Files
Pluto-Genesis-0.6B/last-checkpoint/trainer_state.json

1617 lines
38 KiB
JSON
Raw Normal View History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.8177777777777777,
"eval_steps": 500,
"global_step": 2250,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0008080808080808081,
"grad_norm": 104209.1015625,
"learning_rate": 0.0,
"loss": 0.9057,
"step": 1
},
{
"epoch": 0.00808080808080808,
"grad_norm": 84569.046875,
"learning_rate": 6.060606060606061e-06,
"loss": 0.8345,
"step": 10
},
{
"epoch": 0.01616161616161616,
"grad_norm": 62226.0625,
"learning_rate": 1.2794612794612796e-05,
"loss": 0.7542,
"step": 20
},
{
"epoch": 0.024242424242424242,
"grad_norm": 74986.484375,
"learning_rate": 1.952861952861953e-05,
"loss": 0.5801,
"step": 30
},
{
"epoch": 0.03232323232323232,
"grad_norm": 42189.6953125,
"learning_rate": 2.6262626262626268e-05,
"loss": 0.5045,
"step": 40
},
{
"epoch": 0.04040404040404041,
"grad_norm": 159301.484375,
"learning_rate": 3.2996632996632995e-05,
"loss": 0.6134,
"step": 50
},
{
"epoch": 0.048484848484848485,
"grad_norm": 18917.2421875,
"learning_rate": 3.973063973063973e-05,
"loss": 0.5375,
"step": 60
},
{
"epoch": 0.05656565656565657,
"grad_norm": 17363.826171875,
"learning_rate": 4.6464646464646464e-05,
"loss": 0.4655,
"step": 70
},
{
"epoch": 0.06464646464646465,
"grad_norm": 16705.484375,
"learning_rate": 5.31986531986532e-05,
"loss": 0.428,
"step": 80
},
{
"epoch": 0.07272727272727272,
"grad_norm": 25302.708984375,
"learning_rate": 5.993265993265994e-05,
"loss": 0.3975,
"step": 90
},
{
"epoch": 0.08080808080808081,
"grad_norm": 49937.75390625,
"learning_rate": 6.666666666666667e-05,
"loss": 0.5193,
"step": 100
},
{
"epoch": 0.08888888888888889,
"grad_norm": 15746.19921875,
"learning_rate": 7.340067340067341e-05,
"loss": 0.5242,
"step": 110
},
{
"epoch": 0.09696969696969697,
"grad_norm": 15647.705078125,
"learning_rate": 8.013468013468015e-05,
"loss": 0.4826,
"step": 120
},
{
"epoch": 0.10505050505050505,
"grad_norm": 15548.025390625,
"learning_rate": 8.686868686868688e-05,
"loss": 0.3897,
"step": 130
},
{
"epoch": 0.11313131313131314,
"grad_norm": 20209.09765625,
"learning_rate": 9.36026936026936e-05,
"loss": 0.3776,
"step": 140
},
{
"epoch": 0.12121212121212122,
"grad_norm": 46747.30859375,
"learning_rate": 0.00010033670033670034,
"loss": 0.5127,
"step": 150
},
{
"epoch": 0.1292929292929293,
"grad_norm": 15144.1748046875,
"learning_rate": 0.00010707070707070708,
"loss": 0.5024,
"step": 160
},
{
"epoch": 0.13737373737373737,
"grad_norm": 14970.73046875,
"learning_rate": 0.0001138047138047138,
"loss": 0.4434,
"step": 170
},
{
"epoch": 0.14545454545454545,
"grad_norm": 15086.0517578125,
"learning_rate": 0.00012053872053872056,
"loss": 0.4024,
"step": 180
},
{
"epoch": 0.15353535353535352,
"grad_norm": 17098.69140625,
"learning_rate": 0.00012727272727272728,
"loss": 0.4012,
"step": 190
},
{
"epoch": 0.16161616161616163,
"grad_norm": 41249.6875,
"learning_rate": 0.000134006734006734,
"loss": 0.5027,
"step": 200
},
{
"epoch": 0.1696969696969697,
"grad_norm": 12614.8740234375,
"learning_rate": 0.00014074074074074076,
"loss": 0.4935,
"step": 210
},
{
"epoch": 0.17777777777777778,
"grad_norm": 12711.9716796875,
"learning_rate": 0.00014747474747474747,
"loss": 0.4585,
"step": 220
},
{
"epoch": 0.18585858585858586,
"grad_norm": 13021.1865234375,
"learning_rate": 0.00015420875420875423,
"loss": 0.3589,
"step": 230
},
{
"epoch": 0.19393939393939394,
"grad_norm": 17581.67578125,
"learning_rate": 0.00016094276094276094,
"loss": 0.37,
"step": 240
},
{
"epoch": 0.20202020202020202,
"grad_norm": 31699.576171875,
"learning_rate": 0.00016767676767676768,
"loss": 0.4719,
"step": 250
},
{
"epoch": 0.2101010101010101,
"grad_norm": 11855.990234375,
"learning_rate": 0.00017441077441077442,
"loss": 0.5021,
"step": 260
},
{
