fasthypernet-x5 / checkpoint-3000 /trainer_state.json
aixk's picture
Upload folder using huggingface_hub
8126e84 verified
Raw
History Blame Contribute Delete
52 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.009656720819027879,
"eval_steps": 500,
"global_step": 3000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00013795315455754112,
"grad_norm": 53.4421272277832,
"learning_rate": 4.1999999999999995e-07,
"loss": 8.6569,
"step": 10
},
{
"epoch": 0.00027590630911508224,
"grad_norm": 57.65703582763672,
"learning_rate": 1.0199999999999998e-06,
"loss": 8.6391,
"step": 20
},
{
"epoch": 0.0004138594636726234,
"grad_norm": 93.79612731933594,
"learning_rate": 1.62e-06,
"loss": 8.6097,
"step": 30
},
{
"epoch": 0.0005518126182301645,
"grad_norm": 133.8379669189453,
"learning_rate": 2.22e-06,
"loss": 8.5995,
"step": 40
},
{
"epoch": 0.0006897657727877056,
"grad_norm": 99.24571228027344,
"learning_rate": 2.8199999999999997e-06,
"loss": 8.5699,
"step": 50
},
{
"epoch": 0.0008277189273452468,
"grad_norm": 155.49118041992188,
"learning_rate": 3.42e-06,
"loss": 8.5516,
"step": 60
},
{
"epoch": 0.0009656720819027878,
"grad_norm": 118.79692840576172,
"learning_rate": 4.02e-06,
"loss": 8.5331,
"step": 70
},
{
"epoch": 0.001103625236460329,
"grad_norm": 88.51729583740234,
"learning_rate": 4.62e-06,
"loss": 8.5212,
"step": 80
},
{
"epoch": 0.0012415783910178701,
"grad_norm": 117.01810455322266,
"learning_rate": 5.219999999999999e-06,
"loss": 8.5007,
"step": 90
},
{
"epoch": 0.0013795315455754113,
"grad_norm": 104.06542205810547,
"learning_rate": 5.819999999999999e-06,
"loss": 8.4948,
"step": 100
},
{
"epoch": 0.0015174847001329524,
"grad_norm": 103.02165985107422,
"learning_rate": 6.4199999999999995e-06,
"loss": 8.4871,
"step": 110
},
{
"epoch": 0.0016554378546904936,
"grad_norm": 119.96959686279297,
"learning_rate": 7.02e-06,
"loss": 8.4803,
"step": 120
},
{
"epoch": 0.0017933910092480345,
"grad_norm": 83.34615325927734,
"learning_rate": 7.619999999999999e-06,
"loss": 8.471,
"step": 130
},
{
"epoch": 0.0019313441638055756,
"grad_norm": 76.37083435058594,
"learning_rate": 8.22e-06,
"loss": 8.4655,
"step": 140
},
{
"epoch": 0.002069297318363117,
"grad_norm": 82.49830627441406,
"learning_rate": 8.819999999999999e-06,
"loss": 8.4708,
"step": 150
},
{
"epoch": 0.002207250472920658,
"grad_norm": 69.70299530029297,
"learning_rate": 9.419999999999998e-06,
"loss": 8.4659,
"step": 160
},
{
"epoch": 0.0023452036274781993,
"grad_norm": 103.55642700195312,
"learning_rate": 1.0019999999999999e-05,
"loss": 8.4555,
"step": 170
},
{
"epoch": 0.0024831567820357402,
"grad_norm": 112.4985122680664,
"learning_rate": 1.062e-05,
"loss": 8.4629,
"step": 180
},
{
"epoch": 0.002621109936593281,
"grad_norm": 78.48297882080078,
"learning_rate": 1.122e-05,
"loss": 8.4562,
"step": 190
},
{
"epoch": 0.0027590630911508225,
"grad_norm": 87.78710174560547,
"learning_rate": 1.1819999999999999e-05,
"loss": 8.451,
"step": 200
},
{
"epoch": 3.448828863938528e-05,
"grad_norm": 34.83774948120117,
"learning_rate": 1.2419999999999998e-05,
"loss": 8.4432,
"step": 210
},
{
"epoch": 6.897657727877056e-05,
"grad_norm": 68.05087280273438,
"learning_rate": 1.3019999999999999e-05,
"loss": 8.4542,
"step": 220
},
{
"epoch": 0.00010346486591815585,
"grad_norm": 46.86287307739258,
"learning_rate": 1.362e-05,
"loss": 8.4622,
"step": 230
},
{
"epoch": 0.00013795315455754112,
"grad_norm": 38.558746337890625,
"learning_rate": 1.4219999999999998e-05,
"loss": 8.4458,
"step": 240
},
{
"epoch": 0.0001724414431969264,
"grad_norm": 38.49115753173828,
"learning_rate": 1.4819999999999999e-05,
"loss": 8.4495,
"step": 250
},
{
"epoch": 0.0002069297318363117,
"grad_norm": 29.765907287597656,
"learning_rate": 1.5419999999999998e-05,
"loss": 8.464,
"step": 260
},
{
"epoch": 0.00024141802047569696,
"grad_norm": 48.3050537109375,
"learning_rate": 1.602e-05,
"loss": 8.4601,
"step": 270
},
{
"epoch": 0.00027590630911508224,
"grad_norm": 27.089998245239258,
"learning_rate": 1.6619999999999997e-05,
"loss": 8.4628,
"step": 280
},
{
"epoch": 0.00031039459775446753,
"grad_norm": 36.98786163330078,
"learning_rate": 1.7219999999999998e-05,
"loss": 8.4364,
"step": 290
},
{
"epoch": 0.0003448828863938528,
"grad_norm": 39.274784088134766,
"learning_rate": 1.782e-05,
"loss": 8.4314,
"step": 300
},
{
"epoch": 0.0003793711750332381,
"grad_norm": 29.303476333618164,
"learning_rate": 1.842e-05,
"loss": 8.4153,
"step": 310
},
{
"epoch": 0.0004138594636726234,
"grad_norm": 24.435274124145508,
"learning_rate": 1.9019999999999997e-05,
"loss": 8.4489,
"step": 320
},
{
"epoch": 0.0004483477523120086,
"grad_norm": 25.8426456451416,
"learning_rate": 1.962e-05,
"loss": 8.4297,
"step": 330
},
{
"epoch": 0.0004828360409513939,
"grad_norm": 24.164419174194336,
"learning_rate": 2.022e-05,
"loss": 8.4398,
"step": 340
},
{
"epoch": 0.0005173243295907793,
"grad_norm": 17.73716163635254,
"learning_rate": 2.082e-05,
"loss": 8.4333,
"step": 350
},
{
"epoch": 0.0005518126182301645,
"grad_norm": 34.02337646484375,
"learning_rate": 2.1419999999999998e-05,
