DewiBrynJones's picture
End of training
e5ee54c verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.6299918500407498,
"eval_steps": 1000,
"global_step": 5000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.008149959250203748,
"grad_norm": 34.218711853027344,
"learning_rate": 4.6000000000000004e-07,
"loss": 4.8764,
"step": 25
},
{
"epoch": 0.016299918500407497,
"grad_norm": 28.213876724243164,
"learning_rate": 9.600000000000001e-07,
"loss": 4.6956,
"step": 50
},
{
"epoch": 0.02444987775061125,
"grad_norm": 25.532794952392578,
"learning_rate": 1.46e-06,
"loss": 4.4548,
"step": 75
},
{
"epoch": 0.032599837000814993,
"grad_norm": 20.307037353515625,
"learning_rate": 1.9600000000000003e-06,
"loss": 4.0535,
"step": 100
},
{
"epoch": 0.040749796251018745,
"grad_norm": 20.705177307128906,
"learning_rate": 2.46e-06,
"loss": 3.7212,
"step": 125
},
{
"epoch": 0.0488997555012225,
"grad_norm": 20.950613021850586,
"learning_rate": 2.96e-06,
"loss": 3.3639,
"step": 150
},
{
"epoch": 0.05704971475142624,
"grad_norm": 22.719501495361328,
"learning_rate": 3.46e-06,
"loss": 3.2023,
"step": 175
},
{
"epoch": 0.06519967400162999,
"grad_norm": 21.53318977355957,
"learning_rate": 3.96e-06,
"loss": 3.1417,
"step": 200
},
{
"epoch": 0.07334963325183375,
"grad_norm": 18.566478729248047,
"learning_rate": 4.4600000000000005e-06,
"loss": 2.915,
"step": 225
},
{
"epoch": 0.08149959250203749,
"grad_norm": 21.804105758666992,
"learning_rate": 4.960000000000001e-06,
"loss": 2.8108,
"step": 250
},
{
"epoch": 0.08964955175224124,
"grad_norm": 20.4385929107666,
"learning_rate": 5.460000000000001e-06,
"loss": 2.7508,
"step": 275
},
{
"epoch": 0.097799511002445,
"grad_norm": 20.23823356628418,
"learning_rate": 5.9600000000000005e-06,
"loss": 2.7783,
"step": 300
},
{
"epoch": 0.10594947025264874,
"grad_norm": 18.606101989746094,
"learning_rate": 6.460000000000001e-06,
"loss": 2.6393,
"step": 325
},
{
"epoch": 0.11409942950285248,
"grad_norm": 18.78560447692871,
"learning_rate": 6.96e-06,
"loss": 2.4998,
"step": 350
},
{
"epoch": 0.12224938875305623,
"grad_norm": 18.703861236572266,
"learning_rate": 7.4600000000000006e-06,
"loss": 2.5621,
"step": 375
},
{
"epoch": 0.13039934800325997,
"grad_norm": 17.321996688842773,
"learning_rate": 7.960000000000002e-06,
"loss": 2.3896,
"step": 400
},
{
"epoch": 0.13854930725346373,
"grad_norm": 17.575088500976562,
"learning_rate": 8.46e-06,
"loss": 2.4155,
"step": 425
},
{
"epoch": 0.1466992665036675,
"grad_norm": 19.757349014282227,
"learning_rate": 8.96e-06,
"loss": 2.3126,
"step": 450
},
{
"epoch": 0.15484922575387122,
"grad_norm": 16.2225341796875,
"learning_rate": 9.460000000000001e-06,
"loss": 2.2792,
"step": 475
},
{
"epoch": 0.16299918500407498,
"grad_norm": 20.059782028198242,
"learning_rate": 9.960000000000001e-06,
"loss": 2.2343,
"step": 500
},
{
"epoch": 0.17114914425427874,
"grad_norm": 17.442007064819336,
"learning_rate": 9.94888888888889e-06,
"loss": 2.2066,
"step": 525
},
{
"epoch": 0.17929910350448247,
"grad_norm": 18.0509033203125,
"learning_rate": 9.893333333333334e-06,
"loss": 2.1635,
"step": 550
},
{
"epoch": 0.18744906275468623,
"grad_norm": 16.07720375061035,
"learning_rate": 9.837777777777778e-06,
"loss": 2.086,
"step": 575
},
{
"epoch": 0.19559902200489,
"grad_norm": 19.739547729492188,
"learning_rate": 9.782222222222222e-06,
"loss": 2.0068,
"step": 600
},
{
"epoch": 0.20374898125509372,
"grad_norm": 17.144306182861328,
"learning_rate": 9.726666666666668e-06,
"loss": 1.9867,
"step": 625
},
{
"epoch": 0.21189894050529748,
"grad_norm": 17.193174362182617,
"learning_rate": 9.671111111111112e-06,