"epoch": 0.21818181818181817,
"grad_norm": 11781.5615234375,
"learning_rate": 0.00018114478114478116,
"loss": 0.4663,
"step": 270
},
{
"epoch": 0.22626262626262628,
"grad_norm": 13736.8623046875,
"learning_rate": 0.0001878787878787879,
"loss": 0.376,
"step": 280
},
{
"epoch": 0.23434343434343435,
"grad_norm": 16363.1669921875,
"learning_rate": 0.00019461279461279463,
"loss": 0.4161,
"step": 290
},
{
"epoch": 0.24242424242424243,
"grad_norm": 37522.68359375,
"learning_rate": 0.00019999958426654012,
"loss": 0.4823,
"step": 300
},
{
"epoch": 0.2505050505050505,
"grad_norm": 11292.099609375,
"learning_rate": 0.00019998503395839373,
"loss": 0.5107,
"step": 310
},
{
"epoch": 0.2585858585858586,
"grad_norm": 11292.4189453125,
"learning_rate": 0.00019994970043381107,
"loss": 0.4438,
"step": 320
},
{
"epoch": 0.26666666666666666,
"grad_norm": 11470.115234375,
"learning_rate": 0.00019989359103733426,
"loss": 0.347,
"step": 330
},
{
"epoch": 0.27474747474747474,
"grad_norm": 16092.947265625,
"learning_rate": 0.00019981671743204608,
"loss": 0.3773,
"step": 340
},
{
"epoch": 0.2828282828282828,
"grad_norm": 32128.833984375,
"learning_rate": 0.00019971909559714584,
"loss": 0.4612,
"step": 350
},
{
"epoch": 0.2909090909090909,
"grad_norm": 11273.8564453125,
"learning_rate": 0.0001996007458246276,
"loss": 0.5035,
"step": 360
},
{
"epoch": 0.298989898989899,
"grad_norm": 11598.69140625,
"learning_rate": 0.0001994616927150625,
"loss": 0.4672,
"step": 370
},
{
"epoch": 0.30707070707070705,
"grad_norm": 13716.49609375,
"learning_rate": 0.00019930196517248505,
"loss": 0.3723,
"step": 380
},
{
"epoch": 0.3151515151515151,
"grad_norm": 14737.451171875,
"learning_rate": 0.00019912159639838507,
"loss": 0.3764,
"step": 390
},
{
"epoch": 0.32323232323232326,
"grad_norm": 31768.28515625,
"learning_rate": 0.00019892062388480634,
"loss": 0.4709,
"step": 400
},
{
"epoch": 0.33131313131313134,
"grad_norm": 11241.5498046875,
"learning_rate": 0.0001986990894065534,
"loss": 0.5008,
"step": 410
},
{
"epoch": 0.3393939393939394,
"grad_norm": 10874.2451171875,
"learning_rate": 0.00019845703901250801,
"loss": 0.4413,
"step": 420
},
{
"epoch": 0.3474747474747475,
"grad_norm": 13119.830078125,
"learning_rate": 0.0001981945230160574,
"loss": 0.3699,
"step": 430
},
{
"epoch": 0.35555555555555557,
"grad_norm": 13902.9228515625,
"learning_rate": 0.00019791159598463583,
"loss": 0.4139,
"step": 440
},
{
"epoch": 0.36363636363636365,
"grad_norm": 29348.953125,
"learning_rate": 0.00019760831672838215,
"loss": 0.4733,
"step": 450
},
{
"epoch": 0.3717171717171717,
"grad_norm": 10959.291015625,
"learning_rate": 0.00019728474828791535,
"loss": 0.4805,
"step": 460
},
{
"epoch": 0.3797979797979798,
"grad_norm": 11591.8837890625,
"learning_rate": 0.0001969409579212305,
"loss": 0.4246,
"step": 470
},
{
"epoch": 0.3878787878787879,
"grad_norm": 11706.6015625,
"learning_rate": 0.00019657701708971858,
"loss": 0.3719,
"step": 480
},
{
"epoch": 0.39595959595959596,
"grad_norm": 12885.5693359375,
"learning_rate": 0.00019619300144331213,
"loss": 0.3801,
"step": 490
},
{
"epoch": 0.40404040404040403,
"grad_norm": 31594.91796875,
"learning_rate": 0.00019578899080476036,
"loss": 0.4344,
"step": 500
},
{
"epoch": 0.4121212121212121,
"grad_norm": 10296.2197265625,
"learning_rate": 0.00019536506915303705,
"loss": 0.4824,
"step": 510
},
{
"epoch": 0.4202020202020202,
"grad_norm": 10842.8076171875,
"learning_rate": 0.0001949213246058843,
"loss": 0.4196,
"step": 520
},
{
"epoch": 0.42828282828282827,
"grad_norm": 13196.046875,
"learning_rate": 0.00019445784940149623,
"loss": 0.3769,
"step": 530
},
{
"epoch": 0.43636363636363634,
"grad_norm": 12825.0693359375,
"learning_rate": 0.00019397473987934588,
"loss": 0.3452,
"step": 540
},
{
"epoch": 0.4444444444444444,
"grad_norm": 33373.62109375,