"loss": 8.4488,
"step": 360
},
{
"epoch": 0.0005863009068695498,
"grad_norm": 26.535123825073242,
"learning_rate": 2.202e-05,
"loss": 8.4361,
"step": 370
},
{
"epoch": 0.0006207891955089351,
"grad_norm": 32.27824783325195,
"learning_rate": 2.2619999999999997e-05,
"loss": 8.4116,
"step": 380
},
{
"epoch": 0.0006552774841483203,
"grad_norm": 20.55720329284668,
"learning_rate": 2.3219999999999998e-05,
"loss": 8.4151,
"step": 390
},
{
"epoch": 0.0006897657727877056,
"grad_norm": 30.87930679321289,
"learning_rate": 2.382e-05,
"loss": 8.43,
"step": 400
},
{
"epoch": 0.0007242540614270909,
"grad_norm": 19.80055046081543,
"learning_rate": 2.4419999999999997e-05,
"loss": 8.4345,
"step": 410
},
{
"epoch": 0.0007587423500664762,
"grad_norm": 23.822568893432617,
"learning_rate": 2.5019999999999998e-05,
"loss": 8.413,
"step": 420
},
{
"epoch": 0.0007932306387058614,
"grad_norm": 24.005403518676758,
"learning_rate": 2.562e-05,
"loss": 8.419,
"step": 430
},
{
"epoch": 0.0008277189273452468,
"grad_norm": 30.060989379882812,
"learning_rate": 2.622e-05,
"loss": 8.4314,
"step": 440
},
{
"epoch": 0.000862207215984632,
"grad_norm": 22.84067153930664,
"learning_rate": 2.6759999999999998e-05,
"loss": 8.4154,
"step": 450
},
{
"epoch": 0.0008966955046240172,
"grad_norm": 22.259994506835938,
"learning_rate": 2.7299999999999996e-05,
"loss": 8.4086,
"step": 460
},
{
"epoch": 0.0009311837932634026,
"grad_norm": 11.726574897766113,
"learning_rate": 2.7899999999999997e-05,
"loss": 8.4284,
"step": 470
},
{
"epoch": 0.0009656720819027878,
"grad_norm": 27.165864944458008,
"learning_rate": 2.8499999999999998e-05,
"loss": 8.4169,
"step": 480
},
{
"epoch": 0.001000160370542173,
"grad_norm": 24.19461441040039,
"learning_rate": 2.91e-05,
"loss": 8.4343,
"step": 490
},
{
"epoch": 0.0010346486591815585,
"grad_norm": 10.8844633102417,
"learning_rate": 2.97e-05,
"loss": 8.4233,
"step": 500
},
{
"epoch": 0.0010691369478209437,
"grad_norm": 22.924041748046875,
"learning_rate": 3.0299999999999998e-05,
"loss": 8.394,
"step": 510
},
{
"epoch": 0.001103625236460329,
"grad_norm": 15.292486190795898,
"learning_rate": 3.09e-05,
"loss": 8.4232,
"step": 520
},
{
"epoch": 0.0011381135250997142,
"grad_norm": 18.48368263244629,
"learning_rate": 3.149999999999999e-05,
"loss": 8.4116,
"step": 530
},
{
"epoch": 0.0011726018137390997,
"grad_norm": 17.83504867553711,
"learning_rate": 3.2099999999999994e-05,
"loss": 8.4049,
"step": 540
},
{
"epoch": 0.0012070901023784849,
"grad_norm": 14.123770713806152,
"learning_rate": 3.2699999999999995e-05,
"loss": 8.4031,
"step": 550
},
{
"epoch": 0.0012415783910178701,
"grad_norm": 19.69959831237793,
"learning_rate": 3.3299999999999996e-05,
"loss": 8.4,
"step": 560
},
{
"epoch": 0.0012760666796572554,
"grad_norm": 19.486099243164062,
"learning_rate": 3.39e-05,
"loss": 8.4109,
"step": 570
},
{
"epoch": 0.0013105549682966406,
"grad_norm": 11.047612190246582,
"learning_rate": 3.45e-05,
"loss": 8.4181,
"step": 580
},
{
"epoch": 0.001345043256936026,
"grad_norm": 20.63058090209961,
"learning_rate": 3.51e-05,
"loss": 8.4011,
"step": 590
},
{
"epoch": 0.0013795315455754113,
"grad_norm": 25.09280776977539,
"learning_rate": 3.5699999999999994e-05,
"loss": 8.3946,
"step": 600
},
{
"epoch": 0.0014140198342147965,
"grad_norm": 15.544500350952148,
"learning_rate": 3.6299999999999995e-05,
"loss": 8.4085,
"step": 610
},
{
"epoch": 0.0014485081228541817,
"grad_norm": 17.291248321533203,
"learning_rate": 3.6899999999999996e-05,
"loss": 8.4186,
"step": 620
},
{
"epoch": 0.001482996411493567,
"grad_norm": 20.22525405883789,
"learning_rate": 3.75e-05,
"loss": 8.4061,
"step": 630
},
{
"epoch": 0.0015174847001329524,
"grad_norm": 11.877947807312012,
"learning_rate": 3.81e-05,
"loss": 8.3943,
"step": 640
},
{
"epoch": 0.0015519729887723376,
"grad_norm": 15.03212833404541,
"learning_rate": 3.87e-05,
"loss": 8.4077,
"step": 650
},
{
"epoch": 0.0015864612774117229,
"grad_norm": 11.607198715209961,
"learning_rate": 3.93e-05,
"loss": 8.4003,
"step": 660
},
{
"epoch": 0.0016209495660511081,
"grad_norm": 16.791051864624023,
"learning_rate": 3.99e-05,
"loss": 8.3883,
"step": 670
},
{
"epoch": 0.0016554378546904936,
"grad_norm": 13.70166015625,
"learning_rate": 4.05e-05,
"loss": 8.4086,
"step": 680
},
{
"epoch": 0.0016899261433298788,
"grad_norm": 8.733206748962402,
"learning_rate": 4.11e-05,
"loss": 8.3855,
"step": 690
},
{
"epoch": 0.001724414431969264,
"grad_norm": 18.953189849853516,
"learning_rate": 4.17e-05,
"loss": 8.3839,
"step": 700
},
{
"epoch": 0.0017589027206086493,
"grad_norm": 15.253740310668945,
"learning_rate": 4.229999999999999e-05,
"loss": 8.4009,
"step": 710
},
{
"epoch": 0.0017933910092480345,
"grad_norm": 13.945429801940918,
"learning_rate": 4.289999999999999e-05,
"loss": 8.3869,
"step": 720
},
{
"epoch": 0.00182787929788742,
"grad_norm": 15.738173484802246,
"learning_rate": 4.3499999999999993e-05,
"loss": 8.3847,
"step": 730
},
{
"epoch": 0.0018623675865268052,
"grad_norm": 13.38271713256836,
"learning_rate": 4.4099999999999995e-05,
"loss": 8.3995,
"step": 740
},
{