"loss": 1.9281,
"step": 650
},
{
"epoch": 0.2200488997555012,
"grad_norm": 16.084430694580078,
"learning_rate": 9.615555555555558e-06,
"loss": 1.9546,
"step": 675
},
{
"epoch": 0.22819885900570497,
"grad_norm": 15.92422103881836,
"learning_rate": 9.56e-06,
"loss": 1.9512,
"step": 700
},
{
"epoch": 0.23634881825590873,
"grad_norm": 17.584861755371094,
"learning_rate": 9.504444444444446e-06,
"loss": 1.9593,
"step": 725
},
{
"epoch": 0.24449877750611246,
"grad_norm": 16.338075637817383,
"learning_rate": 9.44888888888889e-06,
"loss": 1.8831,
"step": 750
},
{
"epoch": 0.2526487367563162,
"grad_norm": 16.870452880859375,
"learning_rate": 9.393333333333334e-06,
"loss": 1.888,
"step": 775
},
{
"epoch": 0.26079869600651995,
"grad_norm": 14.461226463317871,
"learning_rate": 9.33777777777778e-06,
"loss": 1.8341,
"step": 800
},
{
"epoch": 0.26894865525672373,
"grad_norm": 18.07695770263672,
"learning_rate": 9.282222222222222e-06,
"loss": 1.854,
"step": 825
},
{
"epoch": 0.27709861450692747,
"grad_norm": 14.117804527282715,
"learning_rate": 9.226666666666668e-06,
"loss": 1.8034,
"step": 850
},
{
"epoch": 0.2852485737571312,
"grad_norm": 17.753490447998047,
"learning_rate": 9.171111111111112e-06,
"loss": 1.7948,
"step": 875
},
{
"epoch": 0.293398533007335,
"grad_norm": 15.982026100158691,
"learning_rate": 9.115555555555556e-06,
"loss": 1.7752,
"step": 900
},
{
"epoch": 0.3015484922575387,
"grad_norm": 15.503934860229492,
"learning_rate": 9.060000000000001e-06,
"loss": 1.7726,
"step": 925
},
{
"epoch": 0.30969845150774244,
"grad_norm": 18.943740844726562,
"learning_rate": 9.004444444444445e-06,
"loss": 1.773,
"step": 950
},
{
"epoch": 0.31784841075794623,
"grad_norm": 14.561786651611328,
"learning_rate": 8.94888888888889e-06,
"loss": 1.7569,
"step": 975
},
{
"epoch": 0.32599837000814996,
"grad_norm": 16.375633239746094,
"learning_rate": 8.893333333333333e-06,
"loss": 1.7363,
"step": 1000
},
{
"epoch": 0.32599837000814996,
"eval_loss": 0.9845153093338013,
"eval_runtime": 457.2834,
"eval_samples_per_second": 8.531,
"eval_steps_per_second": 0.534,
"eval_wer": 0.6733314015261277,
"step": 1000
},
{
"epoch": 0.3341483292583537,
"grad_norm": 14.428193092346191,
"learning_rate": 8.83777777777778e-06,
"loss": 1.731,
"step": 1025
},
{
"epoch": 0.3422982885085575,
"grad_norm": 15.243473052978516,
"learning_rate": 8.782222222222223e-06,
"loss": 1.7255,
"step": 1050
},
{
"epoch": 0.3504482477587612,
"grad_norm": 17.20938491821289,
"learning_rate": 8.726666666666667e-06,
"loss": 1.726,
"step": 1075
},
{
"epoch": 0.35859820700896494,
"grad_norm": 14.864399909973145,
"learning_rate": 8.671111111111113e-06,
"loss": 1.6554,
"step": 1100
},
{
"epoch": 0.36674816625916873,
"grad_norm": 15.314199447631836,
"learning_rate": 8.615555555555555e-06,
"loss": 1.7178,
"step": 1125
},
{
"epoch": 0.37489812550937246,
"grad_norm": 17.271549224853516,
"learning_rate": 8.560000000000001e-06,
"loss": 1.7197,
"step": 1150
},
{
"epoch": 0.3830480847595762,
"grad_norm": 16.468917846679688,
"learning_rate": 8.504444444444445e-06,
"loss": 1.677,
"step": 1175
},
{
"epoch": 0.39119804400978,
"grad_norm": 17.015993118286133,
"learning_rate": 8.448888888888889e-06,
"loss": 1.7056,
"step": 1200
},
{
"epoch": 0.3993480032599837,
"grad_norm": 16.396427154541016,
"learning_rate": 8.393333333333335e-06,
"loss": 1.6895,
"step": 1225
},
{
"epoch": 0.40749796251018744,
"grad_norm": 15.659330368041992,
"learning_rate": 8.337777777777777e-06,
"loss": 1.6117,
"step": 1250
},
{
"epoch": 0.4156479217603912,
"grad_norm": 18.017431259155273,
"learning_rate": 8.282222222222223e-06,
