107,
"step": 4200
},
{
"epoch": 0.842,
"grad_norm": 0.4322859048843384,
"learning_rate": 7.43195422798233e-07,
"loss": 0.0994,
"step": 4210
},
{
"epoch": 0.844,
"grad_norm": 0.41975167393684387,
"learning_rate": 7.249879966715174e-07,
"loss": 0.1166,
"step": 4220
},
{
"epoch": 0.846,
"grad_norm": 0.9153004884719849,
"learning_rate": 7.069889285987025e-07,
"loss": 0.1156,
"step": 4230
},
{
"epoch": 0.848,
"grad_norm": 2.4857826232910156,
"learning_rate": 6.891990958289724e-07,
"loss": 0.1029,
"step": 4240
},
{
"epoch": 0.85,
"grad_norm": 0.6253427267074585,
"learning_rate": 6.716193654136788e-07,
"loss": 0.1108,
"step": 4250
},
{
"epoch": 0.852,
"grad_norm": 0.9512668251991272,
"learning_rate": 6.542505941640803e-07,
"loss": 0.1084,
"step": 4260
},
{
"epoch": 0.854,
"grad_norm": 0.46003222465515137,
"learning_rate": 6.370936286095842e-07,
"loss": 0.1046,
"step": 4270
},
{
"epoch": 0.856,
"grad_norm": 0.37549811601638794,
"learning_rate": 6.201493049564883e-07,
"loss": 0.1058,
"step": 4280
},
{
"epoch": 0.858,
"grad_norm": 0.554378867149353,
"learning_rate": 6.034184490472195e-07,
"loss": 0.1142,
"step": 4290
},
{
"epoch": 0.86,
"grad_norm": 0.6527359485626221,
"learning_rate": 5.869018763200929e-07,
"loss": 0.1076,
"step": 4300
},
{
"epoch": 0.862,
"grad_norm": 0.403913676738739,
"learning_rate": 5.706003917695619e-07,
"loss": 0.1083,
"step": 4310
},
{
"epoch": 0.864,
"grad_norm": 0.4722561240196228,
"learning_rate": 5.545147899069836e-07,
"loss": 0.115,
"step": 4320
},
{
"epoch": 0.866,
"grad_norm": 0.3714640736579895,
"learning_rate": 5.386458547219026e-07,
"loss": 0.1059,
"step": 4330
},
{
"epoch": 0.868,
"grad_norm": 0.7594704627990723,
"learning_rate": 5.229943596438297e-07,
"loss": 0.1132,
"step": 4340
},
{
"epoch": 0.87,
"grad_norm": 0.4244316518306732,
"learning_rate": 5.075610675045567e-07,
"loss": 0.1063,
"step": 4350
},
{
"epoch": 0.872,
"grad_norm": 0.8828716278076172,
"learning_rate": 4.92346730500966e-07,
"loss": 0.1034,
"step": 4360
},
{
"epoch": 0.874,
"grad_norm": 0.46588724851608276,
"learning_rate": 4.773520901583801e-07,
"loss": 0.1064,
"step": 4370
},
{
"epoch": 0.876,
"grad_norm": 0.9723156094551086,
"learning_rate": 4.625778772944156e-07,
"loss": 0.1066,
"step": 4380
},
{
"epoch": 0.878,
"grad_norm": 0.4076969027519226,
"learning_rate": 4.480248119833641e-07,
"loss": 0.1114,
"step": 4390
},
{
"epoch": 0.88,
"grad_norm": 0.37374594807624817,
"learning_rate": 4.33693603521097e-07,
"loss": 0.1127,
"step": 4400
},
{
"epoch": 0.882,
"grad_norm": 0.5135172605514526,
"learning_rate": 4.195849503904975e-07,
"loss": 0.1059,
"step": 4410
},
{
"epoch": 0.884,
"grad_norm": 0.341399222612381,
"learning_rate": 4.056995402274122e-07,
"loss": 0.1063,
"step": 4420
},
{
"epoch": 0.886,
"grad_norm": 0.6200608015060425,
"learning_rate": 3.920380497871473e-07,
"loss": 0.1063,
"step": 4430
},
{
"epoch": 0.888,
"grad_norm": 1.903733730316162,
"learning_rate": 3.7860114491147017e-07,
"loss": 0.1,
"step": 4440
},
{
"epoch": 0.89,
"grad_norm": 1.135728359222412,
"learning_rate": 3.6538948049616886e-07,
"loss": 0.115,
"step": 4450
},
{
"epoch": 0.892,
"grad_norm": 0.3601432740688324,
"learning_rate": 3.524037004591274e-07,
"loss": 0.1061,
"step": 4460
},
{
"epoch": 0.894,
"grad_norm": 0.6803275346755981,
"learning_rate": 3.396444377089453e-07,
"loss": 0.1156,
"step": 4470
},
{
"epoch": 0.896,