"epoch": 0.0018968558751661904,
"grad_norm": 14.793967247009277,
"learning_rate": 4.4699999999999996e-05,
"loss": 8.3833,
"step": 750
},
{
"epoch": 0.0019313441638055756,
"grad_norm": 16.215335845947266,
"learning_rate": 4.5299999999999997e-05,
"loss": 8.3903,
"step": 760
},
{
"epoch": 0.001965832452444961,
"grad_norm": 21.061214447021484,
"learning_rate": 4.59e-05,
"loss": 8.3884,
"step": 770
},
{
"epoch": 0.002000320741084346,
"grad_norm": 14.154373168945312,
"learning_rate": 4.65e-05,
"loss": 8.4027,
"step": 780
},
{
"epoch": 0.0020348090297237316,
"grad_norm": 9.922065734863281,
"learning_rate": 4.709999999999999e-05,
"loss": 8.4037,
"step": 790
},
{
"epoch": 0.002069297318363117,
"grad_norm": 17.40951919555664,
"learning_rate": 4.7699999999999994e-05,
"loss": 8.4032,
"step": 800
},
{
"epoch": 0.002103785607002502,
"grad_norm": 17.422048568725586,
"learning_rate": 4.8299999999999995e-05,
"loss": 8.3983,
"step": 810
},
{
"epoch": 0.0021382738956418875,
"grad_norm": 16.443422317504883,
"learning_rate": 4.8899999999999996e-05,
"loss": 8.3957,
"step": 820
},
{
"epoch": 0.0021727621842812725,
"grad_norm": 24.262035369873047,
"learning_rate": 4.95e-05,
"loss": 8.3895,
"step": 830
},
{
"epoch": 0.002207250472920658,
"grad_norm": 13.685011863708496,
"learning_rate": 5.01e-05,
"loss": 8.3915,
"step": 840
},
{
"epoch": 0.0022417387615600434,
"grad_norm": 10.055377960205078,
"learning_rate": 5.07e-05,
"loss": 8.3822,
"step": 850
},
{
"epoch": 0.0022762270501994284,
"grad_norm": 13.62060546875,
"learning_rate": 5.13e-05,
"loss": 8.3817,
"step": 860
},
{
"epoch": 0.002310715338838814,
"grad_norm": 14.18138313293457,
"learning_rate": 5.1899999999999994e-05,
"loss": 8.3889,
"step": 870
},
{
"epoch": 0.0023452036274781993,
"grad_norm": 6.779985427856445,
"learning_rate": 5.2499999999999995e-05,
"loss": 8.3846,
"step": 880
},
{
"epoch": 0.0023796919161175843,
"grad_norm": 7.078898906707764,
"learning_rate": 5.309999999999999e-05,
"loss": 8.4004,
"step": 890
},
{
"epoch": 0.0024141802047569698,
"grad_norm": 14.189363479614258,
"learning_rate": 5.369999999999999e-05,
"loss": 8.3998,
"step": 900
},
{
"epoch": 0.002448668493396355,
"grad_norm": 16.281265258789062,
"learning_rate": 5.429999999999999e-05,
"loss": 8.3943,
"step": 910
},
{
"epoch": 0.0024831567820357402,
"grad_norm": 14.237013816833496,
"learning_rate": 5.489999999999999e-05,
"loss": 8.382,
"step": 920
},
{
"epoch": 0.0025176450706751257,
"grad_norm": 7.917139530181885,
"learning_rate": 5.5499999999999994e-05,
"loss": 8.3985,
"step": 930
},
{
"epoch": 0.0025521333593145107,
"grad_norm": 13.905786514282227,
"learning_rate": 5.6099999999999995e-05,
"loss": 8.3868,
"step": 940
},
{
"epoch": 0.002586621647953896,
"grad_norm": 14.011921882629395,
"learning_rate": 5.6699999999999996e-05,
"loss": 8.3773,
"step": 950
},
{
"epoch": 0.002621109936593281,
"grad_norm": 10.070409774780273,
"learning_rate": 5.73e-05,
"loss": 8.3894,
"step": 960
},
{
"epoch": 0.0026555982252326666,
"grad_norm": 12.96844482421875,
"learning_rate": 5.79e-05,
"loss": 8.3818,
"step": 970
},
{
"epoch": 0.002690086513872052,
"grad_norm": 14.524611473083496,
"learning_rate": 5.85e-05,
"loss": 8.3793,
"step": 980
},
{
"epoch": 0.002724574802511437,
"grad_norm": 23.980628967285156,
"learning_rate": 5.91e-05,
"loss": 8.3702,
"step": 990
},
{
"epoch": 0.0027590630911508225,
"grad_norm": 79.6168441772461,
"learning_rate": 5.97e-05,
"loss": 8.3695,
"step": 1000
},
{
"epoch": 0.0027935513797902075,
"grad_norm": 11.407526016235352,
"learning_rate": 6.0299999999999995e-05,
"loss": 8.3857,
"step": 1010
},
{
"epoch": 0.002828039668429593,
"grad_norm": 12.697784423828125,
"learning_rate": 6.0899999999999996e-05,
"loss": 8.3714,
"step": 1020
},
{
"epoch": 0.0028625279570689784,
"grad_norm": 7.642915725708008,
"learning_rate": 6.149999999999999e-05,
"loss": 8.3924,
"step": 1030
},
{
"epoch": 0.0028970162457083635,
"grad_norm": 13.278314590454102,
"learning_rate": 6.209999999999999e-05,
"loss": 8.3676,
"step": 1040
},
{
"epoch": 0.002931504534347749,
"grad_norm": 9.300806045532227,
"learning_rate": 6.269999999999999e-05,
"loss": 8.3769,
"step": 1050
},
{
"epoch": 0.002965992822987134,
"grad_norm": 10.766736030578613,
"learning_rate": 6.33e-05,
"loss": 8.3722,
"step": 1060
},
{
"epoch": 0.0030004811116265194,
"grad_norm": 16.214441299438477,
"learning_rate": 6.39e-05,
"loss": 8.3866,
"step": 1070
},
{
"epoch": 0.003034969400265905,
"grad_norm": 12.498828887939453,
"learning_rate": 6.45e-05,
"loss": 8.3778,
"step": 1080
},
{
"epoch": 0.00306945768890529,
"grad_norm": 12.378737449645996,
"learning_rate": 6.503999999999999e-05,
"loss": 8.3748,
"step": 1090
},
{
"epoch": 0.0031039459775446753,
"grad_norm": 7.89847993850708,
"learning_rate": 6.563999999999999e-05,
"loss": 8.3751,
"step": 1100
},
{
"epoch": 0.0031384342661840607,
"grad_norm": 11.73416805267334,
"learning_rate": 6.623999999999999e-05,
"loss": 8.3712,
"step": 1110
},
{
"epoch": 0.0031729225548234458,
"grad_norm": 8.572582244873047,
"learning_rate": 6.683999999999999e-05,
"loss": 8.3753,
"step": 1120
},