"loss": 1.6787,
"step": 1275
},
{
"epoch": 0.42379788101059496,
"grad_norm": 16.356630325317383,
"learning_rate": 8.226666666666667e-06,
"loss": 1.6565,
"step": 1300
},
{
"epoch": 0.4319478402607987,
"grad_norm": 14.949164390563965,
"learning_rate": 8.171111111111113e-06,
"loss": 1.5925,
"step": 1325
},
{
"epoch": 0.4400977995110024,
"grad_norm": 14.068632125854492,
"learning_rate": 8.115555555555557e-06,
"loss": 1.5924,
"step": 1350
},
{
"epoch": 0.4482477587612062,
"grad_norm": 17.71388816833496,
"learning_rate": 8.06e-06,
"loss": 1.6527,
"step": 1375
},
{
"epoch": 0.45639771801140994,
"grad_norm": 15.577716827392578,
"learning_rate": 8.004444444444445e-06,
"loss": 1.6216,
"step": 1400
},
{
"epoch": 0.46454767726161367,
"grad_norm": 14.626981735229492,
"learning_rate": 7.948888888888889e-06,
"loss": 1.5345,
"step": 1425
},
{
"epoch": 0.47269763651181745,
"grad_norm": 16.892454147338867,
"learning_rate": 7.893333333333335e-06,
"loss": 1.5763,
"step": 1450
},
{
"epoch": 0.4808475957620212,
"grad_norm": 18.04649543762207,
"learning_rate": 7.837777777777779e-06,
"loss": 1.6572,
"step": 1475
},
{
"epoch": 0.4889975550122249,
"grad_norm": 15.368581771850586,
"learning_rate": 7.782222222222223e-06,
"loss": 1.5422,
"step": 1500
},
{
"epoch": 0.4971475142624287,
"grad_norm": 14.239161491394043,
"learning_rate": 7.726666666666667e-06,
"loss": 1.5459,
"step": 1525
},
{
"epoch": 0.5052974735126324,
"grad_norm": 16.011531829833984,
"learning_rate": 7.67111111111111e-06,
"loss": 1.5944,
"step": 1550
},
{
"epoch": 0.5134474327628362,
"grad_norm": 15.135168075561523,
"learning_rate": 7.6155555555555564e-06,
"loss": 1.5705,
"step": 1575
},
{
"epoch": 0.5215973920130399,
"grad_norm": 17.60437774658203,
"learning_rate": 7.5600000000000005e-06,
"loss": 1.4844,
"step": 1600
},
{
"epoch": 0.5297473512632437,
"grad_norm": 16.583059310913086,
"learning_rate": 7.504444444444445e-06,
"loss": 1.5222,
"step": 1625
},
{
"epoch": 0.5378973105134475,
"grad_norm": 16.068408966064453,
"learning_rate": 7.44888888888889e-06,
"loss": 1.5425,
"step": 1650
},
{
"epoch": 0.5460472697636511,
"grad_norm": 14.316849708557129,
"learning_rate": 7.393333333333333e-06,
"loss": 1.4927,
"step": 1675
},
{
"epoch": 0.5541972290138549,
"grad_norm": 14.973306655883789,
"learning_rate": 7.337777777777778e-06,
"loss": 1.4961,
"step": 1700
},
{
"epoch": 0.5623471882640587,
"grad_norm": 16.13720703125,
"learning_rate": 7.282222222222222e-06,
"loss": 1.5945,
"step": 1725
},
{
"epoch": 0.5704971475142624,
"grad_norm": 15.234308242797852,
"learning_rate": 7.226666666666667e-06,
"loss": 1.5252,
"step": 1750
},
{
"epoch": 0.5786471067644662,
"grad_norm": 15.829686164855957,
"learning_rate": 7.171111111111112e-06,
"loss": 1.5186,
"step": 1775
},
{
"epoch": 0.58679706601467,
"grad_norm": 15.698533058166504,
"learning_rate": 7.115555555555557e-06,
"loss": 1.5347,
"step": 1800
},
{
"epoch": 0.5949470252648736,
"grad_norm": 16.133045196533203,
"learning_rate": 7.06e-06,
"loss": 1.538,
"step": 1825
},
{
"epoch": 0.6030969845150774,
"grad_norm": 15.596349716186523,
"learning_rate": 7.006666666666667e-06,
"loss": 1.4263,
"step": 1850
},
{
"epoch": 0.6112469437652812,
"grad_norm": 14.529706001281738,
"learning_rate": 6.951111111111112e-06,
"loss": 1.4455,
"step": 1875
},
{
"epoch": 0.6193969030154849,
"grad_norm": 16.734785079956055,
"learning_rate": 6.8955555555555565e-06,
"loss": 1.4879,
"step": 1900
},
{
"epoch": 0.6275468622656887,
"grad_norm": 14.141518592834473,
"learning_rate": 6.8400000000000014e-06,
"loss": 1.4886,
"step": 1925
},
{