"learning_rate": 0.00019347209646015988,
"loss": 0.4759,
"step": 550
},
{
"epoch": 0.45252525252525255,
"grad_norm": 10860.796875,
"learning_rate": 0.00019295002362504455,
"loss": 0.4838,
"step": 560
},
{
"epoch": 0.46060606060606063,
"grad_norm": 11779.4755859375,
"learning_rate": 0.0001924086298937684,
"loss": 0.4058,
"step": 570
},
{
"epoch": 0.4686868686868687,
"grad_norm": 12114.1923828125,
"learning_rate": 0.00019184802780220446,
"loss": 0.3771,
"step": 580
},
{
"epoch": 0.4767676767676768,
"grad_norm": 14090.4404296875,
"learning_rate": 0.00019126833387893852,
"loss": 0.3752,
"step": 590
},
{
"epoch": 0.48484848484848486,
"grad_norm": 36403.37890625,
"learning_rate": 0.00019066966862104696,
"loss": 0.451,
"step": 600
},
{
"epoch": 0.49292929292929294,
"grad_norm": 10452.39453125,
"learning_rate": 0.00019005215646904984,
"loss": 0.5219,
"step": 610
},
{
"epoch": 0.501010101010101,
"grad_norm": 11547.8349609375,
"learning_rate": 0.0001894159257810444,
"loss": 0.4402,
"step": 620
},
{
"epoch": 0.509090909090909,
"grad_norm": 11550.25,
"learning_rate": 0.00018876110880602397,
"loss": 0.3454,
"step": 630
},
{
"epoch": 0.5171717171717172,
"grad_norm": 15633.4873046875,
"learning_rate": 0.00018808784165638836,
"loss": 0.3719,
"step": 640
},
{
"epoch": 0.5252525252525253,
"grad_norm": 41919.3125,
"learning_rate": 0.00018739626427965113,
"loss": 0.4373,
"step": 650
},
{
"epoch": 0.5333333333333333,
"grad_norm": 10477.9384765625,
"learning_rate": 0.00018668652042934947,
"loss": 0.5364,
"step": 660
},
{
"epoch": 0.5414141414141415,
"grad_norm": 10492.9287109375,
"learning_rate": 0.0001859587576351633,
"loss": 0.441,
"step": 670
},
{
"epoch": 0.5494949494949495,
"grad_norm": 11572.525390625,
"learning_rate": 0.0001852131271722492,
"loss": 0.3719,
"step": 680
},
{
"epoch": 0.5575757575757576,
"grad_norm": 14025.73828125,
"learning_rate": 0.00018444978402979583,
"loss": 0.3539,
"step": 690
},
{
"epoch": 0.5656565656565656,
"grad_norm": 26899.578125,
"learning_rate": 0.0001836688868788074,
"loss": 0.4347,
"step": 700
},
{
"epoch": 0.5737373737373738,
"grad_norm": 10082.9345703125,
"learning_rate": 0.0001828705980391218,
"loss": 0.4999,
"step": 710
},
{
"epoch": 0.5818181818181818,
"grad_norm": 11851.2138671875,
"learning_rate": 0.00018205508344567025,
"loss": 0.4459,
"step": 720
},
{
"epoch": 0.5898989898989899,
"grad_norm": 11006.078125,
"learning_rate": 0.00018122251261398553,
"loss": 0.3775,
"step": 730
},
{
"epoch": 0.597979797979798,
"grad_norm": 13371.255859375,
"learning_rate": 0.000180373058604966,
"loss": 0.3625,
"step": 740
},
{
"epoch": 0.6060606060606061,
"grad_norm": 30220.501953125,
"learning_rate": 0.00017950689798890238,
"loss": 0.466,
"step": 750
},
{
"epoch": 0.6141414141414141,
"grad_norm": 9873.701171875,
"learning_rate": 0.00017862421080877555,
"loss": 0.4857,
"step": 760
},
{
"epoch": 0.6222222222222222,
"grad_norm": 11490.2490234375,
"learning_rate": 0.00017772518054283206,
"loss": 0.4424,
"step": 770
},
{
"epoch": 0.6303030303030303,
"grad_norm": 12206.841796875,
"learning_rate": 0.00017680999406644572,
"loss": 0.3741,
"step": 780
},
{
"epoch": 0.6383838383838384,
"grad_norm": 14496.12109375,
"learning_rate": 0.00017587884161327316,
"loss": 0.3856,
"step": 790
},
{
"epoch": 0.6464646464646465,
"grad_norm": 29068.046875,
"learning_rate": 0.0001749319167357112,
"loss": 0.4601,
"step": 800
},
{
"epoch": 0.6545454545454545,
"grad_norm": 9983.5947265625,
"learning_rate": 0.00017396941626466448,
"loss": 0.4729,
"step": 810
},
{
"epoch": 0.6626262626262627,
"grad_norm": 10951.2568359375,
"learning_rate": 0.00017299154026863137,
"loss": 0.4167,
"step": 820
},
{
"epoch": 0.6707070707070707,
"grad_norm": 11942.599609375,
"learning_rate": 0.00017199849201211728,
"loss": 0.3529,
"step": 830
},
{
"epoch": 0.6787878787878788,
"grad_norm": 13497.7216796875,