"grad_norm": 0.5160225033760071,
"learning_rate": 3.271123141140886e-07,
"loss": 0.1056,
"step": 4480
},
{
"epoch": 0.898,
"grad_norm": 0.4760143458843231,
"learning_rate": 3.148079404725801e-07,
"loss": 0.103,
"step": 4490
},
{
"epoch": 0.9,
"grad_norm": 0.5659618377685547,
"learning_rate": 3.027319164822329e-07,
"loss": 0.105,
"step": 4500
},
{
"epoch": 0.902,
"grad_norm": 0.4368997812271118,
"learning_rate": 2.908848307114198e-07,
"loss": 0.1093,
"step": 4510
},
{
"epoch": 0.904,
"grad_norm": 0.4630069434642792,
"learning_rate": 2.792672605703867e-07,
"loss": 0.1052,
"step": 4520
},
{
"epoch": 0.906,
"grad_norm": 0.6788815855979919,
"learning_rate": 2.6787977228311336e-07,
"loss": 0.1077,
"step": 4530
},
{
"epoch": 0.908,
"grad_norm": 0.42769503593444824,
"learning_rate": 2.5672292085971276e-07,
"loss": 0.1125,
"step": 4540
},
{
"epoch": 0.91,
"grad_norm": 0.4623875617980957,
"learning_rate": 2.457972500693834e-07,
"loss": 0.1092,
"step": 4550
},
{
"epoch": 0.912,
"grad_norm": 2.632136821746826,
"learning_rate": 2.351032924139063e-07,
"loss": 0.1037,
"step": 4560
},
{
"epoch": 0.914,
"grad_norm": 0.5099290609359741,
"learning_rate": 2.2464156910168954e-07,
"loss": 0.1083,
"step": 4570
},
{
"epoch": 0.916,
"grad_norm": 0.6304298043251038,
"learning_rate": 2.1441259002236924e-07,
"loss": 0.1119,
"step": 4580
},
{
"epoch": 0.918,
"grad_norm": 0.7745104432106018,
"learning_rate": 2.0441685372195487e-07,
"loss": 0.1065,
"step": 4590
},
{
"epoch": 0.92,
"grad_norm": 0.4200075566768646,
"learning_rate": 1.9465484737853092e-07,
"loss": 0.1054,
"step": 4600
},
{
"epoch": 0.922,
"grad_norm": 0.42068639397621155,
"learning_rate": 1.8512704677851489e-07,
"loss": 0.1067,
"step": 4610
},
{
"epoch": 0.924,
"grad_norm": 2.7736501693725586,
"learning_rate": 1.758339162934658e-07,
"loss": 0.1045,
"step": 4620
},
{
"epoch": 0.926,
"grad_norm": 0.7401975989341736,
"learning_rate": 1.6677590885745388e-07,
"loss": 0.1101,
"step": 4630
},
{
"epoch": 0.928,
"grad_norm": 2.9965381622314453,
"learning_rate": 1.5795346594498162e-07,
"loss": 0.1089,
"step": 4640
},
{
"epoch": 0.93,
"grad_norm": 0.3762933015823364,
"learning_rate": 1.4936701754947104e-07,
"loss": 0.1121,
"step": 4650
},
{
"epoch": 0.932,
"grad_norm": 0.6434100866317749,
"learning_rate": 1.4101698216230254e-07,
"loss": 0.1107,
"step": 4660
},
{
"epoch": 0.934,
"grad_norm": 0.4862038195133209,
"learning_rate": 1.3290376675242022e-07,
"loss": 0.1055,
"step": 4670
},
{
"epoch": 0.936,
"grad_norm": 1.0907853841781616,
"learning_rate": 1.2502776674649776e-07,
"loss": 0.1089,
"step": 4680
},
{
"epoch": 0.938,
"grad_norm": 1.1990622282028198,
"learning_rate": 1.1738936600966366e-07,
"loss": 0.1097,
"step": 4690
},
{
"epoch": 0.94,
"grad_norm": 0.6390599012374878,
"learning_rate": 1.0998893682679479e-07,
"loss": 0.1157,
"step": 4700
},
{
"epoch": 0.942,
"grad_norm": 0.5979716777801514,
"learning_rate": 1.0282683988436792e-07,
"loss": 0.1102,
"step": 4710
},
{
"epoch": 0.944,
"grad_norm": 0.5007372498512268,
"learning_rate": 9.590342425288446e-08,
"loss": 0.0998,
"step": 4720
},
{
"epoch": 0.946,
"grad_norm": 0.3509086072444916,
"learning_rate": 8.921902736985399e-08,
"loss": 0.111,
"step": 4730
},
{
"epoch": 0.948,
"grad_norm": 0.6750816106796265,
"learning_rate": 8.277397502335194e-08,
"loss": 0.1031,
"step": 4740
},
{
"epoch": 0.95,
