so101-tapes-cleanup / trainer_state.json
yuk6ra's picture
5000 step
e3d1e47 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 7.194244604316546,
"eval_steps": 500,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.014388489208633094,
"grad_norm": 11.563289642333984,
"learning_rate": 3.6e-06,
"loss": 1.068,
"step": 10
},
{
"epoch": 0.02877697841726619,
"grad_norm": 2.7710273265838623,
"learning_rate": 7.6e-06,
"loss": 0.5179,
"step": 20
},
{
"epoch": 0.04316546762589928,
"grad_norm": 1.8126420974731445,
"learning_rate": 1.16e-05,
"loss": 0.2288,
"step": 30
},
{
"epoch": 0.05755395683453238,
"grad_norm": 1.836556315422058,
"learning_rate": 1.56e-05,
"loss": 0.1786,
"step": 40
},
{
"epoch": 0.07194244604316546,
"grad_norm": 0.8794971704483032,
"learning_rate": 1.9600000000000002e-05,
"loss": 0.1522,
"step": 50
},
{
"epoch": 0.08633093525179857,
"grad_norm": 1.3779667615890503,
"learning_rate": 2.36e-05,
"loss": 0.1348,
"step": 60
},
{
"epoch": 0.10071942446043165,
"grad_norm": 0.7911031246185303,
"learning_rate": 2.7600000000000003e-05,
"loss": 0.1054,
"step": 70
},
{
"epoch": 0.11510791366906475,
"grad_norm": 0.577821671962738,
"learning_rate": 3.16e-05,
"loss": 0.1169,
"step": 80
},
{
"epoch": 0.12949640287769784,
"grad_norm": 1.065459132194519,
"learning_rate": 3.56e-05,
"loss": 0.0928,
"step": 90
},
{
"epoch": 0.14388489208633093,
"grad_norm": 1.0262460708618164,
"learning_rate": 3.960000000000001e-05,
"loss": 0.0955,
"step": 100
},
{
"epoch": 0.15827338129496402,
"grad_norm": 0.8101535439491272,
"learning_rate": 4.36e-05,
"loss": 0.0839,
"step": 110
},
{
"epoch": 0.17266187050359713,
"grad_norm": 0.9257287383079529,
"learning_rate": 4.76e-05,
"loss": 0.0843,
"step": 120
},
{
"epoch": 0.18705035971223022,
"grad_norm": 0.7149928212165833,
"learning_rate": 5.16e-05,
"loss": 0.0872,
"step": 130
},
{
"epoch": 0.2014388489208633,
"grad_norm": 0.7855506539344788,
"learning_rate": 5.560000000000001e-05,
"loss": 0.0811,
"step": 140
},
{
"epoch": 0.2158273381294964,
"grad_norm": 0.9376429915428162,
"learning_rate": 5.96e-05,
"loss": 0.0826,
"step": 150
},
{
"epoch": 0.2302158273381295,
"grad_norm": 1.1260603666305542,
"learning_rate": 6.36e-05,
"loss": 0.0695,
"step": 160
},
{
"epoch": 0.2446043165467626,
"grad_norm": 1.1379331350326538,
"learning_rate": 6.76e-05,
"loss": 0.0788,
"step": 170
},
{
"epoch": 0.2589928057553957,
"grad_norm": 1.1374833583831787,
"learning_rate": 7.16e-05,
"loss": 0.0881,
"step": 180
},
{
"epoch": 0.2733812949640288,
"grad_norm": 1.0959501266479492,
"learning_rate": 7.560000000000001e-05,
"loss": 0.0805,
"step": 190
},
{
"epoch": 0.28776978417266186,
"grad_norm": 1.0162736177444458,
"learning_rate": 7.960000000000001e-05,
"loss": 0.0753,
"step": 200
},
{
"epoch": 0.302158273381295,
"grad_norm": 0.607140302658081,
"learning_rate": 8.36e-05,
"loss": 0.079,
"step": 210
},
{
"epoch": 0.31654676258992803,
"grad_norm": 0.7478259205818176,
"learning_rate": 8.76e-05,
"loss": 0.0765,
"step": 220
},
{
"epoch": 0.33093525179856115,
"grad_norm": 1.0972402095794678,
"learning_rate": 9.16e-05,
"loss": 0.0719,
"step": 230
},
{
"epoch": 0.34532374100719426,
"grad_norm": 0.951613187789917,
"learning_rate": 9.56e-05,
"loss": 0.0667,
"step": 240
},
{
"epoch": 0.3597122302158273,
"grad_norm": 0.7715170979499817,
"learning_rate": 9.960000000000001e-05,
"loss": 0.0698,
"step": 250
},
{
"epoch": 0.37410071942446044,
"grad_norm": 0.6283460259437561,
"learning_rate": 9.999911419878559e-05,
"loss": 0.0617,
"step": 260
},
{
"epoch": 0.38848920863309355,
"grad_norm": 0.7529840469360352,
"learning_rate": 9.999605221019081e-05,
"loss": 0.0679,
"step": 270
},
{
"epoch": 0.4028776978417266,
"grad_norm": 0.5862048864364624,
"learning_rate": 9.999080323230761e-05,
"loss": 0.072,
"step": 280
},
{
"epoch": 0.4172661870503597,
"grad_norm": 0.7532249689102173,
"learning_rate": 9.998336749474329e-05,
"loss": 0.074,
"step": 290
},
{
"epoch": 0.4316546762589928,
"grad_norm": 0.8006102442741394,
"learning_rate": 9.997374532276107e-05,
"loss": 0.0596,
"step": 300
},
{
"epoch": 0.4460431654676259,
"grad_norm": 0.8201322555541992,
"learning_rate": 9.996193713726596e-05,
"loss": 0.0682,
"step": 310
},
{
"epoch": 0.460431654676259,
"grad_norm": 0.884371280670166,
"learning_rate": 9.994794345478624e-05,
"loss": 0.0687,
"step": 320
},
{
"epoch": 0.4748201438848921,
"grad_norm": 1.0305918455123901,
"learning_rate": 9.99317648874509e-05,
"loss": 0.066,
"step": 330
},
{
"epoch": 0.4892086330935252,
"grad_norm": 1.2047297954559326,
"learning_rate": 9.991340214296292e-05,
"loss": 0.0587,
"step": 340
},
{
"epoch": 0.5035971223021583,
"grad_norm": 1.1814906597137451,
"learning_rate": 9.989285602456819e-05,
"loss": 0.0584,
"step": 350
},
{
"epoch": 0.5179856115107914,
"grad_norm": 1.0936027765274048,
"learning_rate": 9.98701274310205e-05,
"loss": 0.0592,
"step": 360
},
{
"epoch": 0.5323741007194245,
"grad_norm": 0.7059155106544495,
"learning_rate": 9.984521735654218e-05,
"loss": 0.0648,
"step": 370
},
{
"epoch": 0.5467625899280576,
"grad_norm": 0.49328455328941345,
"learning_rate": 9.981812689078057e-05,
"loss": 0.0611,
"step": 380
},
{
"epoch": 0.5611510791366906,
"grad_norm": 0.9345883131027222,
"learning_rate": 9.978885721876041e-05,
"loss": 0.0566,
"step": 390
},
{
"epoch": 0.5755395683453237,
"grad_norm": 0.44096633791923523,
"learning_rate": 9.975740962083198e-05,
"loss": 0.0485,
"step": 400
},
{
"epoch": 0.5899280575539568,
"grad_norm": 0.4832633137702942,
"learning_rate": 9.972378547261504e-05,
"loss": 0.0569,
"step": 410
},
{
"epoch": 0.60431654676259,
"grad_norm": 0.5535807013511658,
"learning_rate": 9.968798624493885e-05,
"loss": 0.0489,
"step": 420
},
{
"epoch": 0.6187050359712231,
"grad_norm": 0.5854157209396362,
"learning_rate": 9.965001350377753e-05,
"loss": 0.0526,
"step": 430
},
{
"epoch": 0.6330935251798561,
"grad_norm": 0.6552462577819824,
"learning_rate": 9.960986891018183e-05,
"loss": 0.0456,
"step": 440
},
{
"epoch": 0.6474820143884892,
"grad_norm": 0.5227953195571899,
"learning_rate": 9.95675542202063e-05,
"loss": 0.0536,
"step": 450
},
{
"epoch": 0.6618705035971223,
"grad_norm": 0.896914005279541,
"learning_rate": 9.952307128483256e-05,
"loss": 0.0522,
"step": 460
},
{
"epoch": 0.6762589928057554,
"grad_norm": 0.49559199810028076,
"learning_rate": 9.947642204988835e-05,
"loss": 0.0612,
"step": 470
},
{
"epoch": 0.6906474820143885,
"grad_norm": 0.6491966843605042,
"learning_rate": 9.942760855596226e-05,
"loss": 0.0485,
"step": 480
},
{
"epoch": 0.7050359712230215,
"grad_norm": 0.7347814440727234,
"learning_rate": 9.937663293831471e-05,
"loss": 0.0444,
"step": 490
},
{
"epoch": 0.7194244604316546,
"grad_norm": 0.8106439113616943,
"learning_rate": 9.932349742678433e-05,
"loss": 0.0491,
"step": 500
},
{
"epoch": 0.7338129496402878,
"grad_norm": 0.5709072947502136,
"learning_rate": 9.926820434569051e-05,
"loss": 0.0471,
"step": 510
},
{
"epoch": 0.7482014388489209,
"grad_norm": 0.4296235144138336,
"learning_rate": 9.921075611373179e-05,
"loss": 0.0461,
"step": 520
},
{
"epoch": 0.762589928057554,
"grad_norm": 0.5036985874176025,
"learning_rate": 9.915115524387988e-05,
"loss": 0.05,
"step": 530
},
{
"epoch": 0.7769784172661871,
"grad_norm": 0.5506060719490051,
"learning_rate": 9.908940434326997e-05,
"loss": 0.0496,
"step": 540
},
{
"epoch": 0.7913669064748201,
"grad_norm": 0.8265791535377502,
"learning_rate": 9.902550611308645e-05,
"loss": 0.0508,
"step": 550
},
{
"epoch": 0.8057553956834532,
"grad_norm": 0.5826058387756348,
"learning_rate": 9.895946334844494e-05,
"loss": 0.0467,
"step": 560
},
{
"epoch": 0.8201438848920863,
"grad_norm": 0.327493280172348,
"learning_rate": 9.889127893826989e-05,
"loss": 0.0487,
"step": 570
},
{
"epoch": 0.8345323741007195,
"grad_norm": 0.6103057265281677,
"learning_rate": 9.882095586516831e-05,
"loss": 0.0474,
"step": 580
},
{
"epoch": 0.8489208633093526,
"grad_norm": 0.6462867856025696,
"learning_rate": 9.874849720529921e-05,
"loss": 0.0454,
"step": 590
},
{
"epoch": 0.8633093525179856,
"grad_norm": 0.49559682607650757,
"learning_rate": 9.867390612823914e-05,
"loss": 0.0465,
"step": 600
},
{
"epoch": 0.8776978417266187,
"grad_norm": 0.5911785960197449,
"learning_rate": 9.859718589684344e-05,
"loss": 0.053,
"step": 610
},
{
"epoch": 0.8920863309352518,
"grad_norm": 0.45888659358024597,
"learning_rate": 9.851833986710353e-05,
"loss": 0.0428,
"step": 620
},
{
"epoch": 0.9064748201438849,
"grad_norm": 0.4031722843647003,
"learning_rate": 9.843737148800023e-05,
"loss": 0.0367,
"step": 630
},
{