"learning_rate": 0.00017099047791338333,
"loss": 0.3538,
"step": 840
},
{
"epoch": 0.6868686868686869,
"grad_norm": 30288.41015625,
"learning_rate": 0.00016996770750153957,
"loss": 0.4399,
"step": 850
},
{
"epoch": 0.694949494949495,
"grad_norm": 10111.71875,
"learning_rate": 0.0001689303933729916,
"loss": 0.4695,
"step": 860
},
{
"epoch": 0.703030303030303,
"grad_norm": 10805.046875,
"learning_rate": 0.0001678787511472496,
"loss": 0.4167,
"step": 870
},
{
"epoch": 0.7111111111111111,
"grad_norm": 11908.6494140625,
"learning_rate": 0.00016681299942210886,
"loss": 0.3479,
"step": 880
},
{
"epoch": 0.7191919191919192,
"grad_norm": 14219.2734375,
"learning_rate": 0.00016573335972821157,
"loss": 0.3743,
"step": 890
},
{
"epoch": 0.7272727272727273,
"grad_norm": 28591.095703125,
"learning_rate": 0.00016464005648299834,
"loss": 0.4201,
"step": 900
},
{
"epoch": 0.7353535353535353,
"grad_norm": 9108.44140625,
"learning_rate": 0.00016353331694406045,
"loss": 0.4607,
"step": 910
},
{
"epoch": 0.7434343434343434,
"grad_norm": 11207.064453125,
"learning_rate": 0.0001624133711619011,
"loss": 0.412,
"step": 920
},
{
"epoch": 0.7515151515151515,
"grad_norm": 11462.8896484375,
"learning_rate": 0.00016128045193211649,
"loss": 0.3511,
"step": 930
},
{
"epoch": 0.7595959595959596,
"grad_norm": 13714.751953125,
"learning_rate": 0.00016013479474700603,
"loss": 0.3561,
"step": 940
},
{
"epoch": 0.7676767676767676,
"grad_norm": 27072.818359375,
"learning_rate": 0.00015897663774662227,
"loss": 0.4284,
"step": 950
},
{
"epoch": 0.7757575757575758,
"grad_norm": 10834.15234375,
"learning_rate": 0.0001578062216692702,
"loss": 0.4669,
"step": 960
},
{
"epoch": 0.7838383838383839,
"grad_norm": 11001.390625,
"learning_rate": 0.00015662378980146687,
"loss": 0.4288,
"step": 970
},
{
"epoch": 0.7919191919191919,
"grad_norm": 13222.6533203125,
"learning_rate": 0.00015542958792737067,
"loss": 0.335,
"step": 980
},
{
"epoch": 0.8,
"grad_norm": 13650.6015625,
"learning_rate": 0.00015422386427769205,
"loss": 0.3506,
"step": 990
},
{
"epoch": 0.8080808080808081,
"grad_norm": 30739.54296875,
"learning_rate": 0.0001530068694780955,
"loss": 0.4666,
"step": 1000
},
{
"epoch": 0.8161616161616162,
"grad_norm": 10002.2568359375,
"learning_rate": 0.0001517788564971037,
"loss": 0.4789,
"step": 1010
},
{
"epoch": 0.8242424242424242,
"grad_norm": 10963.697265625,
"learning_rate": 0.00015054008059351443,
"loss": 0.3991,
"step": 1020
},
{
"epoch": 0.8323232323232324,
"grad_norm": 11823.1181640625,
"learning_rate": 0.00014929079926334182,
"loss": 0.3529,
"step": 1030
},
{
"epoch": 0.8404040404040404,
"grad_norm": 15193.326171875,
"learning_rate": 0.0001480312721862924,
"loss": 0.3713,
"step": 1040
},
{
"epoch": 0.8484848484848485,
"grad_norm": 35250.9921875,
"learning_rate": 0.00014676176117178725,
"loss": 0.4183,
"step": 1050
},
{
"epoch": 0.8565656565656565,
"grad_norm": 10286.982421875,
"learning_rate": 0.0001454825301045412,
"loss": 0.509,
"step": 1060
},
{
"epoch": 0.8646464646464647,
"grad_norm": 10920.2568359375,
"learning_rate": 0.0001441938448897112,
"loss": 0.4191,
"step": 1070
},
{
"epoch": 0.8727272727272727,
"grad_norm": 11989.990234375,
"learning_rate": 0.00014289597339762423,
"loss": 0.3654,
"step": 1080
},
{
"epoch": 0.8808080808080808,
"grad_norm": 14704.974609375,
"learning_rate": 0.00014158918540809685,
"loss": 0.3631,
"step": 1090
},
{
"epoch": 0.8888888888888888,
"grad_norm": 26379.830078125,
"learning_rate": 0.00014027375255435808,
"loss": 0.4477,
"step": 1100
},
{
"epoch": 0.896969696969697,
"grad_norm": 10254.0341796875,
"learning_rate": 0.00013894994826658665,
"loss": 0.4664,
"step": 1110
},
{
"epoch": 0.9050505050505051,
"grad_norm": 11967.30078125,
"learning_rate": 0.00013761804771507497,
"loss": 0.411,
"step": 1120
},
{
"epoch": 0.9131313131313131,