"grad_norm": 0.5669856071472168,
"learning_rate": 7.656858133613498e-08,
"loss": 0.1076,
"step": 4750
},
{
"epoch": 0.952,
"grad_norm": 0.5979962944984436,
"learning_rate": 7.060314875033836e-08,
"loss": 0.1016,
"step": 4760
},
{
"epoch": 0.954,
"grad_norm": 1.0432602167129517,
"learning_rate": 6.487796801272983e-08,
"loss": 0.1137,
"step": 4770
},
{
"epoch": 0.956,
"grad_norm": 0.6175380945205688,
"learning_rate": 5.939331816054161e-08,
"loss": 0.1108,
"step": 4780
},
{
"epoch": 0.958,
"grad_norm": 0.4083493947982788,
"learning_rate": 5.414946650786957e-08,
"loss": 0.1015,
"step": 4790
},
{
"epoch": 0.96,
"grad_norm": 0.4828687310218811,
"learning_rate": 4.914666863264528e-08,
"loss": 0.1038,
"step": 4800
},
{
"epoch": 0.962,
"grad_norm": 0.45067399740219116,
"learning_rate": 4.438516836417994e-08,
"loss": 0.1044,
"step": 4810
},
{
"epoch": 0.964,
"grad_norm": 0.6787665486335754,
"learning_rate": 3.986519777127884e-08,
"loss": 0.1013,
"step": 4820
},
{
"epoch": 0.966,
"grad_norm": 0.529229998588562,
"learning_rate": 3.558697715093207e-08,
"loss": 0.1168,
"step": 4830
},
{
"epoch": 0.968,
"grad_norm": 0.9952864050865173,
"learning_rate": 3.1550715017575895e-08,
"loss": 0.1068,
"step": 4840
},
{
"epoch": 0.97,
"grad_norm": 2.3543412685394287,
"learning_rate": 2.7756608092933678e-08,
"loss": 0.1169,
"step": 4850
},
{
"epoch": 1.0002,
"grad_norm": 0.6844854950904846,
"learning_rate": 2.4204841296424086e-08,
"loss": 0.1223,
"step": 4860
},
{
"epoch": 1.0022,
"grad_norm": 0.5493595600128174,
"learning_rate": 2.0895587736149414e-08,
"loss": 0.1135,
"step": 4870
},
{
"epoch": 1.0042,
"grad_norm": 0.3710947632789612,
"learning_rate": 1.7829008700460116e-08,
"loss": 0.1032,
"step": 4880
},
{
"epoch": 1.0062,
"grad_norm": 0.3379112482070923,
"learning_rate": 1.500525365009109e-08,
"loss": 0.1058,
"step": 4890
},
{
"epoch": 1.0082,
"grad_norm": 0.4570133686065674,
"learning_rate": 1.2424460210881394e-08,
"loss": 0.1146,
"step": 4900
},
{
"epoch": 1.0102,
"grad_norm": 0.41871699690818787,
"learning_rate": 1.008675416706073e-08,
"loss": 0.0956,
"step": 4910
},
{
"epoch": 1.0122,
"grad_norm": 0.34792661666870117,
"learning_rate": 7.992249455124889e-09,
"loss": 0.0974,
"step": 4920
},
{
"epoch": 1.0142,
"grad_norm": 1.3189023733139038,
"learning_rate": 6.141048158277429e-09,
"loss": 0.1055,
"step": 4930
},
{
"epoch": 1.0162,
"grad_norm": 0.6436313390731812,
"learning_rate": 4.533240501459202e-09,
"loss": 0.1092,
"step": 4940
},
{
"epoch": 1.0182,
"grad_norm": 0.40070950984954834,
"learning_rate": 3.1689048469457638e-09,
"loss": 0.1065,
"step": 4950
},
{
"epoch": 1.0202,
"grad_norm": 0.31388986110687256,
"learning_rate": 2.0481076905332074e-09,
"loss": 0.1114,
"step": 4960
},
{
"epoch": 1.0222,
"grad_norm": 0.47767698764801025,
"learning_rate": 1.170903658293532e-09,
"loss": 0.1065,
"step": 4970
},
{
"epoch": 1.0242,
"grad_norm": 0.8886580467224121,
"learning_rate": 5.373355039128836e-10,
"loss": 0.1032,
"step": 4980
},
{
"epoch": 1.0262,
"grad_norm": 0.49497994780540466,
"learning_rate": 1.4743410661044454e-10,
"loss": 0.1148,
"step": 4990
},
{
"epoch": 1.0282,
"grad_norm": 0.4880732297897339,
"learning_rate": 1.2184696296380083e-12,
"loss": 0.1156,
"step": 5000
}
],
"logging_steps": 10,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 6916730815250432.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}