"epoch": 0.920863309352518,
"grad_norm": 0.46807757019996643,
"learning_rate": 9.835428430135271e-05,
"loss": 0.0399,
"step": 640
},
{
"epoch": 0.935251798561151,
"grad_norm": 0.5714170932769775,
"learning_rate": 9.82690819416637e-05,
"loss": 0.0445,
"step": 650
},
{
"epoch": 0.9496402877697842,
"grad_norm": 0.5278415679931641,
"learning_rate": 9.818176813596041e-05,
"loss": 0.0418,
"step": 660
},
{
"epoch": 0.9640287769784173,
"grad_norm": 0.44005468487739563,
"learning_rate": 9.809234670363159e-05,
"loss": 0.0496,
"step": 670
},
{
"epoch": 0.9784172661870504,
"grad_norm": 0.3497026562690735,
"learning_rate": 9.800082155626034e-05,
"loss": 0.0498,
"step": 680
},
{
"epoch": 0.9928057553956835,
"grad_norm": 0.5186319351196289,
"learning_rate": 9.790719669745312e-05,
"loss": 0.0435,
"step": 690
},
{
"epoch": 1.0071942446043165,
"grad_norm": 0.5804557800292969,
"learning_rate": 9.781147622266455e-05,
"loss": 0.046,
"step": 700
},
{
"epoch": 1.0215827338129497,
"grad_norm": 0.526941180229187,
"learning_rate": 9.771366431901831e-05,
"loss": 0.0518,
"step": 710
},
{
"epoch": 1.0359712230215827,
"grad_norm": 0.587197482585907,
"learning_rate": 9.761376526512394e-05,
"loss": 0.0457,
"step": 720
},
{
"epoch": 1.0503597122302157,
"grad_norm": 0.5105483531951904,
"learning_rate": 9.751178343088963e-05,
"loss": 0.044,
"step": 730
},
{
"epoch": 1.064748201438849,
"grad_norm": 0.49746617674827576,
"learning_rate": 9.740772327733123e-05,
"loss": 0.0384,
"step": 740
},
{
"epoch": 1.079136690647482,
"grad_norm": 0.6755368113517761,
"learning_rate": 9.730158935637697e-05,
"loss": 0.0394,
"step": 750
},
{
"epoch": 1.0935251798561152,
"grad_norm": 0.4769406318664551,
"learning_rate": 9.719338631066834e-05,
"loss": 0.0369,
"step": 760
},
{
"epoch": 1.1079136690647482,
"grad_norm": 0.6355546712875366,
"learning_rate": 9.708311887335713e-05,
"loss": 0.0373,
"step": 770
},
{
"epoch": 1.1223021582733812,
"grad_norm": 0.43105027079582214,
"learning_rate": 9.697079186789823e-05,
"loss": 0.0386,
"step": 780
},
{
"epoch": 1.1366906474820144,
"grad_norm": 0.32694926857948303,
"learning_rate": 9.685641020783876e-05,
"loss": 0.0377,
"step": 790
},
{
"epoch": 1.1510791366906474,
"grad_norm": 0.49410852789878845,
"learning_rate": 9.67399788966031e-05,
"loss": 0.0419,
"step": 800
},
{
"epoch": 1.1654676258992807,
"grad_norm": 0.6121776700019836,
"learning_rate": 9.662150302727395e-05,
"loss": 0.0421,
"step": 810
},
{
"epoch": 1.1798561151079137,
"grad_norm": 0.6202641129493713,
"learning_rate": 9.650098778236968e-05,
"loss": 0.0384,
"step": 820
},
{
"epoch": 1.1942446043165469,
"grad_norm": 0.49695542454719543,
"learning_rate": 9.637843843361749e-05,
"loss": 0.0451,
"step": 830
},
{
"epoch": 1.20863309352518,
"grad_norm": 0.4063042402267456,
"learning_rate": 9.62538603417229e-05,
"loss": 0.0379,
"step": 840
},
{
"epoch": 1.223021582733813,
"grad_norm": 0.4808686375617981,
"learning_rate": 9.612725895613526e-05,
"loss": 0.0453,
"step": 850
},
{
"epoch": 1.2374100719424461,
"grad_norm": 0.6141266822814941,
"learning_rate": 9.599863981480926e-05,
"loss": 0.0391,
"step": 860
},
{
"epoch": 1.2517985611510791,
"grad_norm": 0.41138482093811035,
"learning_rate": 9.586800854396283e-05,
"loss": 0.0405,
"step": 870
},
{
"epoch": 1.2661870503597124,
"grad_norm": 0.4762343168258667,
"learning_rate": 9.573537085783095e-05,
"loss": 0.0363,
"step": 880
},
{
"epoch": 1.2805755395683454,
"grad_norm": 0.44630688428878784,
"learning_rate": 9.560073255841571e-05,
"loss": 0.036,
"step": 890
},
{
"epoch": 1.2949640287769784,
"grad_norm": 0.5627872347831726,
"learning_rate": 9.546409953523247e-05,
"loss": 0.0436,
"step": 900
},
{
"epoch": 1.3093525179856116,
"grad_norm": 0.3299891948699951,
"learning_rate": 9.532547776505229e-05,
"loss": 0.0322,
"step": 910
},
{
"epoch": 1.3237410071942446,
"grad_norm": 0.3779972791671753,
"learning_rate": 9.518487331164048e-05,
"loss": 0.0401,
"step": 920
},
{
"epoch": 1.3381294964028778,
"grad_norm": 0.3103158175945282,
"learning_rate": 9.504229232549134e-05,
"loss": 0.039,
"step": 930
},
{
"epoch": 1.3525179856115108,
"grad_norm": 0.6494861841201782,
"learning_rate": 9.489774104355909e-05,
"loss": 0.0399,
"step": 940
},
{
"epoch": 1.3669064748201438,
"grad_norm": 0.44805973768234253,
"learning_rate": 9.475122578898507e-05,
"loss": 0.0433,
"step": 950
},
{
"epoch": 1.381294964028777,
"grad_norm": 0.46079954504966736,
"learning_rate": 9.460275297082119e-05,
"loss": 0.0363,
"step": 960
},
{
"epoch": 1.39568345323741,
"grad_norm": 0.44950559735298157,
"learning_rate": 9.445232908374948e-05,
"loss": 0.0343,
"step": 970
},
{
"epoch": 1.4100719424460433,
"grad_norm": 0.5099852085113525,
"learning_rate": 9.429996070779808e-05,
"loss": 0.0419,
"step": 980
},
{
"epoch": 1.4244604316546763,
"grad_norm": 0.4200413227081299,
"learning_rate": 9.414565450805333e-05,
"loss": 0.0413,
"step": 990
},
{
"epoch": 1.4388489208633093,
"grad_norm": 0.5106402635574341,
"learning_rate": 9.398941723436831e-05,
"loss": 0.0327,
"step": 1000
},
{
"epoch": 1.4532374100719425,
"grad_norm": 0.3222605586051941,
"learning_rate": 9.383125572106752e-05,
"loss": 0.0324,
"step": 1010
},
{
"epoch": 1.4676258992805755,
"grad_norm": 0.49928081035614014,
"learning_rate": 9.367117688664791e-05,
"loss": 0.036,
"step": 1020
},
{
"epoch": 1.4820143884892087,
"grad_norm": 0.28545066714286804,
"learning_rate": 9.35091877334763e-05,
"loss": 0.0354,
"step": 1030
},
{
"epoch": 1.4964028776978417,
"grad_norm": 0.4319049119949341,
"learning_rate": 9.334529534748297e-05,
"loss": 0.0368,
"step": 1040
},
{
"epoch": 1.5107913669064748,
"grad_norm": 0.4933834671974182,
"learning_rate": 9.317950689785188e-05,
"loss": 0.0306,
"step": 1050
},
{
"epoch": 1.5251798561151078,
"grad_norm": 0.49777206778526306,
"learning_rate": 9.301182963670688e-05,
"loss": 0.0362,
"step": 1060
},
{
"epoch": 1.539568345323741,
"grad_norm": 0.3823613226413727,
"learning_rate": 9.284227089879456e-05,
"loss": 0.0365,
"step": 1070
},
{
"epoch": 1.5539568345323742,
"grad_norm": 0.5456088781356812,
"learning_rate": 9.26708381011634e-05,
"loss": 0.035,
"step": 1080
},
{
"epoch": 1.5683453237410072,
"grad_norm": 0.3279576003551483,
"learning_rate": 9.249753874283937e-05,
"loss": 0.027,
"step": 1090
},
{
"epoch": 1.5827338129496402,
"grad_norm": 0.380316823720932,
"learning_rate": 9.232238040449779e-05,
"loss": 0.034,
"step": 1100
},
{
"epoch": 1.5971223021582732,
"grad_norm": 0.3451821804046631,
"learning_rate": 9.214537074813181e-05,
"loss": 0.032,
"step": 1110
},
{
"epoch": 1.6115107913669064,
"grad_norm": 0.38634076714515686,
"learning_rate": 9.196651751671724e-05,
"loss": 0.0361,
"step": 1120
},
{
"epoch": 1.6258992805755397,
"grad_norm": 0.5205132961273193,
"learning_rate": 9.178582853387384e-05,
"loss": 0.0404,
"step": 1130
},
{
"epoch": 1.6402877697841727,
"grad_norm": 0.33734244108200073,
"learning_rate": 9.160331170352304e-05,
"loss": 0.0349,
"step": 1140
},
{
"epoch": 1.6546762589928057,
"grad_norm": 0.5757596492767334,
"learning_rate": 9.141897500954229e-05,
"loss": 0.0387,
"step": 1150
},
{
"epoch": 1.6690647482014387,
"grad_norm": 0.37681758403778076,
"learning_rate": 9.123282651541576e-05,
"loss": 0.0348,
"step": 1160
},
{
"epoch": 1.683453237410072,
"grad_norm": 0.5754590034484863,
"learning_rate": 9.104487436388161e-05,
"loss": 0.0384,
"step": 1170
},
{
"epoch": 1.6978417266187051,
"grad_norm": 0.38294652104377747,
"learning_rate": 9.085512677657582e-05,
"loss": 0.0349,
"step": 1180
},
{
"epoch": 1.7122302158273381,
"grad_norm": 0.3914676010608673,
"learning_rate": 9.066359205367258e-05,
"loss": 0.0362,
"step": 1190
},
{
"epoch": 1.7266187050359711,
"grad_norm": 0.43758222460746765,
"learning_rate": 9.047027857352112e-05,
"loss": 0.041,
"step": 1200
},
{
"epoch": 1.7410071942446042,
"grad_norm": 0.3658917546272278,
"learning_rate": 9.027519479227935e-05,
"loss": 0.0366,
"step": 1210
},
{
"epoch": 1.7553956834532374,
"grad_norm": 0.39443308115005493,
"learning_rate": 9.007834924354383e-05,
"loss": 0.0338,
"step": 1220
},
{
"epoch": 1.7697841726618706,
"grad_norm": 0.4250149726867676,
"learning_rate": 8.987975053797655e-05,
"loss": 0.0389,
"step": 1230
},
{
"epoch": 1.7841726618705036,
"grad_norm": 0.4489368796348572,
"learning_rate": 8.967940736292825e-05,
"loss": 0.0395,
"step": 1240
},
{
"epoch": 1.7985611510791366,
"grad_norm": 0.6220043301582336,
"learning_rate": 8.947732848205846e-05,
"loss": 0.0432,
"step": 1250
},
{
"epoch": 1.8129496402877698,
"grad_norm": 0.463680624961853,
"learning_rate": 8.927352273495204e-05,
"loss": 0.0372,