"grad_norm": 10433.6240234375,
"learning_rate": 0.00013627832775303138,
"loss": 0.3615,
"step": 1130
},
{
"epoch": 0.9212121212121213,
"grad_norm": 15648.314453125,
"learning_rate": 0.0001349310668590324,
"loss": 0.3823,
"step": 1140
},
{
"epoch": 0.9292929292929293,
"grad_norm": 25233.39453125,
"learning_rate": 0.0001335765450791374,
"loss": 0.4271,
"step": 1150
},
{
"epoch": 0.9373737373737374,
"grad_norm": 10123.2548828125,
"learning_rate": 0.0001322150439686773,
"loss": 0.4582,
"step": 1160
},
{
"epoch": 0.9454545454545454,
"grad_norm": 10851.8525390625,
"learning_rate": 0.00013084684653372943,
"loss": 0.4283,
"step": 1170
},
{
"epoch": 0.9535353535353536,
"grad_norm": 11651.4482421875,
"learning_rate": 0.00012947223717229113,
"loss": 0.3522,
"step": 1180
},
{
"epoch": 0.9616161616161616,
"grad_norm": 14655.8388671875,
"learning_rate": 0.00012809150161516376,
"loss": 0.3468,
"step": 1190
},
{
"epoch": 0.9696969696969697,
"grad_norm": 28266.390625,
"learning_rate": 0.00012670492686655966,
"loss": 0.4349,
"step": 1200
},
{
"epoch": 0.9777777777777777,
"grad_norm": 10841.1953125,
"learning_rate": 0.00012531280114444474,
"loss": 0.4611,
"step": 1210
},
{
"epoch": 0.9858585858585859,
"grad_norm": 10848.00390625,
"learning_rate": 0.0001239154138206284,
"loss": 0.361,
"step": 1220
},
{
"epoch": 0.9939393939393939,
"grad_norm": 14088.783203125,
"learning_rate": 0.00012251305536061378,
"loss": 0.3462,
"step": 1230
},
{
"epoch": 1.0016161616161616,
"grad_norm": 9361.748046875,
"learning_rate": 0.00012110601726322071,
"loss": 0.4437,
"step": 1240
},
{
"epoch": 1.0096969696969698,
"grad_norm": 10145.392578125,
"learning_rate": 0.00011969459199999366,
"loss": 0.4057,
"step": 1250
},
{
"epoch": 1.0177777777777777,
"grad_norm": 12363.9423828125,
"learning_rate": 0.00011827907295440788,
"loss": 0.3781,
"step": 1260
},
{
"epoch": 1.0258585858585858,
"grad_norm": 11812.9619140625,
"learning_rate": 0.0001168597543608855,
"loss": 0.3148,
"step": 1270
},
{
"epoch": 1.033939393939394,
"grad_norm": 16864.353515625,
"learning_rate": 0.00011543693124363522,
"loss": 0.3296,
"step": 1280
},
{
"epoch": 1.042020202020202,
"grad_norm": 10625.6259765625,
"learning_rate": 0.00011401089935532745,
"loss": 0.4377,
"step": 1290
},
{
"epoch": 1.0501010101010102,
"grad_norm": 10747.6083984375,
"learning_rate": 0.00011258195511561827,
"loss": 0.4242,
"step": 1300
},
{
"epoch": 1.0581818181818181,
"grad_norm": 10821.7109375,
"learning_rate": 0.00011115039554953466,
"loss": 0.377,
"step": 1310
},
{
"epoch": 1.0662626262626262,
"grad_norm": 12382.2763671875,
"learning_rate": 0.00010971651822573392,
"loss": 0.3006,
"step": 1320
},
{
"epoch": 1.0743434343434344,
"grad_norm": 13722.70703125,
"learning_rate": 0.00010828062119465004,
"loss": 0.3544,
"step": 1330
},
{
"epoch": 1.0824242424242425,
"grad_norm": 10983.6015625,
"learning_rate": 0.00010684300292653994,
"loss": 0.4466,
"step": 1340
},
{
"epoch": 1.0905050505050504,
"grad_norm": 10921.0029296875,
"learning_rate": 0.00010540396224944255,
"loss": 0.4074,
"step": 1350
},
{
"epoch": 1.0985858585858586,
"grad_norm": 12036.0419921875,
"learning_rate": 0.0001039637982870634,
"loss": 0.3634,
"step": 1360
},
{
"epoch": 1.1066666666666667,
"grad_norm": 12630.732421875,
"learning_rate": 0.00010252281039659777,
"loss": 0.3223,
"step": 1370
},
{
"epoch": 1.1147474747474748,
"grad_norm": 16125.1552734375,
"learning_rate": 0.00010108129810650524,
"loss": 0.3141,
"step": 1380
},
{
"epoch": 1.1228282828282827,
"grad_norm": 10188.427734375,
"learning_rate": 9.96395610542489e-05,
"loss": 0.4401,
"step": 1390
},
{
"epoch": 1.1309090909090909,
"grad_norm": 10294.087890625,
"learning_rate": 9.819789892401164e-05,
"loss": 0.4285,
"step": 1400
},
{
"epoch": 1.138989898989899,
"grad_norm": 11054.04296875,
"learning_rate": 9.67566113844028e-05,