{
"epoch": 0.003207410843462831,
"grad_norm": 13.893655776977539,
"learning_rate": 6.743999999999999e-05,
"loss": 8.3667,
"step": 1130
},
{
"epoch": 0.0032418991321022162,
"grad_norm": 7.900502681732178,
"learning_rate": 6.803999999999999e-05,
"loss": 8.3636,
"step": 1140
},
{
"epoch": 0.0032763874207416017,
"grad_norm": 12.325122833251953,
"learning_rate": 6.864e-05,
"loss": 8.3722,
"step": 1150
},
{
"epoch": 0.003310875709380987,
"grad_norm": 12.237879753112793,
"learning_rate": 6.924e-05,
"loss": 8.3573,
"step": 1160
},
{
"epoch": 0.003345363998020372,
"grad_norm": 16.15164566040039,
"learning_rate": 6.984e-05,
"loss": 8.3626,
"step": 1170
},
{
"epoch": 0.0033798522866597576,
"grad_norm": 16.162899017333984,
"learning_rate": 7.044e-05,
"loss": 8.3827,
"step": 1180
},
{
"epoch": 0.0034143405752991426,
"grad_norm": 22.04014015197754,
"learning_rate": 7.104e-05,
"loss": 8.3637,
"step": 1190
},
{
"epoch": 0.003448828863938528,
"grad_norm": 11.339788436889648,
"learning_rate": 7.164e-05,
"loss": 8.3725,
"step": 1200
},
{
"epoch": 0.0034833171525779135,
"grad_norm": 13.985651016235352,
"learning_rate": 7.223999999999999e-05,
"loss": 8.3759,
"step": 1210
},
{
"epoch": 0.0035178054412172985,
"grad_norm": 12.80493450164795,
"learning_rate": 7.283999999999999e-05,
"loss": 8.3635,
"step": 1220
},
{
"epoch": 0.003552293729856684,
"grad_norm": 10.030963897705078,
"learning_rate": 7.343999999999999e-05,
"loss": 8.3563,
"step": 1230
},
{
"epoch": 0.003586782018496069,
"grad_norm": 10.714138984680176,
"learning_rate": 7.403999999999999e-05,
"loss": 8.3529,
"step": 1240
},
{
"epoch": 0.0036212703071354544,
"grad_norm": 14.480472564697266,
"learning_rate": 7.463999999999999e-05,
"loss": 8.3377,
"step": 1250
},
{
"epoch": 0.00365575859577484,
"grad_norm": 6.988164901733398,
"learning_rate": 7.524e-05,
"loss": 8.3632,
"step": 1260
},
{
"epoch": 0.003690246884414225,
"grad_norm": 9.82447338104248,
"learning_rate": 7.584e-05,
"loss": 8.3417,
"step": 1270
},
{
"epoch": 0.0037247351730536104,
"grad_norm": 11.450285911560059,
"learning_rate": 7.644e-05,
"loss": 8.3563,
"step": 1280
},
{
"epoch": 0.003759223461692996,
"grad_norm": 10.065130233764648,
"learning_rate": 7.703999999999998e-05,
"loss": 8.3482,
"step": 1290
},
{
"epoch": 0.003793711750332381,
"grad_norm": 13.670631408691406,
"learning_rate": 7.763999999999998e-05,
"loss": 8.3319,
"step": 1300
},
{
"epoch": 0.0038282000389717663,
"grad_norm": 11.154393196105957,
"learning_rate": 7.823999999999998e-05,
"loss": 8.3378,
"step": 1310
},
{
"epoch": 0.0038626883276111513,
"grad_norm": 13.057133674621582,
"learning_rate": 7.883999999999998e-05,
"loss": 8.3481,
"step": 1320
},
{
"epoch": 0.0038971766162505367,
"grad_norm": 8.13063907623291,
"learning_rate": 7.943999999999998e-05,
"loss": 8.3619,
"step": 1330
},
{
"epoch": 0.003931664904889922,
"grad_norm": 7.690333843231201,
"learning_rate": 8.003999999999999e-05,
"loss": 8.3349,
"step": 1340
},
{
"epoch": 0.003966153193529307,
"grad_norm": 8.945077896118164,
"learning_rate": 8.063999999999999e-05,
"loss": 8.3609,
"step": 1350
},
{
"epoch": 0.004000641482168692,
"grad_norm": 11.839431762695312,
"learning_rate": 8.123999999999999e-05,
"loss": 8.3552,
"step": 1360
},
{
"epoch": 0.004035129770808078,
"grad_norm": 9.798072814941406,
"learning_rate": 8.183999999999999e-05,
"loss": 8.3124,
"step": 1370
},
{
"epoch": 0.004069618059447463,
"grad_norm": 13.58779239654541,
"learning_rate": 8.243999999999999e-05,
"loss": 8.3544,
"step": 1380
},
{
"epoch": 0.004104106348086848,
"grad_norm": 12.656436920166016,
"learning_rate": 8.303999999999999e-05,
"loss": 8.3457,
"step": 1390
},
{
"epoch": 0.004138594636726234,
"grad_norm": 9.099520683288574,
"learning_rate": 8.363999999999999e-05,
"loss": 8.3605,
"step": 1400
},
{
"epoch": 0.004173082925365619,
"grad_norm": 19.523441314697266,
"learning_rate": 8.423999999999999e-05,
"loss": 8.3333,
"step": 1410
},
{
"epoch": 0.004207571214005004,
"grad_norm": 13.605344772338867,
"learning_rate": 8.484e-05,
"loss": 8.3487,
"step": 1420
},
{
"epoch": 0.00424205950264439,
"grad_norm": 15.3053560256958,
"learning_rate": 8.544e-05,
"loss": 8.3397,
"step": 1430
},
{
"epoch": 0.004276547791283775,
"grad_norm": 7.251153469085693,
"learning_rate": 8.604e-05,
"loss": 8.349,
"step": 1440
},
{
"epoch": 0.00431103607992316,
"grad_norm": 9.477354049682617,
"learning_rate": 8.664e-05,
"loss": 8.3419,
"step": 1450
},
{
"epoch": 0.004345524368562545,
"grad_norm": 14.80703067779541,
"learning_rate": 8.724e-05,
"loss": 8.319,
"step": 1460
},
{
"epoch": 0.004380012657201931,
"grad_norm": 14.25675106048584,
"learning_rate": 8.784e-05,
"loss": 8.35,
"step": 1470
},
{
"epoch": 0.004414500945841316,
"grad_norm": 10.686800003051758,
"learning_rate": 8.844e-05,
"loss": 8.3404,
"step": 1480
},
{
"epoch": 0.004448989234480701,
"grad_norm": 9.987860679626465,
"learning_rate": 8.904e-05,
"loss": 8.3332,
"step": 1490
},
{
"epoch": 0.004483477523120087,
"grad_norm": 7.363917350769043,
"learning_rate": 8.963999999999999e-05,
"loss": 8.3299,
"step": 1500
},
{