"step": 1260
},
{
"epoch": 1.8273381294964028,
"grad_norm": 0.3729951083660126,
"learning_rate": 8.906799903673265e-05,
"loss": 0.0402,
"step": 1270
},
{
"epoch": 1.841726618705036,
"grad_norm": 0.45877471566200256,
"learning_rate": 8.88607663776726e-05,
"loss": 0.036,
"step": 1280
},
{
"epoch": 1.856115107913669,
"grad_norm": 0.434235543012619,
"learning_rate": 8.865183382279978e-05,
"loss": 0.0375,
"step": 1290
},
{
"epoch": 1.870503597122302,
"grad_norm": 0.45782503485679626,
"learning_rate": 8.844121051150096e-05,
"loss": 0.0329,
"step": 1300
},
{
"epoch": 1.8848920863309353,
"grad_norm": 0.36859771609306335,
"learning_rate": 8.822890565712211e-05,
"loss": 0.0344,
"step": 1310
},
{
"epoch": 1.8992805755395683,
"grad_norm": 0.46561720967292786,
"learning_rate": 8.801492854656536e-05,
"loss": 0.0355,
"step": 1320
},
{
"epoch": 1.9136690647482015,
"grad_norm": 0.33390310406684875,
"learning_rate": 8.779928853988268e-05,
"loss": 0.0333,
"step": 1330
},
{
"epoch": 1.9280575539568345,
"grad_norm": 0.5821191072463989,
"learning_rate": 8.758199506986655e-05,
"loss": 0.0316,
"step": 1340
},
{
"epoch": 1.9424460431654675,
"grad_norm": 0.33496353030204773,
"learning_rate": 8.73630576416373e-05,
"loss": 0.0299,
"step": 1350
},
{
"epoch": 1.9568345323741008,
"grad_norm": 0.34908604621887207,
"learning_rate": 8.714248583222726e-05,
"loss": 0.0318,
"step": 1360
},
{
"epoch": 1.9712230215827338,
"grad_norm": 0.3325238525867462,
"learning_rate": 8.692028929016196e-05,
"loss": 0.0268,
"step": 1370
},
{
"epoch": 1.985611510791367,
"grad_norm": 0.26410147547721863,
"learning_rate": 8.669647773503797e-05,
"loss": 0.0392,
"step": 1380
},
{
"epoch": 2.0,
"grad_norm": 0.5888432264328003,
"learning_rate": 8.647106095709773e-05,
"loss": 0.03,
"step": 1390
},
{
"epoch": 2.014388489208633,
"grad_norm": 0.2478332668542862,
"learning_rate": 8.624404881680139e-05,
"loss": 0.035,
"step": 1400
},
{
"epoch": 2.028776978417266,
"grad_norm": 0.3821505606174469,
"learning_rate": 8.601545124439535e-05,
"loss": 0.0309,
"step": 1410
},
{
"epoch": 2.0431654676258995,
"grad_norm": 0.6519523859024048,
"learning_rate": 8.5785278239478e-05,
"loss": 0.0364,
"step": 1420
},
{
"epoch": 2.0575539568345325,
"grad_norm": 0.3599141538143158,
"learning_rate": 8.555353987056224e-05,
"loss": 0.0352,
"step": 1430
},
{
"epoch": 2.0719424460431655,
"grad_norm": 0.36180347204208374,
"learning_rate": 8.532024627463505e-05,
"loss": 0.0329,
"step": 1440
},
{
"epoch": 2.0863309352517985,
"grad_norm": 0.3427257537841797,
"learning_rate": 8.508540765671407e-05,
"loss": 0.0334,
"step": 1450
},
{
"epoch": 2.1007194244604315,
"grad_norm": 0.3006877899169922,
"learning_rate": 8.484903428940121e-05,
"loss": 0.0296,
"step": 1460
},
{
"epoch": 2.115107913669065,
"grad_norm": 0.4378103017807007,
"learning_rate": 8.461113651243334e-05,
"loss": 0.0351,
"step": 1470
},
{
"epoch": 2.129496402877698,
"grad_norm": 0.293716162443161,
"learning_rate": 8.437172473222987e-05,
"loss": 0.0378,
"step": 1480
},
{
"epoch": 2.143884892086331,
"grad_norm": 0.533905029296875,
"learning_rate": 8.413080942143767e-05,
"loss": 0.0306,
"step": 1490
},
{
"epoch": 2.158273381294964,
"grad_norm": 0.3864139914512634,
"learning_rate": 8.388840111847288e-05,
"loss": 0.035,
"step": 1500
},
{
"epoch": 2.172661870503597,
"grad_norm": 0.3795413672924042,
"learning_rate": 8.364451042705998e-05,
"loss": 0.0318,
"step": 1510
},
{
"epoch": 2.1870503597122304,
"grad_norm": 0.38925305008888245,
"learning_rate": 8.33991480157679e-05,
"loss": 0.033,
"step": 1520
},
{
"epoch": 2.2014388489208634,
"grad_norm": 0.510567307472229,
"learning_rate": 8.315232461754338e-05,
"loss": 0.0393,
"step": 1530
},
{
"epoch": 2.2158273381294964,
"grad_norm": 0.30273565649986267,
"learning_rate": 8.290405102924144e-05,
"loss": 0.0371,
"step": 1540
},
{
"epoch": 2.2302158273381294,
"grad_norm": 0.47864022850990295,
"learning_rate": 8.265433811115316e-05,
"loss": 0.0371,
"step": 1550
},
{
"epoch": 2.2446043165467624,
"grad_norm": 0.42311903834342957,
"learning_rate": 8.240319678653049e-05,
"loss": 0.0332,
"step": 1560
},
{
"epoch": 2.258992805755396,
"grad_norm": 0.3715522289276123,
"learning_rate": 8.215063804110857e-05,
"loss": 0.0298,
"step": 1570
},
{
"epoch": 2.273381294964029,
"grad_norm": 0.33272427320480347,
"learning_rate": 8.189667292262512e-05,
"loss": 0.0312,
"step": 1580
},
{
"epoch": 2.287769784172662,
"grad_norm": 0.4482347369194031,
"learning_rate": 8.164131254033716e-05,
"loss": 0.0297,
"step": 1590
},
{
"epoch": 2.302158273381295,
"grad_norm": 0.22606757283210754,
"learning_rate": 8.138456806453503e-05,
"loss": 0.0314,
"step": 1600
},
{
"epoch": 2.316546762589928,
"grad_norm": 0.30669596791267395,
"learning_rate": 8.112645072605386e-05,
"loss": 0.0285,
"step": 1610
},
{
"epoch": 2.3309352517985613,
"grad_norm": 0.34195154905319214,
"learning_rate": 8.086697181578222e-05,
"loss": 0.0279,
"step": 1620
},
{
"epoch": 2.3453237410071943,
"grad_norm": 0.46204662322998047,
"learning_rate": 8.060614268416823e-05,
"loss": 0.0289,
"step": 1630
},
{
"epoch": 2.3597122302158273,
"grad_norm": 0.3045788109302521,
"learning_rate": 8.034397474072309e-05,
"loss": 0.0311,
"step": 1640
},
{
"epoch": 2.3741007194244603,
"grad_norm": 0.502484142780304,
"learning_rate": 8.008047945352193e-05,
"loss": 0.0373,
"step": 1650
},
{
"epoch": 2.3884892086330938,
"grad_norm": 0.43514329195022583,
"learning_rate": 7.981566834870225e-05,
"loss": 0.0272,
"step": 1660
},
{
"epoch": 2.402877697841727,
"grad_norm": 0.4066406190395355,
"learning_rate": 7.954955300995961e-05,
"loss": 0.0287,
"step": 1670
},
{
"epoch": 2.41726618705036,
"grad_norm": 0.5418047308921814,
"learning_rate": 7.928214507804104e-05,
"loss": 0.0283,
"step": 1680
},
{
"epoch": 2.431654676258993,
"grad_norm": 0.5154935121536255,
"learning_rate": 7.901345625023576e-05,
"loss": 0.0327,
"step": 1690
},
{
"epoch": 2.446043165467626,
"grad_norm": 0.39965781569480896,
"learning_rate": 7.874349827986354e-05,
"loss": 0.0295,
"step": 1700
},
{
"epoch": 2.460431654676259,
"grad_norm": 0.37923577427864075,
"learning_rate": 7.847228297576053e-05,
"loss": 0.0308,
"step": 1710
},
{
"epoch": 2.4748201438848922,
"grad_norm": 0.38472461700439453,
"learning_rate": 7.819982220176276e-05,
"loss": 0.0348,
"step": 1720
},
{
"epoch": 2.4892086330935252,
"grad_norm": 0.42441073060035706,
"learning_rate": 7.792612787618714e-05,
"loss": 0.035,
"step": 1730
},
{
"epoch": 2.5035971223021583,
"grad_norm": 0.3353954553604126,
"learning_rate": 7.765121197131009e-05,
"loss": 0.0366,
"step": 1740
},
{
"epoch": 2.5179856115107913,
"grad_norm": 0.38094013929367065,
"learning_rate": 7.737508651284391e-05,
"loss": 0.0355,
"step": 1750
},
{
"epoch": 2.5323741007194247,
"grad_norm": 0.3948928117752075,
"learning_rate": 7.709776357941069e-05,
"loss": 0.0337,
"step": 1760
},
{
"epoch": 2.5467625899280577,
"grad_norm": 0.37004798650741577,
"learning_rate": 7.681925530201392e-05,
"loss": 0.0283,
"step": 1770
},
{
"epoch": 2.5611510791366907,
"grad_norm": 0.3940068781375885,
"learning_rate": 7.65395738635079e-05,
"loss": 0.0324,
"step": 1780
},
{
"epoch": 2.5755395683453237,
"grad_norm": 0.4402606785297394,
"learning_rate": 7.62587314980648e-05,
"loss": 0.0349,
"step": 1790
},
{
"epoch": 2.5899280575539567,
"grad_norm": 0.3379155993461609,
"learning_rate": 7.597674049063947e-05,
"loss": 0.0284,
"step": 1800
},
{
"epoch": 2.6043165467625897,
"grad_norm": 0.4218379557132721,
"learning_rate": 7.569361317643211e-05,
"loss": 0.0267,
"step": 1810
},
{
"epoch": 2.618705035971223,
"grad_norm": 0.32986360788345337,
"learning_rate": 7.540936194034865e-05,
"loss": 0.0273,
"step": 1820
},
{
"epoch": 2.633093525179856,
"grad_norm": 0.26728731393814087,
"learning_rate": 7.512399921645901e-05,
"loss": 0.0301,
"step": 1830
},
{
"epoch": 2.647482014388489,
"grad_norm": 0.32634323835372925,
"learning_rate": 7.483753748745317e-05,
"loss": 0.0269,
"step": 1840
},
{
"epoch": 2.661870503597122,
"grad_norm": 0.34928056597709656,
"learning_rate": 7.454998928409516e-05,
"loss": 0.0245,
"step": 1850
},
{
"epoch": 2.6762589928057556,
"grad_norm": 0.6099575757980347,
"learning_rate": 7.426136718467493e-05,
"loss": 0.0344,
"step": 1860
},
{
"epoch": 2.6906474820143886,
"grad_norm": 0.4516599774360657,
"learning_rate": 7.397168381445812e-05,
"loss": 0.0305,
"step": 1870
},
{
"epoch": 2.7050359712230216,
"grad_norm": 0.4836336076259613,
"learning_rate": 7.368095184513377e-05,
"loss": 0.0295,
"step": 1880
},
{
"epoch": 2.7194244604316546,
"grad_norm": 0.33341869711875916,