"epoch": 0.6356968215158925,
"grad_norm": 13.811936378479004,
"learning_rate": 6.784444444444445e-06,
"loss": 1.4527,
"step": 1950
},
{
"epoch": 0.6438467807660961,
"grad_norm": 13.9077787399292,
"learning_rate": 6.7288888888888895e-06,
"loss": 1.4277,
"step": 1975
},
{
"epoch": 0.6519967400162999,
"grad_norm": 15.057381629943848,
"learning_rate": 6.6733333333333335e-06,
"loss": 1.4155,
"step": 2000
},
{
"epoch": 0.6519967400162999,
"eval_loss": 0.8236681818962097,
"eval_runtime": 437.2458,
"eval_samples_per_second": 8.922,
"eval_steps_per_second": 0.558,
"eval_wer": 0.571694194919347,
"step": 2000
},
{
"epoch": 0.6601466992665037,
"grad_norm": 15.232809066772461,
"learning_rate": 6.617777777777778e-06,
"loss": 1.4386,
"step": 2025
},
{
"epoch": 0.6682966585167074,
"grad_norm": 13.116307258605957,
"learning_rate": 6.562222222222223e-06,
"loss": 1.4113,
"step": 2050
},
{
"epoch": 0.6764466177669112,
"grad_norm": 17.645702362060547,
"learning_rate": 6.5066666666666665e-06,
"loss": 1.5237,
"step": 2075
},
{
"epoch": 0.684596577017115,
"grad_norm": 15.816476821899414,
"learning_rate": 6.451111111111111e-06,
"loss": 1.4043,
"step": 2100
},
{
"epoch": 0.6927465362673186,
"grad_norm": 17.2808780670166,
"learning_rate": 6.395555555555556e-06,
"loss": 1.4642,
"step": 2125
},
{
"epoch": 0.7008964955175224,
"grad_norm": 16.934045791625977,
"learning_rate": 6.34e-06,
"loss": 1.5066,
"step": 2150
},
{
"epoch": 0.7090464547677262,
"grad_norm": 16.353330612182617,
"learning_rate": 6.284444444444445e-06,
"loss": 1.4638,
"step": 2175
},
{
"epoch": 0.7171964140179299,
"grad_norm": 13.666918754577637,
"learning_rate": 6.22888888888889e-06,
"loss": 1.4241,
"step": 2200
},
{
"epoch": 0.7253463732681337,
"grad_norm": 16.40180206298828,
"learning_rate": 6.173333333333333e-06,
"loss": 1.4109,
"step": 2225
},
{
"epoch": 0.7334963325183375,
"grad_norm": 15.645243644714355,
"learning_rate": 6.117777777777778e-06,
"loss": 1.4354,
"step": 2250
},
{
"epoch": 0.7416462917685411,
"grad_norm": 14.662041664123535,
"learning_rate": 6.062222222222223e-06,
"loss": 1.4034,
"step": 2275
},
{
"epoch": 0.7497962510187449,
"grad_norm": 17.322463989257812,
"learning_rate": 6.006666666666667e-06,
"loss": 1.4348,
"step": 2300
},
{
"epoch": 0.7579462102689487,
"grad_norm": 15.498252868652344,
"learning_rate": 5.951111111111112e-06,
"loss": 1.4353,
"step": 2325
},
{
"epoch": 0.7660961695191524,
"grad_norm": 15.99593448638916,
"learning_rate": 5.895555555555557e-06,
"loss": 1.3877,
"step": 2350
},
{
"epoch": 0.7742461287693562,
"grad_norm": 15.517157554626465,
"learning_rate": 5.84e-06,
"loss": 1.4029,
"step": 2375
},
{
"epoch": 0.78239608801956,
"grad_norm": 14.675141334533691,
"learning_rate": 5.784444444444445e-06,
"loss": 1.4366,
"step": 2400
},
{
"epoch": 0.7905460472697636,
"grad_norm": 16.114215850830078,
"learning_rate": 5.72888888888889e-06,
"loss": 1.4032,
"step": 2425
},
{
"epoch": 0.7986960065199674,
"grad_norm": 15.094751358032227,
"learning_rate": 5.673333333333334e-06,
"loss": 1.4116,
"step": 2450
},
{
"epoch": 0.8068459657701712,
"grad_norm": 15.857089042663574,
"learning_rate": 5.617777777777779e-06,
"loss": 1.3738,
"step": 2475
},
{
"epoch": 0.8149959250203749,
"grad_norm": 16.033899307250977,
"learning_rate": 5.562222222222222e-06,
"loss": 1.4226,
"step": 2500
},
{
"epoch": 0.8231458842705787,
"grad_norm": 13.293587684631348,
"learning_rate": 5.506666666666667e-06,
"loss": 1.3407,
"step": 2525
},
{
"epoch": 0.8312958435207825,
"grad_norm": 14.656893730163574,
"learning_rate": 5.451111111111112e-06,
"loss": 1.3327,
"step": 2550
},
{
"epoch": 0.8394458027709861,