"epoch": 0.004517965811759472,
"grad_norm": 14.461793899536133,
"learning_rate": 9.023999999999999e-05,
"loss": 8.3306,
"step": 1510
},
{
"epoch": 0.004552454100398857,
"grad_norm": 8.630695343017578,
"learning_rate": 9.083999999999999e-05,
"loss": 8.3417,
"step": 1520
},
{
"epoch": 0.004586942389038243,
"grad_norm": 8.256804466247559,
"learning_rate": 9.143999999999999e-05,
"loss": 8.3341,
"step": 1530
},
{
"epoch": 0.004621430677677628,
"grad_norm": 13.018980026245117,
"learning_rate": 9.203999999999999e-05,
"loss": 8.3408,
"step": 1540
},
{
"epoch": 0.004655918966317013,
"grad_norm": 8.717267036437988,
"learning_rate": 9.264e-05,
"loss": 8.3248,
"step": 1550
},
{
"epoch": 0.004690407254956399,
"grad_norm": 15.733688354492188,
"learning_rate": 9.324e-05,
"loss": 8.3205,
"step": 1560
},
{
"epoch": 0.004724895543595784,
"grad_norm": 15.750919342041016,
"learning_rate": 9.384e-05,
"loss": 8.3227,
"step": 1570
},
{
"epoch": 0.004759383832235169,
"grad_norm": 12.715235710144043,
"learning_rate": 9.444e-05,
"loss": 8.3421,
"step": 1580
},
{
"epoch": 0.004793872120874554,
"grad_norm": 12.098243713378906,
"learning_rate": 9.504e-05,
"loss": 8.3332,
"step": 1590
},
{
"epoch": 0.0048283604095139395,
"grad_norm": 8.364949226379395,
"learning_rate": 9.563999999999999e-05,
"loss": 8.3246,
"step": 1600
},
{
"epoch": 0.0048628486981533246,
"grad_norm": 12.680130004882812,
"learning_rate": 9.623999999999999e-05,
"loss": 8.3347,
"step": 1610
},
{
"epoch": 0.00489733698679271,
"grad_norm": 5.828690052032471,
"learning_rate": 9.683999999999999e-05,
"loss": 8.3288,
"step": 1620
},
{
"epoch": 0.0049318252754320955,
"grad_norm": 10.879535675048828,
"learning_rate": 9.743999999999999e-05,
"loss": 8.3261,
"step": 1630
},
{
"epoch": 0.0049663135640714805,
"grad_norm": 10.933409690856934,
"learning_rate": 9.803999999999999e-05,
"loss": 8.3129,
"step": 1640
},
{
"epoch": 0.0050008018527108655,
"grad_norm": 9.908263206481934,
"learning_rate": 9.863999999999999e-05,
"loss": 8.3215,
"step": 1650
},
{
"epoch": 0.005035290141350251,
"grad_norm": 9.367692947387695,
"learning_rate": 9.923999999999999e-05,
"loss": 8.3161,
"step": 1660
},
{
"epoch": 0.005069778429989636,
"grad_norm": 17.53960418701172,
"learning_rate": 9.983999999999999e-05,
"loss": 8.3095,
"step": 1670
},
{
"epoch": 0.005104266718629021,
"grad_norm": 12.228901863098145,
"learning_rate": 0.00010044,
"loss": 8.3064,
"step": 1680
},
{
"epoch": 0.005138755007268406,
"grad_norm": 9.411287307739258,
"learning_rate": 0.00010103999999999998,
"loss": 8.3047,
"step": 1690
},
{
"epoch": 0.005173243295907792,
"grad_norm": 12.89146900177002,
"learning_rate": 0.00010163999999999998,
"loss": 8.313,
"step": 1700
},
{
"epoch": 0.005207731584547177,
"grad_norm": 13.037481307983398,
"learning_rate": 0.00010223999999999998,
"loss": 8.3005,
"step": 1710
},
{
"epoch": 0.005242219873186562,
"grad_norm": 18.167877197265625,
"learning_rate": 0.00010283999999999998,
"loss": 8.3194,
"step": 1720
},
{
"epoch": 0.005276708161825948,
"grad_norm": 11.53596019744873,
"learning_rate": 0.00010343999999999999,
"loss": 8.3103,
"step": 1730
},
{
"epoch": 0.005311196450465333,
"grad_norm": 14.443565368652344,
"learning_rate": 0.00010403999999999999,
"loss": 8.3078,
"step": 1740
},
{
"epoch": 0.005345684739104718,
"grad_norm": 12.966043472290039,
"learning_rate": 0.00010463999999999999,
"loss": 8.3133,
"step": 1750
},
{
"epoch": 0.005380173027744104,
"grad_norm": 9.95252513885498,
"learning_rate": 0.00010523999999999999,
"loss": 8.3033,
"step": 1760
},
{
"epoch": 0.005414661316383489,
"grad_norm": 15.110918045043945,
"learning_rate": 0.00010583999999999999,
"loss": 8.3018,
"step": 1770
},
{
"epoch": 0.005449149605022874,
"grad_norm": 9.710939407348633,
"learning_rate": 0.00010643999999999999,
"loss": 8.3056,
"step": 1780
},
{
"epoch": 0.00548363789366226,
"grad_norm": 14.128268241882324,
"learning_rate": 0.00010703999999999999,
"loss": 8.3116,
"step": 1790
},
{
"epoch": 0.005518126182301645,
"grad_norm": 12.291546821594238,
"learning_rate": 0.00010763999999999999,
"loss": 8.3087,
"step": 1800
},
{
"epoch": 0.00555261447094103,
"grad_norm": 8.979240417480469,
"learning_rate": 0.00010824,
"loss": 8.297,
"step": 1810
},
{
"epoch": 0.005587102759580415,
"grad_norm": 16.653636932373047,
"learning_rate": 0.00010884,
"loss": 8.3032,
"step": 1820
},
{
"epoch": 0.005621591048219801,
"grad_norm": 7.231003761291504,
"learning_rate": 0.00010944,
"loss": 8.3017,
"step": 1830
},
{
"epoch": 0.005656079336859186,
"grad_norm": 11.044018745422363,
"learning_rate": 0.00011004,
"loss": 8.2998,
"step": 1840
},
{
"epoch": 0.005690567625498571,
"grad_norm": 7.183733940124512,
"learning_rate": 0.00011064,
"loss": 8.2792,
"step": 1850
},
{
"epoch": 0.005725055914137957,
"grad_norm": 6.851602554321289,
"learning_rate": 0.00011124,
"loss": 8.3057,
"step": 1860
},
{
"epoch": 0.005759544202777342,
"grad_norm": 6.955447196960449,
"learning_rate": 0.00011184,
"loss": 8.2914,
"step": 1870
},
{
"epoch": 0.005794032491416727,
"grad_norm": 9.28893756866455,
"learning_rate": 0.00011244,