"learning_rate": 7.338918399426005e-05,
"loss": 0.0278,
"step": 1890
},
{
"epoch": 2.7338129496402876,
"grad_norm": 0.3752714991569519,
"learning_rate": 7.309639302470801e-05,
"loss": 0.0337,
"step": 1900
},
{
"epoch": 2.7482014388489207,
"grad_norm": 0.3812268078327179,
"learning_rate": 7.280259174410312e-05,
"loss": 0.0321,
"step": 1910
},
{
"epoch": 2.762589928057554,
"grad_norm": 0.5472700595855713,
"learning_rate": 7.250779300426517e-05,
"loss": 0.0297,
"step": 1920
},
{
"epoch": 2.776978417266187,
"grad_norm": 0.3447122275829315,
"learning_rate": 7.22120097006461e-05,
"loss": 0.0315,
"step": 1930
},
{
"epoch": 2.79136690647482,
"grad_norm": 0.3902391195297241,
"learning_rate": 7.191525477176577e-05,
"loss": 0.0315,
"step": 1940
},
{
"epoch": 2.805755395683453,
"grad_norm": 0.3023565709590912,
"learning_rate": 7.161754119864616e-05,
"loss": 0.0285,
"step": 1950
},
{
"epoch": 2.8201438848920866,
"grad_norm": 0.27357932925224304,
"learning_rate": 7.131888200424339e-05,
"loss": 0.0228,
"step": 1960
},
{
"epoch": 2.8345323741007196,
"grad_norm": 0.3916592299938202,
"learning_rate": 7.101929025287816e-05,
"loss": 0.0372,
"step": 1970
},
{
"epoch": 2.8489208633093526,
"grad_norm": 0.47285720705986023,
"learning_rate": 7.071877904966423e-05,
"loss": 0.0322,
"step": 1980
},
{
"epoch": 2.8633093525179856,
"grad_norm": 0.2980084717273712,
"learning_rate": 7.04173615399351e-05,
"loss": 0.0286,
"step": 1990
},
{
"epoch": 2.8776978417266186,
"grad_norm": 0.21847301721572876,
"learning_rate": 7.011505090866913e-05,
"loss": 0.0259,
"step": 2000
},
{
"epoch": 2.8920863309352516,
"grad_norm": 0.27048084139823914,
"learning_rate": 6.981186037991271e-05,
"loss": 0.0242,
"step": 2010
},
{
"epoch": 2.906474820143885,
"grad_norm": 0.3246220648288727,
"learning_rate": 6.950780321620174e-05,
"loss": 0.0259,
"step": 2020
},
{
"epoch": 2.920863309352518,
"grad_norm": 0.30013322830200195,
"learning_rate": 6.920289271798157e-05,
"loss": 0.0381,
"step": 2030
},
{
"epoch": 2.935251798561151,
"grad_norm": 0.39500758051872253,
"learning_rate": 6.889714222302517e-05,
"loss": 0.0349,
"step": 2040
},
{
"epoch": 2.949640287769784,
"grad_norm": 0.2924429178237915,
"learning_rate": 6.85905651058497e-05,
"loss": 0.0271,
"step": 2050
},
{
"epoch": 2.9640287769784175,
"grad_norm": 0.3150521516799927,
"learning_rate": 6.82831747771314e-05,
"loss": 0.0276,
"step": 2060
},
{
"epoch": 2.9784172661870505,
"grad_norm": 0.4109116494655609,
"learning_rate": 6.797498468311907e-05,
"loss": 0.0298,
"step": 2070
},
{
"epoch": 2.9928057553956835,
"grad_norm": 0.27924907207489014,
"learning_rate": 6.766600830504585e-05,
"loss": 0.0266,
"step": 2080
},
{
"epoch": 3.0071942446043165,
"grad_norm": 0.39713388681411743,
"learning_rate": 6.735625915853942e-05,
"loss": 0.0259,
"step": 2090
},
{
"epoch": 3.0215827338129495,
"grad_norm": 0.28341811895370483,
"learning_rate": 6.70457507930309e-05,
"loss": 0.0237,
"step": 2100
},
{
"epoch": 3.0359712230215825,
"grad_norm": 0.46310800313949585,
"learning_rate": 6.673449679116215e-05,
"loss": 0.0284,
"step": 2110
},
{
"epoch": 3.050359712230216,
"grad_norm": 0.48776841163635254,
"learning_rate": 6.642251076819148e-05,
"loss": 0.0281,
"step": 2120
},
{
"epoch": 3.064748201438849,
"grad_norm": 0.2813694477081299,
"learning_rate": 6.610980637139827e-05,
"loss": 0.0283,
"step": 2130
},
{
"epoch": 3.079136690647482,
"grad_norm": 0.3164334297180176,
"learning_rate": 6.579639727948583e-05,
"loss": 0.0245,
"step": 2140
},
{
"epoch": 3.093525179856115,
"grad_norm": 0.35484516620635986,
"learning_rate": 6.548229720198315e-05,
"loss": 0.0334,
"step": 2150
},
{
"epoch": 3.1079136690647484,
"grad_norm": 0.3275477886199951,
"learning_rate": 6.516751987864517e-05,
"loss": 0.0249,
"step": 2160
},
{
"epoch": 3.1223021582733814,
"grad_norm": 0.36483341455459595,
"learning_rate": 6.485207907885175e-05,
"loss": 0.0274,
"step": 2170
},
{
"epoch": 3.1366906474820144,
"grad_norm": 0.3230241537094116,
"learning_rate": 6.453598860100536e-05,
"loss": 0.024,
"step": 2180
},
{
"epoch": 3.1510791366906474,
"grad_norm": 0.41761744022369385,
"learning_rate": 6.421926227192749e-05,
"loss": 0.0258,
"step": 2190
},
{
"epoch": 3.1654676258992804,
"grad_norm": 0.4524862766265869,
"learning_rate": 6.390191394625381e-05,
"loss": 0.0288,
"step": 2200
},
{
"epoch": 3.1798561151079134,
"grad_norm": 0.29386061429977417,
"learning_rate": 6.358395750582817e-05,
"loss": 0.0237,
"step": 2210
},
{
"epoch": 3.194244604316547,
"grad_norm": 0.3071223795413971,
"learning_rate": 6.326540685909532e-05,
"loss": 0.0248,
"step": 2220
},
{
"epoch": 3.20863309352518,
"grad_norm": 0.3864443004131317,
"learning_rate": 6.294627594049249e-05,
"loss": 0.0258,
"step": 2230
},
{
"epoch": 3.223021582733813,
"grad_norm": 0.3610457181930542,
"learning_rate": 6.262657870983989e-05,
"loss": 0.025,
"step": 2240
},
{
"epoch": 3.237410071942446,
"grad_norm": 0.255758136510849,
"learning_rate": 6.230632915173009e-05,
"loss": 0.0262,
"step": 2250
},
{
"epoch": 3.2517985611510793,
"grad_norm": 0.2611458897590637,
"learning_rate": 6.198554127491622e-05,
"loss": 0.0246,
"step": 2260
},
{
"epoch": 3.2661870503597124,
"grad_norm": 0.36741286516189575,
"learning_rate": 6.166422911169923e-05,
"loss": 0.0259,
"step": 2270
},
{
"epoch": 3.2805755395683454,
"grad_norm": 0.3749394118785858,
"learning_rate": 6.1342406717314e-05,
"loss": 0.0259,
"step": 2280
},
{
"epoch": 3.2949640287769784,
"grad_norm": 0.28398364782333374,
"learning_rate": 6.102008816931466e-05,
"loss": 0.026,
"step": 2290
},
{
"epoch": 3.3093525179856114,
"grad_norm": 0.40850111842155457,
"learning_rate": 6.069728756695866e-05,
"loss": 0.0186,
"step": 2300
},
{
"epoch": 3.3237410071942444,
"grad_norm": 0.34636157751083374,
"learning_rate": 6.037401903059008e-05,
"loss": 0.027,
"step": 2310
},
{
"epoch": 3.338129496402878,
"grad_norm": 0.25287190079689026,
"learning_rate": 6.005029670102195e-05,
"loss": 0.0246,
"step": 2320
},
{
"epoch": 3.352517985611511,
"grad_norm": 0.3979804515838623,
"learning_rate": 5.972613473891766e-05,
"loss": 0.0273,
"step": 2330
},
{
"epoch": 3.366906474820144,
"grad_norm": 0.2626373767852783,
"learning_rate": 5.940154732417158e-05,
"loss": 0.0285,
"step": 2340
},
{
"epoch": 3.381294964028777,
"grad_norm": 0.5092778205871582,
"learning_rate": 5.907654865528876e-05,
"loss": 0.0243,
"step": 2350
},
{
"epoch": 3.3956834532374103,
"grad_norm": 0.33727580308914185,
"learning_rate": 5.875115294876381e-05,
"loss": 0.0277,
"step": 2360
},
{
"epoch": 3.4100719424460433,
"grad_norm": 0.2765815854072571,
"learning_rate": 5.842537443845908e-05,
"loss": 0.0243,
"step": 2370
},
{
"epoch": 3.4244604316546763,
"grad_norm": 0.2593759000301361,
"learning_rate": 5.809922737498198e-05,
"loss": 0.0206,
"step": 2380
},
{
"epoch": 3.4388489208633093,
"grad_norm": 0.30707883834838867,
"learning_rate": 5.777272602506165e-05,
"loss": 0.0286,
"step": 2390
},
{
"epoch": 3.4532374100719423,
"grad_norm": 0.34865009784698486,
"learning_rate": 5.744588467092483e-05,
"loss": 0.0291,
"step": 2400
},
{
"epoch": 3.4676258992805753,
"grad_norm": 0.19412925839424133,
"learning_rate": 5.7118717609671194e-05,
"loss": 0.024,
"step": 2410
},
{
"epoch": 3.4820143884892087,
"grad_norm": 0.3889428973197937,
"learning_rate": 5.679123915264786e-05,
"loss": 0.0239,
"step": 2420
},
{
"epoch": 3.4964028776978417,
"grad_norm": 0.2873934805393219,
"learning_rate": 5.646346362482342e-05,
"loss": 0.0217,
"step": 2430
},
{
"epoch": 3.5107913669064748,
"grad_norm": 0.4242481589317322,
"learning_rate": 5.613540536416132e-05,
"loss": 0.0255,
"step": 2440
},
{
"epoch": 3.5251798561151078,
"grad_norm": 0.5621308088302612,
"learning_rate": 5.5807078720992645e-05,
"loss": 0.0242,
"step": 2450
},
{
"epoch": 3.539568345323741,
"grad_norm": 0.47139909863471985,
"learning_rate": 5.547849805738836e-05,
"loss": 0.0238,
"step": 2460
},
{
"epoch": 3.553956834532374,
"grad_norm": 0.5155444741249084,
"learning_rate": 5.514967774653118e-05,
"loss": 0.0236,
"step": 2470
},
{
"epoch": 3.568345323741007,
"grad_norm": 0.19469717144966125,
"learning_rate": 5.482063217208674e-05,
"loss": 0.0251,
"step": 2480
},
{
"epoch": 3.58273381294964,
"grad_norm": 0.45209935307502747,
"learning_rate": 5.449137572757439e-05,
"loss": 0.0224,
"step": 2490
},
{
"epoch": 3.597122302158273,
"grad_norm": 0.3203049600124359,
"learning_rate": 5.4161922815737696e-05,
"loss": 0.0216,
"step": 2500
},
{
"epoch": 3.6115107913669062,
"grad_norm": 0.3630366027355194,
"learning_rate": 5.3832287847914276e-05,
"loss": 0.0239,
"step": 2510