"loss": 0.3721,
"step": 1410
},
{
"epoch": 1.1470707070707071,
"grad_norm": 11459.974609375,
"learning_rate": 9.53159980261681e-05,
"loss": 0.3283,
"step": 1420
},
{
"epoch": 1.1551515151515153,
"grad_norm": 15478.9228515625,
"learning_rate": 9.387635829991562e-05,
"loss": 0.3227,
"step": 1430
},
{
"epoch": 1.1632323232323232,
"grad_norm": 9957.7587890625,
"learning_rate": 9.243799145387128e-05,
"loss": 0.4417,
"step": 1440
},
{
"epoch": 1.1713131313131313,
"grad_norm": 9921.2607421875,
"learning_rate": 9.100119647167575e-05,
"loss": 0.4225,
"step": 1450
},
{
"epoch": 1.1793939393939394,
"grad_norm": 12321.8955078125,
"learning_rate": 8.956627201023713e-05,
"loss": 0.3676,
"step": 1460
},
{
"epoch": 1.1874747474747476,
"grad_norm": 12257.5107421875,
"learning_rate": 8.813351633765082e-05,
"loss": 0.3097,
"step": 1470
},
{
"epoch": 1.1955555555555555,
"grad_norm": 16210.310546875,
"learning_rate": 8.670322727120073e-05,
"loss": 0.3377,
"step": 1480
},
{
"epoch": 1.2036363636363636,
"grad_norm": 12172.9990234375,
"learning_rate": 8.527570211545391e-05,
"loss": 0.4439,
"step": 1490
},
{
"epoch": 1.2117171717171717,
"grad_norm": 9363.861328125,
"learning_rate": 8.385123760046196e-05,
"loss": 0.4176,
"step": 1500
},
{
"epoch": 1.2197979797979799,
"grad_norm": 11533.712890625,
"learning_rate": 8.243012982008159e-05,
"loss": 0.3772,
"step": 1510
},
{
"epoch": 1.2278787878787878,
"grad_norm": 12586.962890625,
"learning_rate": 8.10126741704276e-05,
"loss": 0.2936,
"step": 1520
},
{
"epoch": 1.235959595959596,
"grad_norm": 18315.810546875,
"learning_rate": 7.959916528847116e-05,
"loss": 0.3284,
"step": 1530
},
{
"epoch": 1.244040404040404,
"grad_norm": 10804.5888671875,
"learning_rate": 7.81898969907952e-05,
"loss": 0.4056,
"step": 1540
},
{
"epoch": 1.2521212121212122,
"grad_norm": 10015.4306640625,
"learning_rate": 7.678516221252106e-05,
"loss": 0.4028,
"step": 1550
},
{
"epoch": 1.26020202020202,
"grad_norm": 11953.2783203125,
"learning_rate": 7.538525294641772e-05,
"loss": 0.3554,
"step": 1560
},
{
"epoch": 1.2682828282828282,
"grad_norm": 12607.8125,
"learning_rate": 7.39904601822076e-05,
"loss": 0.2623,
"step": 1570
},
{
"epoch": 1.2763636363636364,
"grad_norm": 14767.171875,
"learning_rate": 7.260107384608011e-05,
"loss": 0.3027,
"step": 1580
},
{
"epoch": 1.2844444444444445,
"grad_norm": 10422.837890625,
"learning_rate": 7.121738274042698e-05,
"loss": 0.3977,
"step": 1590
},
{
"epoch": 1.2925252525252526,
"grad_norm": 10670.5595703125,
"learning_rate": 6.983967448381061e-05,
"loss": 0.3809,
"step": 1600
},
{
"epoch": 1.3006060606060605,
"grad_norm": 11297.861328125,
"learning_rate": 6.846823545117876e-05,
"loss": 0.3838,
"step": 1610
},
{
"epoch": 1.3086868686868687,
"grad_norm": 12071.4716796875,
"learning_rate": 6.710335071433794e-05,
"loss": 0.2819,
"step": 1620
},
{
"epoch": 1.3167676767676768,
"grad_norm": 14401.484375,
"learning_rate": 6.574530398269709e-05,
"loss": 0.2986,
"step": 1630
},
{
"epoch": 1.3248484848484847,
"grad_norm": 10620.7119140625,
"learning_rate": 6.439437754429518e-05,
"loss": 0.4074,
"step": 1640
},
{
"epoch": 1.3329292929292929,
"grad_norm": 10562.8427734375,
"learning_rate": 6.305085220712354e-05,
"loss": 0.3818,
"step": 1650
},
{
"epoch": 1.341010101010101,
"grad_norm": 11316.8154296875,
"learning_rate": 6.171500724075657e-05,
"loss": 0.347,
"step": 1660
},
{
"epoch": 1.3490909090909091,
"grad_norm": 12836.2236328125,
"learning_rate": 6.0387120318301495e-05,
"loss": 0.2832,
"step": 1670
},
{
"epoch": 1.3571717171717173,
"grad_norm": 17111.091796875,
"learning_rate": 5.9067467458680545e-05,
"loss": 0.336,
"step": 1680
},
{
"epoch": 1.3652525252525254,
"grad_norm": 10938.0673828125,
"learning_rate": 5.7756322969256727e-05,
"loss": 0.385,
"step": 1690
},
{
"epoch": 1.3733333333333333,
"grad_norm": 10721.283203125,