"loss": 8.296,
"step": 1880
},
{
"epoch": 0.005828520780056113,
"grad_norm": 12.825960159301758,
"learning_rate": 0.00011304,
"loss": 8.29,
"step": 1890
},
{
"epoch": 0.005863009068695498,
"grad_norm": 7.817082405090332,
"learning_rate": 0.00011364,
"loss": 8.2943,
"step": 1900
},
{
"epoch": 0.005897497357334883,
"grad_norm": 5.764162063598633,
"learning_rate": 0.00011424,
"loss": 8.2943,
"step": 1910
},
{
"epoch": 0.005931985645974268,
"grad_norm": 6.886749267578125,
"learning_rate": 0.00011483999999999998,
"loss": 8.2861,
"step": 1920
},
{
"epoch": 0.005966473934613654,
"grad_norm": 6.364310264587402,
"learning_rate": 0.00011543999999999998,
"loss": 8.2987,
"step": 1930
},
{
"epoch": 0.006000962223253039,
"grad_norm": 6.065140247344971,
"learning_rate": 0.00011603999999999998,
"loss": 8.2828,
"step": 1940
},
{
"epoch": 0.006035450511892424,
"grad_norm": 11.439751625061035,
"learning_rate": 0.00011663999999999998,
"loss": 8.2804,
"step": 1950
},
{
"epoch": 0.00606993880053181,
"grad_norm": 7.701569080352783,
"learning_rate": 0.00011723999999999998,
"loss": 8.2792,
"step": 1960
},
{
"epoch": 0.006104427089171195,
"grad_norm": 27.84369659423828,
"learning_rate": 0.00011783999999999998,
"loss": 8.283,
"step": 1970
},
{
"epoch": 0.00613891537781058,
"grad_norm": 16.438413619995117,
"learning_rate": 0.00011843999999999998,
"loss": 8.3069,
"step": 1980
},
{
"epoch": 0.006173403666449966,
"grad_norm": 10.777158737182617,
"learning_rate": 0.00011903999999999999,
"loss": 8.2894,
"step": 1990
},
{
"epoch": 0.006207891955089351,
"grad_norm": 11.02175235748291,
"learning_rate": 0.00011963999999999999,
"loss": 8.2933,
"step": 2000
},
{
"epoch": 0.006242380243728736,
"grad_norm": 12.038606643676758,
"learning_rate": 0.00012023999999999999,
"loss": 8.2855,
"step": 2010
},
{
"epoch": 0.0062768685323681215,
"grad_norm": 15.360584259033203,
"learning_rate": 0.00012083999999999999,
"loss": 8.2823,
"step": 2020
},
{
"epoch": 0.0063113568210075065,
"grad_norm": 8.902900695800781,
"learning_rate": 0.00012143999999999999,
"loss": 8.2887,
"step": 2030
},
{
"epoch": 0.0063458451096468915,
"grad_norm": 8.834257125854492,
"learning_rate": 0.00012203999999999999,
"loss": 8.267,
"step": 2040
},
{
"epoch": 0.0063803333982862765,
"grad_norm": 8.919024467468262,
"learning_rate": 0.00012263999999999998,
"loss": 8.2935,
"step": 2050
},
{
"epoch": 0.006414821686925662,
"grad_norm": 11.514122009277344,
"learning_rate": 0.00012324,
"loss": 8.2532,
"step": 2060
},
{
"epoch": 0.0064493099755650474,
"grad_norm": 14.257649421691895,
"learning_rate": 0.00012383999999999998,
"loss": 8.2589,
"step": 2070
},
{
"epoch": 0.0064837982642044325,
"grad_norm": 9.895721435546875,
"learning_rate": 0.00012444,
"loss": 8.2733,
"step": 2080
},
{
"epoch": 0.006518286552843818,
"grad_norm": 9.748128890991211,
"learning_rate": 0.00012503999999999998,
"loss": 8.2575,
"step": 2090
},
{
"epoch": 0.006552774841483203,
"grad_norm": 9.160621643066406,
"learning_rate": 0.00012564,
"loss": 8.285,
"step": 2100
},
{
"epoch": 0.006587263130122588,
"grad_norm": 14.095819473266602,
"learning_rate": 0.00012623999999999998,
"loss": 8.2613,
"step": 2110
},
{
"epoch": 0.006621751418761974,
"grad_norm": 10.618462562561035,
"learning_rate": 0.00012684,
"loss": 8.2808,
"step": 2120
},
{
"epoch": 0.006656239707401359,
"grad_norm": 9.176691055297852,
"learning_rate": 0.00012743999999999999,
"loss": 8.2753,
"step": 2130
},
{
"epoch": 0.006690727996040744,
"grad_norm": 12.300357818603516,
"learning_rate": 0.00012804,
"loss": 8.2777,
"step": 2140
},
{
"epoch": 0.00672521628468013,
"grad_norm": 15.15916919708252,
"learning_rate": 0.00012864,
"loss": 8.2636,
"step": 2150
},
{
"epoch": 0.006759704573319515,
"grad_norm": 13.030776977539062,
"learning_rate": 0.00012924,
"loss": 8.262,
"step": 2160
},
{
"epoch": 0.0067941928619589,
"grad_norm": 9.670841217041016,
"learning_rate": 0.00012984,
"loss": 8.2866,
"step": 2170
},
{
"epoch": 0.006828681150598285,
"grad_norm": 15.426817893981934,
"learning_rate": 0.00013044,
"loss": 8.2659,
"step": 2180
},
{
"epoch": 0.006863169439237671,
"grad_norm": 11.12787914276123,
"learning_rate": 0.00013104,
"loss": 8.2738,
"step": 2190
},
{
"epoch": 0.006897657727877056,
"grad_norm": 8.150566101074219,
"learning_rate": 0.00013164,
"loss": 8.2478,
"step": 2200
},
{
"epoch": 0.006932146016516441,
"grad_norm": 10.044174194335938,
"learning_rate": 0.00013224,
"loss": 8.2494,
"step": 2210
},
{
"epoch": 0.006966634305155827,
"grad_norm": 7.456984996795654,
"learning_rate": 0.00013284,
"loss": 8.2632,
"step": 2220
},
{
"epoch": 0.007001122593795212,
"grad_norm": 11.210662841796875,
"learning_rate": 0.00013343999999999997,
"loss": 8.2636,
"step": 2230
},
{
"epoch": 0.007035610882434597,
"grad_norm": 13.365647315979004,
"learning_rate": 0.00013403999999999998,
"loss": 8.2558,
"step": 2240
},
{
"epoch": 0.007070099171073983,
"grad_norm": 10.76402473449707,
"learning_rate": 0.00013463999999999997,
"loss": 8.2466,
"step": 2250
},
{
"epoch": 0.007104587459713368,
"grad_norm": 12.834994316101074,