},
{
"epoch": 3.6258992805755397,
"grad_norm": 0.18389952182769775,
"learning_rate": 5.35024852434055e-05,
"loss": 0.0218,
"step": 2520
},
{
"epoch": 3.6402877697841727,
"grad_norm": 0.239175945520401,
"learning_rate": 5.317252942884567e-05,
"loss": 0.0243,
"step": 2530
},
{
"epoch": 3.6546762589928057,
"grad_norm": 0.28686219453811646,
"learning_rate": 5.284243483757109e-05,
"loss": 0.0214,
"step": 2540
},
{
"epoch": 3.6690647482014387,
"grad_norm": 0.3766411542892456,
"learning_rate": 5.2512215908988484e-05,
"loss": 0.0273,
"step": 2550
},
{
"epoch": 3.683453237410072,
"grad_norm": 0.4206293523311615,
"learning_rate": 5.218188708794357e-05,
"loss": 0.0245,
"step": 2560
},
{
"epoch": 3.697841726618705,
"grad_norm": 0.40152212977409363,
"learning_rate": 5.18514628240891e-05,
"loss": 0.0227,
"step": 2570
},
{
"epoch": 3.712230215827338,
"grad_norm": 0.32241809368133545,
"learning_rate": 5.1520957571252795e-05,
"loss": 0.0241,
"step": 2580
},
{
"epoch": 3.726618705035971,
"grad_norm": 0.41372379660606384,
"learning_rate": 5.1190385786805106e-05,
"loss": 0.0257,
"step": 2590
},
{
"epoch": 3.741007194244604,
"grad_norm": 0.3384152054786682,
"learning_rate": 5.085976193102677e-05,
"loss": 0.0196,
"step": 2600
},
{
"epoch": 3.755395683453237,
"grad_norm": 0.35520708560943604,
"learning_rate": 5.052910046647634e-05,
"loss": 0.0203,
"step": 2610
},
{
"epoch": 3.7697841726618706,
"grad_norm": 0.21003730595111847,
"learning_rate": 5.0198415857357464e-05,
"loss": 0.02,
"step": 2620
},
{
"epoch": 3.7841726618705036,
"grad_norm": 0.306309312582016,
"learning_rate": 4.9867722568886223e-05,
"loss": 0.0207,
"step": 2630
},
{
"epoch": 3.7985611510791366,
"grad_norm": 0.44406765699386597,
"learning_rate": 4.9537035066658314e-05,
"loss": 0.019,
"step": 2640
},
{
"epoch": 3.81294964028777,
"grad_norm": 0.21013295650482178,
"learning_rate": 4.920636781601638e-05,
"loss": 0.0218,
"step": 2650
},
{
"epoch": 3.827338129496403,
"grad_norm": 0.44460663199424744,
"learning_rate": 4.88757352814172e-05,
"loss": 0.0287,
"step": 2660
},
{
"epoch": 3.841726618705036,
"grad_norm": 0.33253446221351624,
"learning_rate": 4.8545151925798924e-05,
"loss": 0.0234,
"step": 2670
},
{
"epoch": 3.856115107913669,
"grad_norm": 0.38294294476509094,
"learning_rate": 4.821463220994848e-05,
"loss": 0.0256,
"step": 2680
},
{
"epoch": 3.870503597122302,
"grad_norm": 0.2662079930305481,
"learning_rate": 4.788419059186895e-05,
"loss": 0.02,
"step": 2690
},
{
"epoch": 3.884892086330935,
"grad_norm": 0.32366669178009033,
"learning_rate": 4.7553841526147205e-05,
"loss": 0.0211,
"step": 2700
},
{
"epoch": 3.899280575539568,
"grad_norm": 0.4299909174442291,
"learning_rate": 4.722359946332156e-05,
"loss": 0.0226,
"step": 2710
},
{
"epoch": 3.9136690647482015,
"grad_norm": 0.4672795236110687,
"learning_rate": 4.6893478849249654e-05,
"loss": 0.0239,
"step": 2720
},
{
"epoch": 3.9280575539568345,
"grad_norm": 0.44387346506118774,
"learning_rate": 4.656349412447664e-05,
"loss": 0.02,
"step": 2730
},
{
"epoch": 3.9424460431654675,
"grad_norm": 0.4087350368499756,
"learning_rate": 4.623365972360337e-05,
"loss": 0.0238,
"step": 2740
},
{
"epoch": 3.956834532374101,
"grad_norm": 0.22277550399303436,
"learning_rate": 4.590399007465503e-05,
"loss": 0.0179,
"step": 2750
},
{
"epoch": 3.971223021582734,
"grad_norm": 0.3167475759983063,
"learning_rate": 4.557449959845005e-05,
"loss": 0.0215,
"step": 2760
},
{
"epoch": 3.985611510791367,
"grad_norm": 0.292919397354126,
"learning_rate": 4.524520270796927e-05,
"loss": 0.0198,
"step": 2770
},
{
"epoch": 4.0,
"grad_norm": 0.3293292224407196,
"learning_rate": 4.491611380772545e-05,
"loss": 0.0198,
"step": 2780
},
{
"epoch": 4.014388489208633,
"grad_norm": 0.2932709753513336,
"learning_rate": 4.458724729313318e-05,
"loss": 0.0193,
"step": 2790
},
{
"epoch": 4.028776978417266,
"grad_norm": 0.29950132966041565,
"learning_rate": 4.42586175498792e-05,
"loss": 0.0224,
"step": 2800
},
{
"epoch": 4.043165467625899,
"grad_norm": 0.2784114480018616,
"learning_rate": 4.3930238953293094e-05,
"loss": 0.0236,
"step": 2810
},
{
"epoch": 4.057553956834532,
"grad_norm": 0.23218660056591034,
"learning_rate": 4.360212586771847e-05,
"loss": 0.0177,
"step": 2820
},
{
"epoch": 4.071942446043165,
"grad_norm": 0.2871486246585846,
"learning_rate": 4.327429264588463e-05,
"loss": 0.022,
"step": 2830
},
{
"epoch": 4.086330935251799,
"grad_norm": 0.24163144826889038,
"learning_rate": 4.2946753628278725e-05,
"loss": 0.0203,
"step": 2840
},
{
"epoch": 4.100719424460432,
"grad_norm": 0.23145246505737305,
"learning_rate": 4.2619523142518474e-05,
"loss": 0.0169,
"step": 2850
},
{
"epoch": 4.115107913669065,
"grad_norm": 0.2641954720020294,
"learning_rate": 4.229261550272539e-05,
"loss": 0.018,
"step": 2860
},
{
"epoch": 4.129496402877698,
"grad_norm": 0.27292731404304504,
"learning_rate": 4.196604500889868e-05,
"loss": 0.0217,
"step": 2870
},
{
"epoch": 4.143884892086331,
"grad_norm": 0.26087990403175354,
"learning_rate": 4.163982594628969e-05,
"loss": 0.019,
"step": 2880
},
{
"epoch": 4.158273381294964,
"grad_norm": 0.3011043965816498,
"learning_rate": 4.131397258477702e-05,
"loss": 0.0184,
"step": 2890
},
{
"epoch": 4.172661870503597,
"grad_norm": 0.22835054993629456,
"learning_rate": 4.0988499178242315e-05,
"loss": 0.0205,
"step": 2900
},
{
"epoch": 4.18705035971223,
"grad_norm": 0.3175669014453888,
"learning_rate": 4.066341996394678e-05,
"loss": 0.0237,
"step": 2910
},
{
"epoch": 4.201438848920863,
"grad_norm": 0.23583489656448364,
"learning_rate": 4.033874916190833e-05,
"loss": 0.0177,
"step": 2920
},
{
"epoch": 4.215827338129497,
"grad_norm": 0.34252455830574036,
"learning_rate": 4.001450097427966e-05,
"loss": 0.0197,
"step": 2930
},
{
"epoch": 4.23021582733813,
"grad_norm": 0.33203423023223877,
"learning_rate": 3.9690689584726894e-05,
"loss": 0.0187,
"step": 2940
},
{
"epoch": 4.244604316546763,
"grad_norm": 0.2036760449409485,
"learning_rate": 3.936732915780923e-05,
"loss": 0.019,
"step": 2950
},
{
"epoch": 4.258992805755396,
"grad_norm": 0.23969554901123047,
"learning_rate": 3.904443383835929e-05,
"loss": 0.0204,
"step": 2960
},
{
"epoch": 4.273381294964029,
"grad_norm": 0.31450337171554565,
"learning_rate": 3.872201775086437e-05,
"loss": 0.0189,
"step": 2970
},
{
"epoch": 4.287769784172662,
"grad_norm": 0.26025325059890747,
"learning_rate": 3.8400094998848616e-05,
"loss": 0.022,
"step": 2980
},
{
"epoch": 4.302158273381295,
"grad_norm": 0.2698114514350891,
"learning_rate": 3.807867966425611e-05,
"loss": 0.021,
"step": 2990
},
{
"epoch": 4.316546762589928,
"grad_norm": 0.2846289873123169,
"learning_rate": 3.775778580683481e-05,
"loss": 0.0197,
"step": 3000
},
{
"epoch": 4.330935251798561,
"grad_norm": 0.17230279743671417,
"learning_rate": 3.743742746352156e-05,
"loss": 0.0186,
"step": 3010
},
{
"epoch": 4.345323741007194,
"grad_norm": 0.40015915036201477,
"learning_rate": 3.711761864782817e-05,
"loss": 0.0223,
"step": 3020
},
{
"epoch": 4.359712230215827,
"grad_norm": 0.29461169242858887,
"learning_rate": 3.679837334922825e-05,
"loss": 0.0171,
"step": 3030
},
{
"epoch": 4.374100719424461,
"grad_norm": 0.21421046555042267,
"learning_rate": 3.647970553254538e-05,
"loss": 0.0165,
"step": 3040
},
{
"epoch": 4.388489208633094,
"grad_norm": 0.2178724706172943,
"learning_rate": 3.61616291373422e-05,
"loss": 0.0181,
"step": 3050
},
{
"epoch": 4.402877697841727,
"grad_norm": 0.3000968396663666,
"learning_rate": 3.584415807731065e-05,
"loss": 0.0247,
"step": 3060
},
{
"epoch": 4.41726618705036,
"grad_norm": 0.34208056330680847,
"learning_rate": 3.552730623966337e-05,
"loss": 0.0242,
"step": 3070
},
{
"epoch": 4.431654676258993,
"grad_norm": 0.34540867805480957,
"learning_rate": 3.521108748452617e-05,
"loss": 0.0184,
"step": 3080
},
{
"epoch": 4.446043165467626,
"grad_norm": 0.2662169933319092,
"learning_rate": 3.489551564433186e-05,
"loss": 0.0237,
"step": 3090
},
{
"epoch": 4.460431654676259,
"grad_norm": 0.27601611614227295,
"learning_rate": 3.4580604523215006e-05,
"loss": 0.0211,
"step": 3100
},
{
"epoch": 4.474820143884892,
"grad_norm": 0.2639255225658417,
"learning_rate": 3.4266367896408216e-05,
"loss": 0.0208,
"step": 3110
},
{
"epoch": 4.489208633093525,
"grad_norm": 0.23080821335315704,
"learning_rate": 3.3952819509639534e-05,
"loss": 0.0252,
"step": 3120
},
{
"epoch": 4.503597122302159,
"grad_norm": 0.35403913259506226,
"learning_rate": 3.3639973078531165e-05,
"loss": 0.0188,
"step": 3130
},
{
"epoch": 4.517985611510792,
"grad_norm": 0.28776633739471436,