"grad_norm": 15.4556245803833,
"learning_rate": 5.3955555555555565e-06,
"loss": 1.3217,
"step": 2575
},
{
"epoch": 0.8475957620211899,
"grad_norm": 13.271966934204102,
"learning_rate": 5.3400000000000005e-06,
"loss": 1.3906,
"step": 2600
},
{
"epoch": 0.8557457212713936,
"grad_norm": 16.06142234802246,
"learning_rate": 5.2844444444444454e-06,
"loss": 1.3629,
"step": 2625
},
{
"epoch": 0.8638956805215974,
"grad_norm": 16.935834884643555,
"learning_rate": 5.228888888888889e-06,
"loss": 1.4123,
"step": 2650
},
{
"epoch": 0.8720456397718012,
"grad_norm": 16.714982986450195,
"learning_rate": 5.1733333333333335e-06,
"loss": 1.3887,
"step": 2675
},
{
"epoch": 0.8801955990220048,
"grad_norm": 13.961587905883789,
"learning_rate": 5.117777777777778e-06,
"loss": 1.3757,
"step": 2700
},
{
"epoch": 0.8883455582722086,
"grad_norm": 14.335005760192871,
"learning_rate": 5.062222222222222e-06,
"loss": 1.3393,
"step": 2725
},
{
"epoch": 0.8964955175224124,
"grad_norm": 13.483233451843262,
"learning_rate": 5.006666666666667e-06,
"loss": 1.3782,
"step": 2750
},
{
"epoch": 0.9046454767726161,
"grad_norm": 15.308544158935547,
"learning_rate": 4.951111111111111e-06,
"loss": 1.3422,
"step": 2775
},
{
"epoch": 0.9127954360228199,
"grad_norm": 11.132999420166016,
"learning_rate": 4.895555555555556e-06,
"loss": 1.3534,
"step": 2800
},
{
"epoch": 0.9209453952730237,
"grad_norm": 15.973653793334961,
"learning_rate": 4.84e-06,
"loss": 1.4114,
"step": 2825
},
{
"epoch": 0.9290953545232273,
"grad_norm": 14.977301597595215,
"learning_rate": 4.784444444444445e-06,
"loss": 1.3283,
"step": 2850
},
{
"epoch": 0.9372453137734311,
"grad_norm": 16.090097427368164,
"learning_rate": 4.728888888888889e-06,
"loss": 1.2659,
"step": 2875
},
{
"epoch": 0.9453952730236349,
"grad_norm": 13.608882904052734,
"learning_rate": 4.673333333333333e-06,
"loss": 1.2446,
"step": 2900
},
{
"epoch": 0.9535452322738386,
"grad_norm": 15.259020805358887,
"learning_rate": 4.617777777777778e-06,
"loss": 1.2993,
"step": 2925
},
{
"epoch": 0.9616951915240424,
"grad_norm": 13.211849212646484,
"learning_rate": 4.562222222222222e-06,
"loss": 1.3024,
"step": 2950
},
{
"epoch": 0.9698451507742462,
"grad_norm": 16.002653121948242,
"learning_rate": 4.506666666666667e-06,
"loss": 1.3546,
"step": 2975
},
{
"epoch": 0.9779951100244498,
"grad_norm": 13.069381713867188,
"learning_rate": 4.451111111111112e-06,
"loss": 1.3809,
"step": 3000
},
{
"epoch": 0.9779951100244498,
"eval_loss": 0.7566213607788086,
"eval_runtime": 438.5427,
"eval_samples_per_second": 8.895,
"eval_steps_per_second": 0.556,
"eval_wer": 0.5274075147300299,
"step": 3000
},
{
"epoch": 0.9861450692746536,
"grad_norm": 14.120579719543457,
"learning_rate": 4.395555555555556e-06,
"loss": 1.3394,
"step": 3025
},
{
"epoch": 0.9942950285248574,
"grad_norm": 20.488895416259766,
"learning_rate": 4.34e-06,
"loss": 1.3911,
"step": 3050
},
{
"epoch": 1.0024449877750612,
"grad_norm": 13.28896713256836,
"learning_rate": 4.284444444444445e-06,
"loss": 1.2861,
"step": 3075
},
{
"epoch": 1.0105949470252649,
"grad_norm": 13.970645904541016,
"learning_rate": 4.228888888888889e-06,
"loss": 1.146,
"step": 3100
},
{
"epoch": 1.0187449062754685,
"grad_norm": 13.756807327270508,
"learning_rate": 4.173333333333334e-06,
"loss": 1.1688,
"step": 3125
},
{
"epoch": 1.0268948655256724,
"grad_norm": 14.576444625854492,
"learning_rate": 4.117777777777779e-06,
"loss": 1.1595,
"step": 3150
},
{
"epoch": 1.0350448247758761,
"grad_norm": 12.942944526672363,
"learning_rate": 4.062222222222223e-06,
"loss": 1.2039,
"step": 3175
},
{