"learning_rate": 5.6453959388815316e-05,
"loss": 0.3742,
"step": 1700
},
{
"epoch": 1.3814141414141414,
"grad_norm": 10335.0322265625,
"learning_rate": 5.516064743091313e-05,
"loss": 0.3389,
"step": 1710
},
{
"epoch": 1.3894949494949496,
"grad_norm": 14728.2900390625,
"learning_rate": 5.387665592760709e-05,
"loss": 0.3067,
"step": 1720
},
{
"epoch": 1.3975757575757575,
"grad_norm": 16770.470703125,
"learning_rate": 5.260225177357393e-05,
"loss": 0.2756,
"step": 1730
},
{
"epoch": 1.4056565656565656,
"grad_norm": 10549.6943359375,
"learning_rate": 5.133769987063256e-05,
"loss": 0.3576,
"step": 1740
},
{
"epoch": 1.4137373737373737,
"grad_norm": 10353.4306640625,
"learning_rate": 5.008326307268074e-05,
"loss": 0.3805,
"step": 1750
},
{
"epoch": 1.4218181818181819,
"grad_norm": 11408.3955078125,
"learning_rate": 4.883920213105741e-05,
"loss": 0.3474,
"step": 1760
},
{
"epoch": 1.42989898989899,
"grad_norm": 13688.7353515625,
"learning_rate": 4.760577564034202e-05,
"loss": 0.2768,
"step": 1770
},
{
"epoch": 1.437979797979798,
"grad_norm": 16796.970703125,
"learning_rate": 4.638323998460229e-05,
"loss": 0.2742,
"step": 1780
},
{
"epoch": 1.446060606060606,
"grad_norm": 10752.8583984375,
"learning_rate": 4.5171849284101275e-05,
"loss": 0.3919,
"step": 1790
},
{
"epoch": 1.4541414141414142,
"grad_norm": 10992.8544921875,
"learning_rate": 4.397185534247521e-05,
"loss": 0.3836,
"step": 1800
},
{
"epoch": 1.462222222222222,
"grad_norm": 12128.2666015625,
"learning_rate": 4.278350759439266e-05,
"loss": 0.3237,
"step": 1810
},
{
"epoch": 1.4703030303030302,
"grad_norm": 12960.751953125,
"learning_rate": 4.160705305370641e-05,
"loss": 0.2916,
"step": 1820
},
{
"epoch": 1.4783838383838384,
"grad_norm": 17203.375,
"learning_rate": 4.044273626210801e-05,
"loss": 0.2793,
"step": 1830
},
{
"epoch": 1.4864646464646465,
"grad_norm": 11018.6611328125,
"learning_rate": 3.929079923829703e-05,
"loss": 0.3863,
"step": 1840
},
{
"epoch": 1.4945454545454546,
"grad_norm": 10717.3359375,
"learning_rate": 3.815148142767371e-05,
"loss": 0.3995,
"step": 1850
},
{
"epoch": 1.5026262626262628,
"grad_norm": 10796.341796875,
"learning_rate": 3.70250196525675e-05,
"loss": 0.3491,
"step": 1860
},
{
"epoch": 1.5107070707070707,
"grad_norm": 13137.72265625,
"learning_rate": 3.591164806301019e-05,
"loss": 0.2664,
"step": 1870
},
{
"epoch": 1.5187878787878788,
"grad_norm": 16772.97265625,
"learning_rate": 3.481159808806471e-05,
"loss": 0.2843,
"step": 1880
},
{
"epoch": 1.5268686868686867,
"grad_norm": 11630.0029296875,
"learning_rate": 3.37250983877197e-05,
"loss": 0.3866,
"step": 1890
},
{
"epoch": 1.5349494949494948,
"grad_norm": 12221.330078125,
"learning_rate": 3.265237480535915e-05,
"loss": 0.4233,
"step": 1900
},
{
"epoch": 1.543030303030303,
"grad_norm": 12206.5517578125,
"learning_rate": 3.1593650320818335e-05,
"loss": 0.3498,
"step": 1910
},
{
"epoch": 1.551111111111111,
"grad_norm": 12800.09375,
"learning_rate": 3.054914500403396e-05,
"loss": 0.2765,
"step": 1920
},
{
"epoch": 1.5591919191919192,
"grad_norm": 16723.20703125,
"learning_rate": 2.9519075969300224e-05,
"loss": 0.2721,
"step": 1930
},
{
"epoch": 1.5672727272727274,
"grad_norm": 11790.9482421875,
"learning_rate": 2.8503657330138357e-05,
"loss": 0.3701,
"step": 1940
},
{
"epoch": 1.5753535353535355,
"grad_norm": 11674.5712890625,
"learning_rate": 2.7503100154790363e-05,
"loss": 0.3906,
"step": 1950
},
{
"epoch": 1.5834343434343434,
"grad_norm": 12388.21875,
"learning_rate": 2.6517612422345906e-05,
"loss": 0.3612,
"step": 1960
},
{
"epoch": 1.5915151515151515,
"grad_norm": 14171.6044921875,
"learning_rate": 2.5547398979510682e-05,
"loss": 0.2858,
"step": 1970
},
{
"epoch": 1.5995959595959595,
"grad_norm": 17654.046875,
"learning_rate": 2.4592661498026814e-05,
"loss": 0.2897,
"step": 1980
},
{