"learning_rate": 3.332784228799947e-05,
"loss": 0.0218,
"step": 3140
},
{
"epoch": 4.532374100719425,
"grad_norm": 0.1710900068283081,
"learning_rate": 3.301644079165638e-05,
"loss": 0.0218,
"step": 3150
},
{
"epoch": 4.546762589928058,
"grad_norm": 0.2477635145187378,
"learning_rate": 3.27057822112122e-05,
"loss": 0.0146,
"step": 3160
},
{
"epoch": 4.561151079136691,
"grad_norm": 0.48362207412719727,
"learning_rate": 3.239588013587958e-05,
"loss": 0.0229,
"step": 3170
},
{
"epoch": 4.575539568345324,
"grad_norm": 0.27333328127861023,
"learning_rate": 3.208674812177926e-05,
"loss": 0.0176,
"step": 3180
},
{
"epoch": 4.589928057553957,
"grad_norm": 0.2521935701370239,
"learning_rate": 3.177839969134698e-05,
"loss": 0.0195,
"step": 3190
},
{
"epoch": 4.60431654676259,
"grad_norm": 0.30661070346832275,
"learning_rate": 3.1470848332742e-05,
"loss": 0.0234,
"step": 3200
},
{
"epoch": 4.618705035971223,
"grad_norm": 0.23590540885925293,
"learning_rate": 3.116410749925708e-05,
"loss": 0.0188,
"step": 3210
},
{
"epoch": 4.633093525179856,
"grad_norm": 0.2483549416065216,
"learning_rate": 3.085819060872995e-05,
"loss": 0.0173,
"step": 3220
},
{
"epoch": 4.647482014388489,
"grad_norm": 0.27481165528297424,
"learning_rate": 3.055311104295648e-05,
"loss": 0.0225,
"step": 3230
},
{
"epoch": 4.661870503597123,
"grad_norm": 0.1598542034626007,
"learning_rate": 3.024888214710517e-05,
"loss": 0.0143,
"step": 3240
},
{
"epoch": 4.676258992805756,
"grad_norm": 0.3678795099258423,
"learning_rate": 2.994551722913349e-05,
"loss": 0.0116,
"step": 3250
},
{
"epoch": 4.690647482014389,
"grad_norm": 0.25717994570732117,
"learning_rate": 2.9643029559205727e-05,
"loss": 0.022,
"step": 3260
},
{
"epoch": 4.705035971223022,
"grad_norm": 0.2855883538722992,
"learning_rate": 2.934143236911248e-05,
"loss": 0.0154,
"step": 3270
},
{
"epoch": 4.719424460431655,
"grad_norm": 0.14246642589569092,
"learning_rate": 2.90407388516919e-05,
"loss": 0.0157,
"step": 3280
},
{
"epoch": 4.733812949640288,
"grad_norm": 0.3501797020435333,
"learning_rate": 2.8740962160252495e-05,
"loss": 0.018,
"step": 3290
},
{
"epoch": 4.748201438848921,
"grad_norm": 0.2768447697162628,
"learning_rate": 2.844211540799797e-05,
"loss": 0.0163,
"step": 3300
},
{
"epoch": 4.762589928057554,
"grad_norm": 0.20550833642482758,
"learning_rate": 2.8144211667453368e-05,
"loss": 0.0146,
"step": 3310
},
{
"epoch": 4.7769784172661875,
"grad_norm": 0.2292393296957016,
"learning_rate": 2.7847263969893344e-05,
"loss": 0.0207,
"step": 3320
},
{
"epoch": 4.7913669064748206,
"grad_norm": 0.3661034107208252,
"learning_rate": 2.7551285304772206e-05,
"loss": 0.0161,
"step": 3330
},
{
"epoch": 4.805755395683454,
"grad_norm": 0.3321722447872162,
"learning_rate": 2.7256288619155567e-05,
"loss": 0.0196,
"step": 3340
},
{
"epoch": 4.820143884892087,
"grad_norm": 0.2642754018306732,
"learning_rate": 2.6962286817154158e-05,
"loss": 0.0169,
"step": 3350
},
{
"epoch": 4.83453237410072,
"grad_norm": 0.25726601481437683,
"learning_rate": 2.6669292759359166e-05,
"loss": 0.0179,
"step": 3360
},
{
"epoch": 4.848920863309353,
"grad_norm": 0.2713489830493927,
"learning_rate": 2.637731926227993e-05,
"loss": 0.0255,
"step": 3370
},
{
"epoch": 4.863309352517986,
"grad_norm": 0.3417958617210388,
"learning_rate": 2.6086379097783033e-05,
"loss": 0.0182,
"step": 3380
},
{
"epoch": 4.877697841726619,
"grad_norm": 0.26902711391448975,
"learning_rate": 2.579648499253377e-05,
"loss": 0.0187,
"step": 3390
},
{
"epoch": 4.892086330935252,
"grad_norm": 0.26709261536598206,
"learning_rate": 2.5507649627439466e-05,
"loss": 0.0198,
"step": 3400
},
{
"epoch": 4.906474820143885,
"grad_norm": 0.42128634452819824,
"learning_rate": 2.5219885637094653e-05,
"loss": 0.0204,
"step": 3410
},
{
"epoch": 4.920863309352518,
"grad_norm": 0.2843369245529175,
"learning_rate": 2.4933205609228533e-05,
"loss": 0.0197,
"step": 3420
},
{
"epoch": 4.935251798561151,
"grad_norm": 0.2533762454986572,
"learning_rate": 2.464762208415419e-05,
"loss": 0.0147,
"step": 3430
},
{
"epoch": 4.9496402877697845,
"grad_norm": 0.23326824605464935,
"learning_rate": 2.4363147554220213e-05,
"loss": 0.0165,
"step": 3440
},
{
"epoch": 4.9640287769784175,
"grad_norm": 0.4002981185913086,
"learning_rate": 2.407979446326411e-05,
"loss": 0.0201,
"step": 3450
},
{
"epoch": 4.9784172661870505,
"grad_norm": 0.3233909010887146,
"learning_rate": 2.379757520606799e-05,
"loss": 0.0182,
"step": 3460
},
{
"epoch": 4.9928057553956835,
"grad_norm": 0.3028855323791504,
"learning_rate": 2.3516502127816455e-05,
"loss": 0.0187,
"step": 3470
},
{
"epoch": 5.0071942446043165,
"grad_norm": 0.2609952390193939,
"learning_rate": 2.323658752355647e-05,
"loss": 0.0149,
"step": 3480
},
{
"epoch": 5.0215827338129495,
"grad_norm": 0.3774644732475281,
"learning_rate": 2.2957843637659654e-05,
"loss": 0.0182,
"step": 3490
},
{
"epoch": 5.0359712230215825,
"grad_norm": 0.25385966897010803,
"learning_rate": 2.2680282663286552e-05,
"loss": 0.024,
"step": 3500
},
{
"epoch": 5.0503597122302155,
"grad_norm": 0.2839617133140564,
"learning_rate": 2.2403916741853364e-05,
"loss": 0.0192,
"step": 3510
},
{
"epoch": 5.0647482014388485,
"grad_norm": 0.24836570024490356,
"learning_rate": 2.2128757962500817e-05,
"loss": 0.0168,
"step": 3520
},
{
"epoch": 5.079136690647482,
"grad_norm": 0.2822698950767517,
"learning_rate": 2.1854818361565275e-05,
"loss": 0.0147,
"step": 3530
},
{
"epoch": 5.093525179856115,
"grad_norm": 0.24144606292247772,
"learning_rate": 2.1582109922052364e-05,
"loss": 0.0186,
"step": 3540
},
{
"epoch": 5.107913669064748,
"grad_norm": 0.1764133721590042,
"learning_rate": 2.1310644573112635e-05,
"loss": 0.0188,
"step": 3550
},
{
"epoch": 5.122302158273381,
"grad_norm": 0.25632283091545105,
"learning_rate": 2.1040434189519924e-05,
"loss": 0.0169,
"step": 3560
},
{
"epoch": 5.136690647482014,
"grad_norm": 0.3180352747440338,
"learning_rate": 2.0771490591151733e-05,
"loss": 0.017,
"step": 3570
},
{
"epoch": 5.151079136690647,
"grad_norm": 0.1721346378326416,
"learning_rate": 2.0503825542472317e-05,
"loss": 0.0159,
"step": 3580
},
{
"epoch": 5.16546762589928,
"grad_norm": 0.2239348292350769,
"learning_rate": 2.023745075201805e-05,
"loss": 0.0181,
"step": 3590
},
{
"epoch": 5.179856115107913,
"grad_norm": 0.20928189158439636,
"learning_rate": 1.9972377871885157e-05,
"loss": 0.0156,
"step": 3600
},
{
"epoch": 5.194244604316546,
"grad_norm": 0.3971610963344574,
"learning_rate": 1.970861849722017e-05,
"loss": 0.0149,
"step": 3610
},
{
"epoch": 5.2086330935251794,
"grad_norm": 0.19644194841384888,
"learning_rate": 1.9446184165712587e-05,
"loss": 0.0143,
"step": 3620
},
{
"epoch": 5.223021582733813,
"grad_norm": 0.42992860078811646,
"learning_rate": 1.9185086357090214e-05,
"loss": 0.019,
"step": 3630
},
{
"epoch": 5.237410071942446,
"grad_norm": 0.13897356390953064,
"learning_rate": 1.8925336492617057e-05,
"loss": 0.0138,
"step": 3640
},
{
"epoch": 5.251798561151079,
"grad_norm": 0.21071375906467438,
"learning_rate": 1.8666945934593666e-05,
"loss": 0.0159,
"step": 3650
},
{
"epoch": 5.266187050359712,
"grad_norm": 0.18721044063568115,
"learning_rate": 1.8409925985860126e-05,
"loss": 0.0152,
"step": 3660
},
{
"epoch": 5.280575539568345,
"grad_norm": 0.3037004768848419,
"learning_rate": 1.8154287889301603e-05,
"loss": 0.0147,
"step": 3670
},
{
"epoch": 5.294964028776978,
"grad_norm": 0.2435615211725235,
"learning_rate": 1.7900042827356612e-05,
"loss": 0.0153,
"step": 3680
},
{
"epoch": 5.309352517985611,
"grad_norm": 0.21572694182395935,
"learning_rate": 1.76472019215278e-05,
"loss": 0.0193,
"step": 3690
},
{
"epoch": 5.323741007194244,
"grad_norm": 0.22225253283977509,
"learning_rate": 1.739577623189545e-05,
"loss": 0.0178,
"step": 3700
},
{
"epoch": 5.338129496402877,
"grad_norm": 0.24780559539794922,
"learning_rate": 1.7145776756633768e-05,
"loss": 0.017,
"step": 3710
},
{
"epoch": 5.35251798561151,
"grad_norm": 0.3017253577709198,
"learning_rate": 1.6897214431529646e-05,
"loss": 0.0218,
"step": 3720
},
{
"epoch": 5.366906474820144,
"grad_norm": 0.23846878111362457,
"learning_rate": 1.6650100129504475e-05,
"loss": 0.0147,
"step": 3730
},
{
"epoch": 5.381294964028777,
"grad_norm": 0.14102913439273834,
"learning_rate": 1.6404444660138335e-05,
"loss": 0.0174,
"step": 3740
},
{
"epoch": 5.39568345323741,
"grad_norm": 0.284797728061676,
"learning_rate": 1.616025876919725e-05,
"loss": 0.0211,
"step": 3750
},
{
"epoch": 5.410071942446043,
"grad_norm": 0.27457743883132935,
"learning_rate": 1.5917553138163172e-05,
"loss": 0.0165,
"step": 3760
},
{