"epoch": 1.0431947840260798,
"grad_norm": 13.682048797607422,
"learning_rate": 4.006666666666667e-06,
"loss": 1.18,
"step": 3200
},
{
"epoch": 1.0513447432762837,
"grad_norm": 14.890068054199219,
"learning_rate": 3.951111111111112e-06,
"loss": 1.1905,
"step": 3225
},
{
"epoch": 1.0594947025264874,
"grad_norm": 14.280926704406738,
"learning_rate": 3.895555555555556e-06,
"loss": 1.1572,
"step": 3250
},
{
"epoch": 1.067644661776691,
"grad_norm": 12.664419174194336,
"learning_rate": 3.8400000000000005e-06,
"loss": 1.1405,
"step": 3275
},
{
"epoch": 1.075794621026895,
"grad_norm": 13.912672996520996,
"learning_rate": 3.784444444444445e-06,
"loss": 1.1779,
"step": 3300
},
{
"epoch": 1.0839445802770986,
"grad_norm": 13.749354362487793,
"learning_rate": 3.728888888888889e-06,
"loss": 1.1534,
"step": 3325
},
{
"epoch": 1.0920945395273023,
"grad_norm": 13.6649169921875,
"learning_rate": 3.673333333333334e-06,
"loss": 1.1381,
"step": 3350
},
{
"epoch": 1.1002444987775062,
"grad_norm": 13.478617668151855,
"learning_rate": 3.617777777777778e-06,
"loss": 1.2095,
"step": 3375
},
{
"epoch": 1.1083944580277099,
"grad_norm": 13.082009315490723,
"learning_rate": 3.5622222222222224e-06,
"loss": 1.1965,
"step": 3400
},
{
"epoch": 1.1165444172779135,
"grad_norm": 13.995879173278809,
"learning_rate": 3.5066666666666673e-06,
"loss": 1.1418,
"step": 3425
},
{
"epoch": 1.1246943765281174,
"grad_norm": 13.161800384521484,
"learning_rate": 3.4511111111111113e-06,
"loss": 1.1888,
"step": 3450
},
{
"epoch": 1.132844335778321,
"grad_norm": 12.4346342086792,
"learning_rate": 3.3955555555555558e-06,
"loss": 1.1425,
"step": 3475
},
{
"epoch": 1.1409942950285248,
"grad_norm": 14.341651916503906,
"learning_rate": 3.3400000000000006e-06,
"loss": 1.136,
"step": 3500
},
{
"epoch": 1.1491442542787287,
"grad_norm": 11.676952362060547,
"learning_rate": 3.2844444444444447e-06,
"loss": 1.1345,
"step": 3525
},
{
"epoch": 1.1572942135289324,
"grad_norm": 14.452948570251465,
"learning_rate": 3.228888888888889e-06,
"loss": 1.1602,
"step": 3550
},
{
"epoch": 1.165444172779136,
"grad_norm": 13.384020805358887,
"learning_rate": 3.173333333333334e-06,
"loss": 1.159,
"step": 3575
},
{
"epoch": 1.17359413202934,
"grad_norm": 13.295368194580078,
"learning_rate": 3.117777777777778e-06,
"loss": 1.1146,
"step": 3600
},
{
"epoch": 1.1817440912795436,
"grad_norm": 13.34139633178711,
"learning_rate": 3.0622222222222225e-06,
"loss": 1.1496,
"step": 3625
},
{
"epoch": 1.1898940505297473,
"grad_norm": 15.428239822387695,
"learning_rate": 3.0066666666666674e-06,
"loss": 1.2243,
"step": 3650
},
{
"epoch": 1.1980440097799512,
"grad_norm": 13.666820526123047,
"learning_rate": 2.9511111111111114e-06,
"loss": 1.1792,
"step": 3675
},
{
"epoch": 1.2061939690301549,
"grad_norm": 11.906725883483887,
"learning_rate": 2.895555555555556e-06,
"loss": 1.135,
"step": 3700
},
{
"epoch": 1.2143439282803585,
"grad_norm": 14.56628131866455,
"learning_rate": 2.84e-06,
"loss": 1.134,
"step": 3725
},
{
"epoch": 1.2224938875305624,
"grad_norm": 14.18961238861084,
"learning_rate": 2.784444444444445e-06,
"loss": 1.1115,
"step": 3750
},
{
"epoch": 1.230643846780766,
"grad_norm": 14.074260711669922,
"learning_rate": 2.7288888888888893e-06,
"loss": 1.1592,
"step": 3775
},
{
"epoch": 1.2387938060309698,
"grad_norm": 14.50068473815918,
"learning_rate": 2.6733333333333333e-06,
"loss": 1.1568,
"step": 3800
},
{
"epoch": 1.2469437652811737,
"grad_norm": 14.552042007446289,
"learning_rate": 2.617777777777778e-06,
"loss": 1.1456,
"step": 3825
},
{
"epoch": 1.2550937245313774,
"grad_norm": 13.851117134094238,