"learning_rate": 0.00013523999999999999,
"loss": 8.247,
"step": 2260
},
{
"epoch": 0.007139075748352753,
"grad_norm": 13.960427284240723,
"learning_rate": 0.00013583999999999997,
"loss": 8.2624,
"step": 2270
},
{
"epoch": 0.007173564036992138,
"grad_norm": 11.685848236083984,
"learning_rate": 0.00013644,
"loss": 8.255,
"step": 2280
},
{
"epoch": 0.007208052325631524,
"grad_norm": 10.026966094970703,
"learning_rate": 0.00013703999999999998,
"loss": 8.2459,
"step": 2290
},
{
"epoch": 0.007242540614270909,
"grad_norm": 8.979503631591797,
"learning_rate": 0.00013764,
"loss": 8.2485,
"step": 2300
},
{
"epoch": 0.007277028902910294,
"grad_norm": 12.863329887390137,
"learning_rate": 0.00013823999999999998,
"loss": 8.2427,
"step": 2310
},
{
"epoch": 0.00731151719154968,
"grad_norm": 8.985013961791992,
"learning_rate": 0.00013884,
"loss": 8.2594,
"step": 2320
},
{
"epoch": 0.007346005480189065,
"grad_norm": 14.681097030639648,
"learning_rate": 0.00013943999999999998,
"loss": 8.2528,
"step": 2330
},
{
"epoch": 0.00738049376882845,
"grad_norm": 17.21650505065918,
"learning_rate": 0.00014004,
"loss": 8.2437,
"step": 2340
},
{
"epoch": 0.007414982057467836,
"grad_norm": 5.822660446166992,
"learning_rate": 0.00014063999999999998,
"loss": 8.2456,
"step": 2350
},
{
"epoch": 0.007449470346107221,
"grad_norm": 5.644645690917969,
"learning_rate": 0.00014124,
"loss": 8.2322,
"step": 2360
},
{
"epoch": 0.007483958634746606,
"grad_norm": 5.099442005157471,
"learning_rate": 0.00014183999999999998,
"loss": 8.2522,
"step": 2370
},
{
"epoch": 0.007518446923385992,
"grad_norm": 10.48246955871582,
"learning_rate": 0.00014244,
"loss": 8.2434,
"step": 2380
},
{
"epoch": 0.007552935212025377,
"grad_norm": 9.684501647949219,
"learning_rate": 0.00014303999999999999,
"loss": 8.2283,
"step": 2390
},
{
"epoch": 0.007587423500664762,
"grad_norm": 14.485626220703125,
"learning_rate": 0.00014364,
"loss": 8.2252,
"step": 2400
},
{
"epoch": 0.007621911789304147,
"grad_norm": 9.568318367004395,
"learning_rate": 0.00014424,
"loss": 8.2169,
"step": 2410
},
{
"epoch": 0.0076564000779435325,
"grad_norm": 13.435136795043945,
"learning_rate": 0.00014484,
"loss": 8.2358,
"step": 2420
},
{
"epoch": 0.0076908883665829176,
"grad_norm": 8.654492378234863,
"learning_rate": 0.00014544,
"loss": 8.2305,
"step": 2430
},
{
"epoch": 0.007725376655222303,
"grad_norm": 8.88370418548584,
"learning_rate": 0.00014603999999999998,
"loss": 8.2422,
"step": 2440
},
{
"epoch": 0.0077598649438616885,
"grad_norm": 16.867353439331055,
"learning_rate": 0.00014664,
"loss": 8.2255,
"step": 2450
},
{
"epoch": 0.0077943532325010735,
"grad_norm": 15.305095672607422,
"learning_rate": 0.00014723999999999998,
"loss": 8.2227,
"step": 2460
},
{
"epoch": 0.007828841521140458,
"grad_norm": 10.0158052444458,
"learning_rate": 0.00014784,
"loss": 8.2215,
"step": 2470
},
{
"epoch": 0.007863329809779844,
"grad_norm": 13.85173225402832,
"learning_rate": 0.00014843999999999998,
"loss": 8.2297,
"step": 2480
},
{
"epoch": 0.007897818098419229,
"grad_norm": 7.829097747802734,
"learning_rate": 0.00014904,
"loss": 8.2212,
"step": 2490
},
{
"epoch": 0.007932306387058614,
"grad_norm": 14.449021339416504,
"learning_rate": 0.00014963999999999998,
"loss": 8.2244,
"step": 2500
},
{
"epoch": 0.007966794675698,
"grad_norm": 10.232049942016602,
"learning_rate": 0.00015024,
"loss": 8.2225,
"step": 2510
},
{
"epoch": 0.008001282964337384,
"grad_norm": 7.4921875,
"learning_rate": 0.00015083999999999999,
"loss": 8.2141,
"step": 2520
},
{
"epoch": 0.00803577125297677,
"grad_norm": 9.413374900817871,
"learning_rate": 0.00015144,
"loss": 8.2167,
"step": 2530
},
{
"epoch": 0.008070259541616156,
"grad_norm": 8.397259712219238,
"learning_rate": 0.00015204,
"loss": 8.2153,
"step": 2540
},
{
"epoch": 0.00810474783025554,
"grad_norm": 12.561037063598633,
"learning_rate": 0.00015264,
"loss": 8.2165,
"step": 2550
},
{
"epoch": 0.008139236118894926,
"grad_norm": 7.44576358795166,
"learning_rate": 0.00015324,
"loss": 8.2223,
"step": 2560
},
{
"epoch": 0.008173724407534312,
"grad_norm": 7.779847621917725,
"learning_rate": 0.00015384,
"loss": 8.2148,
"step": 2570
},
{
"epoch": 0.008208212696173696,
"grad_norm": 7.283217430114746,
"learning_rate": 0.00015444,
"loss": 8.2217,
"step": 2580
},
{
"epoch": 0.008242700984813082,
"grad_norm": 5.9629106521606445,
"learning_rate": 0.00015504,
"loss": 8.2161,
"step": 2590
},
{
"epoch": 0.008277189273452468,
"grad_norm": 8.903937339782715,
"learning_rate": 0.00015564,
"loss": 8.2195,
"step": 2600
},
{
"epoch": 0.008311677562091852,
"grad_norm": 13.877178192138672,
"learning_rate": 0.00015624,
"loss": 8.2247,
"step": 2610
},
{
"epoch": 0.008346165850731238,
"grad_norm": 17.147628784179688,
"learning_rate": 0.00015684,
"loss": 8.2117,
"step": 2620
},
{
"epoch": 0.008380654139370624,
"grad_norm": 11.52202320098877,
"learning_rate": 0.00015744,
"loss": 8.2079,
"step": 2630
},
{
"epoch": 0.008415142428010008,
"grad_norm": 13.7362642288208,
"learning_rate": 0.00015804,
"loss": 8.1936,
"step": 2640
},
{
"epoch": 0.008449630716649394,
"grad_norm": 9.204730987548828,