"epoch": 5.424460431654676,
"grad_norm": 0.198410302400589,
"learning_rate": 1.5676338383766632e-05,
"loss": 0.022,
"step": 3770
},
{
"epoch": 5.438848920863309,
"grad_norm": 0.21444785594940186,
"learning_rate": 1.5436625057522447e-05,
"loss": 0.017,
"step": 3780
},
{
"epoch": 5.453237410071942,
"grad_norm": 0.21764877438545227,
"learning_rate": 1.519842364526804e-05,
"loss": 0.0127,
"step": 3790
},
{
"epoch": 5.467625899280575,
"grad_norm": 0.2532919645309448,
"learning_rate": 1.4961744566704855e-05,
"loss": 0.0155,
"step": 3800
},
{
"epoch": 5.482014388489208,
"grad_norm": 0.11257531493902206,
"learning_rate": 1.4726598174942551e-05,
"loss": 0.0161,
"step": 3810
},
{
"epoch": 5.496402877697841,
"grad_norm": 0.22773367166519165,
"learning_rate": 1.4492994756046035e-05,
"loss": 0.0182,
"step": 3820
},
{
"epoch": 5.510791366906475,
"grad_norm": 0.23635350167751312,
"learning_rate": 1.4260944528585645e-05,
"loss": 0.0178,
"step": 3830
},
{
"epoch": 5.525179856115108,
"grad_norm": 0.3650851845741272,
"learning_rate": 1.4030457643190048e-05,
"loss": 0.0121,
"step": 3840
},
{
"epoch": 5.539568345323741,
"grad_norm": 0.19080981612205505,
"learning_rate": 1.3801544182102311e-05,
"loss": 0.0141,
"step": 3850
},
{
"epoch": 5.553956834532374,
"grad_norm": 0.27812474966049194,
"learning_rate": 1.3574214158738763e-05,
"loss": 0.0152,
"step": 3860
},
{
"epoch": 5.568345323741007,
"grad_norm": 0.1835435926914215,
"learning_rate": 1.3348477517251101e-05,
"loss": 0.0168,
"step": 3870
},
{
"epoch": 5.58273381294964,
"grad_norm": 0.2268984466791153,
"learning_rate": 1.312434413209131e-05,
"loss": 0.0161,
"step": 3880
},
{
"epoch": 5.597122302158273,
"grad_norm": 0.23231583833694458,
"learning_rate": 1.2901823807579727e-05,
"loss": 0.0116,
"step": 3890
},
{
"epoch": 5.611510791366906,
"grad_norm": 0.2471722662448883,
"learning_rate": 1.2680926277476245e-05,
"loss": 0.0186,
"step": 3900
},
{
"epoch": 5.625899280575539,
"grad_norm": 0.20653215050697327,
"learning_rate": 1.2461661204554397e-05,
"loss": 0.0147,
"step": 3910
},
{
"epoch": 5.640287769784173,
"grad_norm": 0.2517624795436859,
"learning_rate": 1.2244038180178835e-05,
"loss": 0.0169,
"step": 3920
},
{
"epoch": 5.654676258992806,
"grad_norm": 0.2775188684463501,
"learning_rate": 1.2028066723885612e-05,
"loss": 0.0129,
"step": 3930
},
{
"epoch": 5.669064748201439,
"grad_norm": 0.20716650784015656,
"learning_rate": 1.1813756282965888e-05,
"loss": 0.0163,
"step": 3940
},
{
"epoch": 5.683453237410072,
"grad_norm": 0.2366328239440918,
"learning_rate": 1.1601116232052638e-05,
"loss": 0.0148,
"step": 3950
},
{
"epoch": 5.697841726618705,
"grad_norm": 0.23200540244579315,
"learning_rate": 1.1390155872710517e-05,
"loss": 0.0158,
"step": 3960
},
{
"epoch": 5.712230215827338,
"grad_norm": 0.2675670087337494,
"learning_rate": 1.1180884433029087e-05,
"loss": 0.0152,
"step": 3970
},
{
"epoch": 5.726618705035971,
"grad_norm": 0.3245616853237152,
"learning_rate": 1.097331106721904e-05,
"loss": 0.0162,
"step": 3980
},
{
"epoch": 5.741007194244604,
"grad_norm": 0.23680998384952545,
"learning_rate": 1.0767444855211862e-05,
"loss": 0.0141,
"step": 3990
},
{
"epoch": 5.755395683453237,
"grad_norm": 0.20226994156837463,
"learning_rate": 1.0563294802262558e-05,
"loss": 0.0171,
"step": 4000
},
{
"epoch": 5.76978417266187,
"grad_norm": 0.15609662234783173,
"learning_rate": 1.0360869838555809e-05,
"loss": 0.0143,
"step": 4010
},
{
"epoch": 5.784172661870503,
"grad_norm": 0.2188311517238617,
"learning_rate": 1.0160178818815313e-05,
"loss": 0.0141,
"step": 4020
},
{
"epoch": 5.798561151079137,
"grad_norm": 0.25170716643333435,
"learning_rate": 9.961230521916387e-06,
"loss": 0.015,
"step": 4030
},
{
"epoch": 5.81294964028777,
"grad_norm": 0.25536754727363586,
"learning_rate": 9.764033650502074e-06,
"loss": 0.0162,
"step": 4040
},
{
"epoch": 5.827338129496403,
"grad_norm": 0.14454156160354614,
"learning_rate": 9.568596830602344e-06,
"loss": 0.0144,
"step": 4050
},
{
"epoch": 5.841726618705036,
"grad_norm": 0.2269306182861328,
"learning_rate": 9.37492861125681e-06,
"loss": 0.0146,
"step": 4060
},
{
"epoch": 5.856115107913669,
"grad_norm": 0.19262368977069855,
"learning_rate": 9.183037464140804e-06,
"loss": 0.015,
"step": 4070
},
{
"epoch": 5.870503597122302,
"grad_norm": 0.26915743947029114,
"learning_rate": 8.992931783194735e-06,
"loss": 0.0165,
"step": 4080
},
{
"epoch": 5.884892086330935,
"grad_norm": 0.15057407319545746,
"learning_rate": 8.80461988425696e-06,
"loss": 0.0156,
"step": 4090
},
{
"epoch": 5.899280575539568,
"grad_norm": 0.291503369808197,
"learning_rate": 8.618110004699974e-06,
"loss": 0.0159,
"step": 4100
},
{
"epoch": 5.913669064748201,
"grad_norm": 0.22148099541664124,
"learning_rate": 8.4334103030701e-06,
"loss": 0.0158,
"step": 4110
},
{
"epoch": 5.928057553956835,
"grad_norm": 0.2366560995578766,
"learning_rate": 8.25052885873066e-06,
"loss": 0.0135,
"step": 4120
},
{
"epoch": 5.942446043165468,
"grad_norm": 0.29662200808525085,
"learning_rate": 8.06947367150846e-06,
"loss": 0.0163,
"step": 4130
},
{
"epoch": 5.956834532374101,
"grad_norm": 0.17638437449932098,
"learning_rate": 7.890252661343938e-06,
"loss": 0.0145,
"step": 4140
},
{
"epoch": 5.971223021582734,
"grad_norm": 0.22053952515125275,
"learning_rate": 7.712873667944681e-06,
"loss": 0.0167,
"step": 4150
},
{
"epoch": 5.985611510791367,
"grad_norm": 0.14171098172664642,
"learning_rate": 7.537344450442469e-06,
"loss": 0.0085,
"step": 4160
},
{
"epoch": 6.0,
"grad_norm": 0.4520989954471588,
"learning_rate": 7.36367268705393e-06,
"loss": 0.0152,
"step": 4170
},
{
"epoch": 6.014388489208633,
"grad_norm": 0.20241068303585052,
"learning_rate": 7.1918659747446e-06,
"loss": 0.0175,
"step": 4180
},
{
"epoch": 6.028776978417266,
"grad_norm": 0.27699771523475647,
"learning_rate": 7.021931828896666e-06,
"loss": 0.0197,
"step": 4190
},
{
"epoch": 6.043165467625899,
"grad_norm": 0.14625288546085358,
"learning_rate": 6.8538776829801584e-06,
"loss": 0.016,
"step": 4200
},
{
"epoch": 6.057553956834532,
"grad_norm": 0.27774307131767273,
"learning_rate": 6.687710888227849e-06,
"loss": 0.0125,
"step": 4210
},
{
"epoch": 6.071942446043165,
"grad_norm": 0.11720925569534302,
"learning_rate": 6.5234387133136565e-06,
"loss": 0.0142,
"step": 4220
},
{
"epoch": 6.086330935251799,
"grad_norm": 0.1398768275976181,
"learning_rate": 6.361068344034665e-06,
"loss": 0.0115,
"step": 4230
},
{
"epoch": 6.100719424460432,
"grad_norm": 0.20375216007232666,
"learning_rate": 6.200606882996846e-06,
"loss": 0.0133,
"step": 4240
},
{
"epoch": 6.115107913669065,
"grad_norm": 0.2346504032611847,
"learning_rate": 6.042061349304312e-06,
"loss": 0.0165,
"step": 4250
},
{
"epoch": 6.129496402877698,
"grad_norm": 0.14118801057338715,
"learning_rate": 5.885438678252342e-06,
"loss": 0.0191,
"step": 4260
},
{
"epoch": 6.143884892086331,
"grad_norm": 0.2191145122051239,
"learning_rate": 5.730745721023939e-06,
"loss": 0.0124,
"step": 4270
},
{
"epoch": 6.158273381294964,
"grad_norm": 0.17568112909793854,
"learning_rate": 5.577989244390192e-06,
"loss": 0.01,
"step": 4280
},
{
"epoch": 6.172661870503597,
"grad_norm": 0.23272404074668884,
"learning_rate": 5.4271759304142635e-06,
"loss": 0.0128,
"step": 4290
},
{
"epoch": 6.18705035971223,
"grad_norm": 0.1852826327085495,
"learning_rate": 5.278312376159051e-06,
"loss": 0.0115,
"step": 4300
},
{
"epoch": 6.201438848920863,
"grad_norm": 0.13249315321445465,
"learning_rate": 5.1314050933986944e-06,
"loss": 0.0126,
"step": 4310
},
{
"epoch": 6.215827338129497,
"grad_norm": 0.14491771161556244,
"learning_rate": 4.986460508333634e-06,
"loss": 0.0123,
"step": 4320
},
{
"epoch": 6.23021582733813,
"grad_norm": 0.17501996457576752,
"learning_rate": 4.843484961309597e-06,
"loss": 0.0106,
"step": 4330
},
{
"epoch": 6.244604316546763,
"grad_norm": 0.16884444653987885,
"learning_rate": 4.702484706540161e-06,
"loss": 0.0106,
"step": 4340
},
{
"epoch": 6.258992805755396,
"grad_norm": 0.22701726853847504,
"learning_rate": 4.563465911833259e-06,
"loss": 0.0117,
"step": 4350
},
{
"epoch": 6.273381294964029,
"grad_norm": 0.2593272924423218,
"learning_rate": 4.426434658321344e-06,
"loss": 0.0119,
"step": 4360
},
{
"epoch": 6.287769784172662,
"grad_norm": 0.1882023960351944,
"learning_rate": 4.2913969401953466e-06,
"loss": 0.015,
"step": 4370
},
{
"epoch": 6.302158273381295,
"grad_norm": 0.1382090449333191,
"learning_rate": 4.15835866444253e-06,
"loss": 0.0094,
"step": 4380
},
{
"epoch": 6.316546762589928,
"grad_norm": 0.15261417627334595,
"learning_rate": 4.027325650588043e-06,