"learning_rate": 2.5622222222222226e-06,
"loss": 1.1469,
"step": 3850
},
{
"epoch": 1.263243683781581,
"grad_norm": 13.306814193725586,
"learning_rate": 2.5066666666666667e-06,
"loss": 1.1595,
"step": 3875
},
{
"epoch": 1.271393643031785,
"grad_norm": 14.614799499511719,
"learning_rate": 2.451111111111111e-06,
"loss": 1.148,
"step": 3900
},
{
"epoch": 1.2795436022819886,
"grad_norm": 13.00090217590332,
"learning_rate": 2.3955555555555556e-06,
"loss": 1.1127,
"step": 3925
},
{
"epoch": 1.2876935615321923,
"grad_norm": 12.845821380615234,
"learning_rate": 2.3400000000000005e-06,
"loss": 1.108,
"step": 3950
},
{
"epoch": 1.295843520782396,
"grad_norm": 13.80776596069336,
"learning_rate": 2.2844444444444445e-06,
"loss": 1.1851,
"step": 3975
},
{
"epoch": 1.3039934800325998,
"grad_norm": 12.853269577026367,
"learning_rate": 2.228888888888889e-06,
"loss": 1.0882,
"step": 4000
},
{
"epoch": 1.3039934800325998,
"eval_loss": 0.7265446186065674,
"eval_runtime": 440.5369,
"eval_samples_per_second": 8.855,
"eval_steps_per_second": 0.554,
"eval_wer": 0.5286148942335555,
"step": 4000
},
{
"epoch": 1.3121434392828035,
"grad_norm": 13.121480941772461,
"learning_rate": 2.1733333333333334e-06,
"loss": 1.1081,
"step": 4025
},
{
"epoch": 1.3202933985330074,
"grad_norm": 14.560317039489746,
"learning_rate": 2.117777777777778e-06,
"loss": 1.1418,
"step": 4050
},
{
"epoch": 1.328443357783211,
"grad_norm": 13.478865623474121,
"learning_rate": 2.0622222222222223e-06,
"loss": 1.1296,
"step": 4075
},
{
"epoch": 1.3365933170334148,
"grad_norm": 12.93165397644043,
"learning_rate": 2.006666666666667e-06,
"loss": 1.1609,
"step": 4100
},
{
"epoch": 1.3447432762836184,
"grad_norm": 13.513744354248047,
"learning_rate": 1.9511111111111113e-06,
"loss": 1.099,
"step": 4125
},
{
"epoch": 1.3528932355338223,
"grad_norm": 13.625751495361328,
"learning_rate": 1.8955555555555557e-06,
"loss": 1.0996,
"step": 4150
},
{
"epoch": 1.361043194784026,
"grad_norm": 13.141559600830078,
"learning_rate": 1.8400000000000002e-06,
"loss": 1.1522,
"step": 4175
},
{
"epoch": 1.36919315403423,
"grad_norm": 13.436763763427734,
"learning_rate": 1.7844444444444444e-06,
"loss": 1.188,
"step": 4200
},
{
"epoch": 1.3773431132844336,
"grad_norm": 13.628213882446289,
"learning_rate": 1.728888888888889e-06,
"loss": 1.0837,
"step": 4225
},
{
"epoch": 1.3854930725346373,
"grad_norm": 12.989063262939453,
"learning_rate": 1.6733333333333335e-06,
"loss": 1.1027,
"step": 4250
},
{
"epoch": 1.393643031784841,
"grad_norm": 13.773276329040527,
"learning_rate": 1.6177777777777778e-06,
"loss": 1.1276,
"step": 4275
},
{
"epoch": 1.4017929910350448,
"grad_norm": 13.921293258666992,
"learning_rate": 1.5622222222222225e-06,
"loss": 1.0993,
"step": 4300
},
{
"epoch": 1.4099429502852485,
"grad_norm": 13.18332290649414,
"learning_rate": 1.506666666666667e-06,
"loss": 1.0903,
"step": 4325
},
{
"epoch": 1.4180929095354524,
"grad_norm": 14.134808540344238,
"learning_rate": 1.4511111111111112e-06,
"loss": 1.0599,
"step": 4350
},
{
"epoch": 1.426242868785656,
"grad_norm": 13.665396690368652,
"learning_rate": 1.3955555555555556e-06,
"loss": 1.0808,
"step": 4375
},
{
"epoch": 1.4343928280358598,
"grad_norm": 14.74458122253418,
"learning_rate": 1.34e-06,
"loss": 1.1597,
"step": 4400
},
{
"epoch": 1.4425427872860634,
"grad_norm": 13.300642013549805,
"learning_rate": 1.2844444444444445e-06,
"loss": 1.1345,
"step": 4425
},
{
"epoch": 1.4506927465362673,
"grad_norm": 12.414579391479492,
"learning_rate": 1.228888888888889e-06,
"loss": 1.1019,
"step": 4450
},
{
"epoch": 1.458842705786471,
"grad_norm": 16.401771545410156,