"learning_rate": 0.00015864,
"loss": 8.201,
"step": 2650
},
{
"epoch": 0.00848411900528878,
"grad_norm": 9.73067855834961,
"learning_rate": 0.00015924,
"loss": 8.2129,
"step": 2660
},
{
"epoch": 0.008518607293928164,
"grad_norm": 13.568399429321289,
"learning_rate": 0.00015984000000000001,
"loss": 8.1985,
"step": 2670
},
{
"epoch": 0.00855309558256755,
"grad_norm": 8.648536682128906,
"learning_rate": 0.00016044,
"loss": 8.2032,
"step": 2680
},
{
"epoch": 0.008587583871206934,
"grad_norm": 12.637685775756836,
"learning_rate": 0.00016104,
"loss": 8.1951,
"step": 2690
},
{
"epoch": 0.00862207215984632,
"grad_norm": 9.516839027404785,
"learning_rate": 0.00016163999999999998,
"loss": 8.1942,
"step": 2700
},
{
"epoch": 0.008656560448485706,
"grad_norm": 11.678093910217285,
"learning_rate": 0.00016223999999999996,
"loss": 8.1984,
"step": 2710
},
{
"epoch": 0.00869104873712509,
"grad_norm": 8.335277557373047,
"learning_rate": 0.00016283999999999998,
"loss": 8.1956,
"step": 2720
},
{
"epoch": 0.008725537025764476,
"grad_norm": 13.28842544555664,
"learning_rate": 0.00016343999999999997,
"loss": 8.185,
"step": 2730
},
{
"epoch": 0.008760025314403862,
"grad_norm": 19.442224502563477,
"learning_rate": 0.00016403999999999998,
"loss": 8.1824,
"step": 2740
},
{
"epoch": 0.008794513603043246,
"grad_norm": 14.967656135559082,
"learning_rate": 0.00016463999999999997,
"loss": 8.1977,
"step": 2750
},
{
"epoch": 0.008829001891682632,
"grad_norm": 11.715154647827148,
"learning_rate": 0.00016523999999999998,
"loss": 8.1949,
"step": 2760
},
{
"epoch": 0.008863490180322018,
"grad_norm": 8.84989070892334,
"learning_rate": 0.00016583999999999997,
"loss": 8.1863,
"step": 2770
},
{
"epoch": 0.008897978468961402,
"grad_norm": 10.982098579406738,
"learning_rate": 0.00016643999999999999,
"loss": 8.1668,
"step": 2780
},
{
"epoch": 0.008932466757600788,
"grad_norm": 12.693875312805176,
"learning_rate": 0.00016703999999999997,
"loss": 8.1866,
"step": 2790
},
{
"epoch": 0.008966955046240174,
"grad_norm": 6.943259239196777,
"learning_rate": 0.00016764,
"loss": 8.1953,
"step": 2800
},
{
"epoch": 0.009001443334879558,
"grad_norm": 11.37366008758545,
"learning_rate": 0.00016823999999999997,
"loss": 8.1878,
"step": 2810
},
{
"epoch": 0.009035931623518944,
"grad_norm": 12.237605094909668,
"learning_rate": 0.00016883999999999996,
"loss": 8.1766,
"step": 2820
},
{
"epoch": 0.00907041991215833,
"grad_norm": 16.808801651000977,
"learning_rate": 0.00016943999999999998,
"loss": 8.1807,
"step": 2830
},
{
"epoch": 0.009104908200797714,
"grad_norm": 11.261263847351074,
"learning_rate": 0.00017003999999999996,
"loss": 8.1648,
"step": 2840
},
{
"epoch": 0.0091393964894371,
"grad_norm": 17.05980682373047,
"learning_rate": 0.00017063999999999998,
"loss": 8.1839,
"step": 2850
},
{
"epoch": 0.009173884778076485,
"grad_norm": 15.804736137390137,
"learning_rate": 0.00017123999999999997,
"loss": 8.167,
"step": 2860
},
{
"epoch": 0.00920837306671587,
"grad_norm": 7.427157402038574,
"learning_rate": 0.00017183999999999998,
"loss": 8.1892,
"step": 2870
},
{
"epoch": 0.009242861355355255,
"grad_norm": 12.42757797241211,
"learning_rate": 0.00017243999999999997,
"loss": 8.1538,
"step": 2880
},
{
"epoch": 0.009277349643994641,
"grad_norm": 17.209657669067383,
"learning_rate": 0.00017303999999999998,
"loss": 8.1743,
"step": 2890
},
{
"epoch": 0.009311837932634025,
"grad_norm": 13.741491317749023,
"learning_rate": 0.00017363999999999997,
"loss": 8.1646,
"step": 2900
},
{
"epoch": 0.009346326221273411,
"grad_norm": 14.243498802185059,
"learning_rate": 0.00017423999999999999,
"loss": 8.1661,
"step": 2910
},
{
"epoch": 0.009380814509912797,
"grad_norm": 9.208465576171875,
"learning_rate": 0.00017483999999999997,
"loss": 8.1586,
"step": 2920
},
{
"epoch": 0.009415302798552181,
"grad_norm": 15.108938217163086,
"learning_rate": 0.00017544,
"loss": 8.1642,
"step": 2930
},
{
"epoch": 0.009449791087191567,
"grad_norm": 13.513765335083008,
"learning_rate": 0.00017603999999999997,
"loss": 8.1649,
"step": 2940
},
{
"epoch": 0.009484279375830951,
"grad_norm": 16.385805130004883,
"learning_rate": 0.00017664,
"loss": 8.1444,
"step": 2950
},
{
"epoch": 0.009518767664470337,
"grad_norm": 9.07113265991211,
"learning_rate": 0.00017723999999999998,
"loss": 8.1709,
"step": 2960
},
{
"epoch": 0.009553255953109723,
"grad_norm": 10.152043342590332,
"learning_rate": 0.00017784,
"loss": 8.1571,
"step": 2970
},
{
"epoch": 0.009587744241749107,
"grad_norm": 10.640069007873535,
"learning_rate": 0.00017843999999999998,
"loss": 8.1708,
"step": 2980
},
{
"epoch": 0.009622232530388493,
"grad_norm": 12.887789726257324,
"learning_rate": 0.00017904,
"loss": 8.1552,
"step": 2990
},
{
"epoch": 0.009656720819027879,
"grad_norm": 19.61029624938965,
"learning_rate": 0.00017963999999999998,
"loss": 8.1308,
"step": 3000
}
],
"logging_steps": 10,
"max_steps": 10000000,
"num_input_tokens_seen": 0,
"num_train_epochs": 35,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.4140324297801728e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}