"loss": 0.0127,
"step": 4390
},
{
"epoch": 6.330935251798561,
"grad_norm": 0.27737364172935486,
"learning_rate": 3.898303630440419e-06,
"loss": 0.0165,
"step": 4400
},
{
"epoch": 6.345323741007194,
"grad_norm": 0.22711853682994843,
"learning_rate": 3.7712982478407877e-06,
"loss": 0.0151,
"step": 4410
},
{
"epoch": 6.359712230215827,
"grad_norm": 0.18514472246170044,
"learning_rate": 3.6463150584160053e-06,
"loss": 0.0118,
"step": 4420
},
{
"epoch": 6.374100719424461,
"grad_norm": 0.16549082100391388,
"learning_rate": 3.5233595293356957e-06,
"loss": 0.0147,
"step": 4430
},
{
"epoch": 6.388489208633094,
"grad_norm": 0.31707775592803955,
"learning_rate": 3.4024370390730033e-06,
"loss": 0.0129,
"step": 4440
},
{
"epoch": 6.402877697841727,
"grad_norm": 0.12159635871648788,
"learning_rate": 3.2835528771693992e-06,
"loss": 0.0168,
"step": 4450
},
{
"epoch": 6.41726618705036,
"grad_norm": 0.2887188196182251,
"learning_rate": 3.1667122440032505e-06,
"loss": 0.0158,
"step": 4460
},
{
"epoch": 6.431654676258993,
"grad_norm": 0.20540013909339905,
"learning_rate": 3.051920250562351e-06,
"loss": 0.0109,
"step": 4470
},
{
"epoch": 6.446043165467626,
"grad_norm": 0.16006773710250854,
"learning_rate": 2.939181918220385e-06,
"loss": 0.0122,
"step": 4480
},
{
"epoch": 6.460431654676259,
"grad_norm": 0.23204688727855682,
"learning_rate": 2.8285021785172226e-06,
"loss": 0.0122,
"step": 4490
},
{
"epoch": 6.474820143884892,
"grad_norm": 0.12096382677555084,
"learning_rate": 2.7198858729432288e-06,
"loss": 0.0137,
"step": 4500
},
{
"epoch": 6.489208633093525,
"grad_norm": 0.34355470538139343,
"learning_rate": 2.6133377527274905e-06,
"loss": 0.0151,
"step": 4510
},
{
"epoch": 6.503597122302159,
"grad_norm": 0.22738580405712128,
"learning_rate": 2.5088624786299366e-06,
"loss": 0.0131,
"step": 4520
},
{
"epoch": 6.517985611510792,
"grad_norm": 0.13188156485557556,
"learning_rate": 2.406464620737531e-06,
"loss": 0.0152,
"step": 4530
},
{
"epoch": 6.532374100719425,
"grad_norm": 0.1573144793510437,
"learning_rate": 2.3061486582642734e-06,
"loss": 0.0114,
"step": 4540
},
{
"epoch": 6.546762589928058,
"grad_norm": 0.21495701372623444,
"learning_rate": 2.2079189793553667e-06,
"loss": 0.0129,
"step": 4550
},
{
"epoch": 6.561151079136691,
"grad_norm": 0.2134663611650467,
"learning_rate": 2.111779880895165e-06,
"loss": 0.0104,
"step": 4560
},
{
"epoch": 6.575539568345324,
"grad_norm": 0.1690753549337387,
"learning_rate": 2.01773556831929e-06,
"loss": 0.0137,
"step": 4570
},
{
"epoch": 6.589928057553957,
"grad_norm": 0.2286052703857422,
"learning_rate": 1.9257901554306513e-06,
"loss": 0.0125,
"step": 4580
},
{
"epoch": 6.60431654676259,
"grad_norm": 0.20296066999435425,
"learning_rate": 1.835947664219445e-06,
"loss": 0.0118,
"step": 4590
},
{
"epoch": 6.618705035971223,
"grad_norm": 0.09227417409420013,
"learning_rate": 1.748212024687307e-06,
"loss": 0.0094,
"step": 4600
},
{
"epoch": 6.633093525179856,
"grad_norm": 0.2194000482559204,
"learning_rate": 1.6625870746753147e-06,
"loss": 0.0156,
"step": 4610
},
{
"epoch": 6.647482014388489,
"grad_norm": 0.17536112666130066,
"learning_rate": 1.5790765596961853e-06,
"loss": 0.0122,
"step": 4620
},
{
"epoch": 6.661870503597123,
"grad_norm": 0.16536195576190948,
"learning_rate": 1.4976841327703717e-06,
"loss": 0.0142,
"step": 4630
},
{
"epoch": 6.676258992805756,
"grad_norm": 0.21622546017169952,
"learning_rate": 1.4184133542663014e-06,
"loss": 0.0109,
"step": 4640
},
{
"epoch": 6.690647482014389,
"grad_norm": 0.28809091448783875,
"learning_rate": 1.341267691744641e-06,
"loss": 0.0141,
"step": 4650
},
{
"epoch": 6.705035971223022,
"grad_norm": 0.3993728458881378,
"learning_rate": 1.2662505198065666e-06,
"loss": 0.0178,
"step": 4660
},
{
"epoch": 6.719424460431655,
"grad_norm": 0.1959897130727768,
"learning_rate": 1.193365119946216e-06,
"loss": 0.0134,
"step": 4670
},
{
"epoch": 6.733812949640288,
"grad_norm": 0.21339170634746552,
"learning_rate": 1.1226146804070859e-06,
"loss": 0.0113,
"step": 4680
},
{
"epoch": 6.748201438848921,
"grad_norm": 0.1273609697818756,
"learning_rate": 1.0540022960426111e-06,
"loss": 0.0124,
"step": 4690
},
{
"epoch": 6.762589928057554,
"grad_norm": 0.214540496468544,
"learning_rate": 9.875309681807443e-07,
"loss": 0.0149,
"step": 4700
},
{
"epoch": 6.7769784172661875,
"grad_norm": 0.15893138945102692,
"learning_rate": 9.232036044927061e-07,
"loss": 0.0125,
"step": 4710
},
{
"epoch": 6.7913669064748206,
"grad_norm": 0.20806622505187988,
"learning_rate": 8.610230188657919e-07,
"loss": 0.0145,
"step": 4720
},
{
"epoch": 6.805755395683454,
"grad_norm": 0.19652938842773438,
"learning_rate": 8.009919312802372e-07,
"loss": 0.0147,
"step": 4730
},
{
"epoch": 6.820143884892087,
"grad_norm": 0.08473528921604156,
"learning_rate": 7.431129676902904e-07,
"loss": 0.0116,
"step": 4740
},
{
"epoch": 6.83453237410072,
"grad_norm": 0.16946707665920258,
"learning_rate": 6.873886599093215e-07,
"loss": 0.0148,
"step": 4750
},
{
"epoch": 6.848920863309353,
"grad_norm": 0.1452234834432602,
"learning_rate": 6.338214454990776e-07,
"loss": 0.0109,
"step": 4760
},
{
"epoch": 6.863309352517986,
"grad_norm": 0.17169806361198425,
"learning_rate": 5.82413667663051e-07,
"loss": 0.0126,
"step": 4770
},
{
"epoch": 6.877697841726619,
"grad_norm": 0.1109093502163887,
"learning_rate": 5.331675751439725e-07,
"loss": 0.0137,
"step": 4780
},
{
"epoch": 6.892086330935252,
"grad_norm": 0.16190867125988007,
"learning_rate": 4.86085322125479e-07,
"loss": 0.0122,
"step": 4790
},
{
"epoch": 6.906474820143885,
"grad_norm": 0.1729852259159088,
"learning_rate": 4.411689681378284e-07,
"loss": 0.0112,
"step": 4800
},
{
"epoch": 6.920863309352518,
"grad_norm": 0.27040132880210876,
"learning_rate": 3.9842047796786466e-07,
"loss": 0.0154,
"step": 4810
},
{
"epoch": 6.935251798561151,
"grad_norm": 0.12134522199630737,
"learning_rate": 3.578417215730323e-07,
"loss": 0.013,
"step": 4820
},
{
"epoch": 6.9496402877697845,
"grad_norm": 0.1886853128671646,
"learning_rate": 3.1943447399958027e-07,
"loss": 0.0121,
"step": 4830
},
{
"epoch": 6.9640287769784175,
"grad_norm": 0.29213500022888184,
"learning_rate": 2.8320041530495724e-07,
"loss": 0.0195,
"step": 4840
},
{
"epoch": 6.9784172661870505,
"grad_norm": 0.18699689209461212,
"learning_rate": 2.491411304842539e-07,
"loss": 0.013,
"step": 4850
},
{
"epoch": 6.9928057553956835,
"grad_norm": 0.19202850759029388,
"learning_rate": 2.1725810940094183e-07,
"loss": 0.0138,
"step": 4860
},
{
"epoch": 7.0071942446043165,
"grad_norm": 0.10336049646139145,
"learning_rate": 1.8755274672164202e-07,
"loss": 0.0127,
"step": 4870
},
{
"epoch": 7.0215827338129495,
"grad_norm": 0.1390511691570282,
"learning_rate": 1.600263418551573e-07,
"loss": 0.0134,
"step": 4880
},
{
"epoch": 7.0359712230215825,
"grad_norm": 0.1635769009590149,
"learning_rate": 1.346800988955954e-07,
"loss": 0.0115,
"step": 4890
},
{
"epoch": 7.0503597122302155,
"grad_norm": 0.1559794843196869,
"learning_rate": 1.1151512656975005e-07,
"loss": 0.0144,
"step": 4900
},
{
"epoch": 7.0647482014388485,
"grad_norm": 0.10500802844762802,
"learning_rate": 9.053243818853973e-08,
"loss": 0.0146,
"step": 4910
},
{
"epoch": 7.079136690647482,
"grad_norm": 0.11405812203884125,
"learning_rate": 7.173295160273763e-08,
"loss": 0.0114,
"step": 4920
},
{
"epoch": 7.093525179856115,
"grad_norm": 0.18420521914958954,
"learning_rate": 5.511748916279258e-08,
"loss": 0.0112,
"step": 4930
},
{
"epoch": 7.107913669064748,
"grad_norm": 0.16781188547611237,
"learning_rate": 4.068677768285234e-08,
"loss": 0.0137,
"step": 4940
},
{
"epoch": 7.122302158273381,
"grad_norm": 0.15142551064491272,
"learning_rate": 2.844144840898344e-08,
"loss": 0.0124,
"step": 4950
},
{
"epoch": 7.136690647482014,
"grad_norm": 0.15345831215381622,
"learning_rate": 1.8382036991559936e-08,
"loss": 0.012,
"step": 4960
},
{
"epoch": 7.151079136690647,
"grad_norm": 0.27668800950050354,
"learning_rate": 1.0508983461832156e-08,
"loss": 0.0132,
"step": 4970
},
{
"epoch": 7.16546762589928,
"grad_norm": 0.26003187894821167,
"learning_rate": 4.822632212653222e-09,
"loss": 0.0195,
"step": 4980
},
{
"epoch": 7.179856115107913,
"grad_norm": 0.31761491298675537,
"learning_rate": 1.3232319834632912e-09,
"loss": 0.0123,
"step": 4990
},
{
"epoch": 7.194244604316546,
"grad_norm": 0.26118502020835876,
"learning_rate": 1.0935849353854721e-11,
"loss": 0.0128,
"step": 5000
}
],
"logging_steps": 10,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 8,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}