"epoch": 1.6076767676767676,
"grad_norm": 10777.1025390625,
"learning_rate": 2.3653598432752254e-05,
"loss": 0.393,
"step": 1990
},
{
"epoch": 1.6157575757575757,
"grad_norm": 12413.0361328125,
"learning_rate": 2.273040498040956e-05,
"loss": 0.3766,
"step": 2000
},
{
"epoch": 1.6238383838383839,
"grad_norm": 11548.69921875,
"learning_rate": 2.1823273039011584e-05,
"loss": 0.3526,
"step": 2010
},
{
"epoch": 1.631919191919192,
"grad_norm": 13187.0625,
"learning_rate": 2.0932391167972974e-05,
"loss": 0.288,
"step": 2020
},
{
"epoch": 1.6400000000000001,
"grad_norm": 17303.259765625,
"learning_rate": 2.005794454891554e-05,
"loss": 0.303,
"step": 2030
},
{
"epoch": 1.648080808080808,
"grad_norm": 9522.2451171875,
"learning_rate": 1.9200114947175753e-05,
"loss": 0.3714,
"step": 2040
},
{
"epoch": 1.6561616161616162,
"grad_norm": 11601.1962890625,
"learning_rate": 1.83590806740226e-05,
"loss": 0.3746,
"step": 2050
},
{
"epoch": 1.664242424242424,
"grad_norm": 10855.3818359375,
"learning_rate": 1.753501654959302e-05,
"loss": 0.3258,
"step": 2060
},
{
"epoch": 1.6723232323232322,
"grad_norm": 13471.251953125,
"learning_rate": 1.6728093866553287e-05,
"loss": 0.2771,
"step": 2070
},
{
"epoch": 1.6804040404040403,
"grad_norm": 16066.009765625,
"learning_rate": 1.5938480354493556e-05,
"loss": 0.2701,
"step": 2080
},
{
"epoch": 1.6884848484848485,
"grad_norm": 10615.1748046875,
"learning_rate": 1.5166340145063085e-05,
"loss": 0.3644,
"step": 2090
},
{
"epoch": 1.6965656565656566,
"grad_norm": 11259.359375,
"learning_rate": 1.441183373785321e-05,
"loss": 0.3572,
"step": 2100
},
{
"epoch": 1.7046464646464647,
"grad_norm": 10914.65625,
"learning_rate": 1.3675117967035444e-05,
"loss": 0.3295,
"step": 2110
},
{
"epoch": 1.7127272727272729,
"grad_norm": 13878.3046875,
"learning_rate": 1.2956345968761375e-05,
"loss": 0.2676,
"step": 2120
},
{
"epoch": 1.7208080808080808,
"grad_norm": 15814.9521484375,
"learning_rate": 1.2255667149331285e-05,
"loss": 0.2896,
"step": 2130
},
{
"epoch": 1.728888888888889,
"grad_norm": 8977.2587890625,
"learning_rate": 1.15732271541381e-05,
"loss": 0.3492,
"step": 2140
},
{
"epoch": 1.7369696969696968,
"grad_norm": 10733.3017578125,
"learning_rate": 1.090916783739302e-05,
"loss": 0.3667,
"step": 2150
},
{
"epoch": 1.745050505050505,
"grad_norm": 12199.140625,
"learning_rate": 1.0263627232639283e-05,
"loss": 0.3361,
"step": 2160
},
{
"epoch": 1.753131313131313,
"grad_norm": 11585.9658203125,
"learning_rate": 9.636739524060001e-06,
"loss": 0.2498,
"step": 2170
},
{
"epoch": 1.7612121212121212,
"grad_norm": 17156.82421875,
"learning_rate": 9.028635018586252e-06,
"loss": 0.2915,
"step": 2180
},
{
"epoch": 1.7692929292929294,
"grad_norm": 9348.0810546875,
"learning_rate": 8.439440118811027e-06,
"loss": 0.3564,
"step": 2190
},
{
"epoch": 1.7773737373737375,
"grad_norm": 10844.87109375,
"learning_rate": 7.869277296714828e-06,
"loss": 0.3632,
"step": 2200
},
{
"epoch": 1.7854545454545454,
"grad_norm": 11805.4248046875,
"learning_rate": 7.318265068208108e-06,
"loss": 0.3421,
"step": 2210
},
{
"epoch": 1.7935353535353535,
"grad_norm": 12437.6533203125,
"learning_rate": 6.786517968496287e-06,
"loss": 0.2627,
"step": 2220
},
{
"epoch": 1.8016161616161617,
"grad_norm": 16116.330078125,
"learning_rate": 6.274146528272085e-06,
"loss": 0.2891,
"step": 2230
},
{
"epoch": 1.8096969696969696,
"grad_norm": 10148.9169921875,
"learning_rate": 5.7812572507401795e-06,
"loss": 0.3845,
"step": 2240
},
{
"epoch": 1.8177777777777777,
"grad_norm": 10441.8017578125,
"learning_rate": 5.3079525894791925e-06,
"loss": 0.3749,
"step": 2250
}
],
"logging_steps": 10,
"max_steps": 2476,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 250,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.900007361179812e+17,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}