"learning_rate": 1.1733333333333335e-06,
"loss": 1.1428,
"step": 4475
},
{
"epoch": 1.466992665036675,
"grad_norm": 14.687612533569336,
"learning_rate": 1.117777777777778e-06,
"loss": 1.1126,
"step": 4500
},
{
"epoch": 1.4751426242868786,
"grad_norm": 12.354652404785156,
"learning_rate": 1.0622222222222222e-06,
"loss": 1.145,
"step": 4525
},
{
"epoch": 1.4832925835370823,
"grad_norm": 14.474491119384766,
"learning_rate": 1.0066666666666668e-06,
"loss": 1.1362,
"step": 4550
},
{
"epoch": 1.491442542787286,
"grad_norm": 13.144413948059082,
"learning_rate": 9.511111111111111e-07,
"loss": 1.1293,
"step": 4575
},
{
"epoch": 1.4995925020374898,
"grad_norm": 12.695611953735352,
"learning_rate": 8.955555555555557e-07,
"loss": 1.0901,
"step": 4600
},
{
"epoch": 1.5077424612876935,
"grad_norm": 16.430057525634766,
"learning_rate": 8.400000000000001e-07,
"loss": 1.1373,
"step": 4625
},
{
"epoch": 1.5158924205378974,
"grad_norm": 13.695208549499512,
"learning_rate": 7.844444444444445e-07,
"loss": 1.12,
"step": 4650
},
{
"epoch": 1.524042379788101,
"grad_norm": 14.039958000183105,
"learning_rate": 7.28888888888889e-07,
"loss": 1.1394,
"step": 4675
},
{
"epoch": 1.5321923390383048,
"grad_norm": 14.492539405822754,
"learning_rate": 6.733333333333334e-07,
"loss": 1.1385,
"step": 4700
},
{
"epoch": 1.5403422982885084,
"grad_norm": 15.331225395202637,
"learning_rate": 6.177777777777778e-07,
"loss": 1.119,
"step": 4725
},
{
"epoch": 1.5484922575387123,
"grad_norm": 12.952714920043945,
"learning_rate": 5.622222222222223e-07,
"loss": 1.1479,
"step": 4750
},
{
"epoch": 1.556642216788916,
"grad_norm": 13.869738578796387,
"learning_rate": 5.066666666666667e-07,
"loss": 1.0884,
"step": 4775
},
{
"epoch": 1.56479217603912,
"grad_norm": 13.278375625610352,
"learning_rate": 4.511111111111111e-07,
"loss": 1.1049,
"step": 4800
},
{
"epoch": 1.5729421352893236,
"grad_norm": 12.596354484558105,
"learning_rate": 3.9555555555555557e-07,
"loss": 1.0696,
"step": 4825
},
{
"epoch": 1.5810920945395273,
"grad_norm": 13.701985359191895,
"learning_rate": 3.4000000000000003e-07,
"loss": 1.1376,
"step": 4850
},
{
"epoch": 1.589242053789731,
"grad_norm": 12.712823867797852,
"learning_rate": 2.844444444444445e-07,
"loss": 1.1158,
"step": 4875
},
{
"epoch": 1.5973920130399348,
"grad_norm": 12.259002685546875,
"learning_rate": 2.2888888888888892e-07,
"loss": 1.094,
"step": 4900
},
{
"epoch": 1.6055419722901385,
"grad_norm": 14.35509204864502,
"learning_rate": 1.7333333333333335e-07,
"loss": 1.1516,
"step": 4925
},
{
"epoch": 1.6136919315403424,
"grad_norm": 12.94543170928955,
"learning_rate": 1.1777777777777778e-07,
"loss": 1.0555,
"step": 4950
},
{
"epoch": 1.621841890790546,
"grad_norm": 14.701851844787598,
"learning_rate": 6.222222222222223e-08,
"loss": 1.0818,
"step": 4975
},
{
"epoch": 1.6299918500407498,
"grad_norm": 13.011249542236328,
"learning_rate": 6.666666666666667e-09,
"loss": 1.0903,
"step": 5000
},
{
"epoch": 1.6299918500407498,
"eval_loss": 0.7135367393493652,
"eval_runtime": 442.8745,
"eval_samples_per_second": 8.808,
"eval_steps_per_second": 0.551,
"eval_wer": 0.508886313145948,
"step": 5000
},
{
"epoch": 1.6299918500407498,
"step": 5000,
"total_flos": 1.037708050563072e+19,
"train_loss": 1.5516241161346436,
"train_runtime": 17280.0888,
"train_samples_per_second": 9.259,
"train_steps_per_second": 0.289
}
],
"logging_steps": 25,
"max_steps": 5000,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"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": 1.037708050563072e+19,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}