{
"best_metric": 1.18520188331604,
"best_model_checkpoint": "batoula187/wav2vec2-xlsr-arabic2/checkpoint-1800",
"epoch": 17.206766917293233,
"eval_steps": 200,
"global_step": 12200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.42283298097251587,
"grad_norm": 0.43856075406074524,
"learning_rate": 0.00011999999999999999,
"loss": 0.1174,
"step": 200
},
{
"epoch": 0.42283298097251587,
"eval_loss": 1.2291227579116821,
"eval_runtime": 64.2106,
"eval_samples_per_second": 16.321,
"eval_steps_per_second": 2.04,
"eval_wer": 0.6153393487642213,
"step": 200
},
{
"epoch": 0.8456659619450317,
"grad_norm": 5.825935363769531,
"learning_rate": 0.00023999999999999998,
"loss": 0.098,
"step": 400
},
{
"epoch": 0.8456659619450317,
"eval_loss": 1.232483148574829,
"eval_runtime": 64.9125,
"eval_samples_per_second": 16.145,
"eval_steps_per_second": 2.018,
"eval_wer": 0.6275009807767752,
"step": 400
},
{
"epoch": 1.2684989429175475,
"grad_norm": 0.9385707378387451,
"learning_rate": 0.00029780861943024103,
"loss": 0.1301,
"step": 600
},
{
"epoch": 1.2684989429175475,
"eval_loss": 1.1968672275543213,
"eval_runtime": 63.9557,
"eval_samples_per_second": 16.386,
"eval_steps_per_second": 2.048,
"eval_wer": 0.6127893291486858,
"step": 600
},
{
"epoch": 1.6913319238900635,
"grad_norm": 0.5655494332313538,
"learning_rate": 0.00029346968590211834,
"loss": 0.1514,
"step": 800
},
{
"epoch": 1.6913319238900635,
"eval_loss": 1.2292981147766113,
"eval_runtime": 64.6544,
"eval_samples_per_second": 16.209,
"eval_steps_per_second": 2.026,
"eval_wer": 0.6488819144762652,
"step": 800
},
{
"epoch": 2.1141649048625792,
"grad_norm": 1.0750213861465454,
"learning_rate": 0.0002890869247626004,
"loss": 0.1494,
"step": 1000
},
{
"epoch": 2.1141649048625792,
"eval_loss": 1.3062251806259155,
"eval_runtime": 63.5313,
"eval_samples_per_second": 16.496,
"eval_steps_per_second": 2.062,
"eval_wer": 0.670066692820714,
"step": 1000
},
{
"epoch": 2.536997885835095,
"grad_norm": 0.8127875328063965,
"learning_rate": 0.0002847041636230825,
"loss": 0.1382,
"step": 1200
},
{
"epoch": 2.536997885835095,
"eval_loss": 1.2222816944122314,
"eval_runtime": 63.8482,
"eval_samples_per_second": 16.414,
"eval_steps_per_second": 2.052,
"eval_wer": 0.6261278932914869,
"step": 1200
},
{
"epoch": 2.9598308668076108,
"grad_norm": 0.8811041116714478,
"learning_rate": 0.00028032140248356464,
"loss": 0.1382,
"step": 1400
},
{
"epoch": 2.9598308668076108,
"eval_loss": 1.3115968704223633,
"eval_runtime": 65.0526,
"eval_samples_per_second": 16.11,
"eval_steps_per_second": 2.014,
"eval_wer": 0.6506473126716359,
"step": 1400
},
{
"epoch": 3.382663847780127,
"grad_norm": 0.7833051085472107,
"learning_rate": 0.0002759386413440467,
"loss": 0.1239,
"step": 1600
},
{
"epoch": 3.382663847780127,
"eval_loss": 1.1977170705795288,
"eval_runtime": 64.8745,
"eval_samples_per_second": 16.154,
"eval_steps_per_second": 2.019,
"eval_wer": 0.6188701451549627,
"step": 1600
},
{
"epoch": 3.8054968287526427,
"grad_norm": 0.4267895817756653,
"learning_rate": 0.0002715558802045288,
"loss": 0.1228,
"step": 1800
},
{
"epoch": 3.8054968287526427,
"eval_loss": 1.18520188331604,
"eval_runtime": 64.2457,
"eval_samples_per_second": 16.312,
"eval_steps_per_second": 2.039,
"eval_wer": 0.6280894468418988,
"step": 1800
},
{
"epoch": 4.2283298097251585,
"grad_norm": 0.7231135964393616,
"learning_rate": 0.00026717311906501094,
"loss": 0.1117,
"step": 2000
},
{
"epoch": 4.2283298097251585,
"eval_loss": 1.3370016813278198,
"eval_runtime": 64.7812,
"eval_samples_per_second": 16.178,
"eval_steps_per_second": 2.022,
"eval_wer": 0.6494703805413887,
"step": 2000
},
{
"epoch": 4.651162790697675,
"grad_norm": 0.40635955333709717,
"learning_rate": 0.000262790357925493,
"loss": 0.1118,
"step": 2200
},
{
"epoch": 4.651162790697675,
"eval_loss": 1.3265222311019897,
"eval_runtime": 63.5949,
"eval_samples_per_second": 16.479,
"eval_steps_per_second": 2.06,
"eval_wer": 0.6431934091800706,
"step": 2200
},
{
"epoch": 5.07399577167019,
"grad_norm": 0.9417168498039246,
"learning_rate": 0.0002584075967859751,
"loss": 0.1101,
"step": 2400
},
{
"epoch": 5.07399577167019,
"eval_loss": 1.345849633216858,
"eval_runtime": 64.4165,
"eval_samples_per_second": 16.269,
"eval_steps_per_second": 2.034,
"eval_wer": 0.6310317771675167,
"step": 2400
},
{
"epoch": 5.496828752642706,
"grad_norm": 0.7866037487983704,
"learning_rate": 0.00025402483564645724,
"loss": 0.1328,
"step": 2600
},
{
"epoch": 5.496828752642706,
"eval_loss": 1.2545326948165894,
"eval_runtime": 64.0398,
"eval_samples_per_second": 16.365,
"eval_steps_per_second": 2.046,
"eval_wer": 0.6341702628481758,
"step": 2600
},
{
"epoch": 5.9196617336152215,
"grad_norm": 0.47020798921585083,
"learning_rate": 0.00024964207450693936,
"loss": 0.1384,
"step": 2800
},
{
"epoch": 5.9196617336152215,
"eval_loss": 1.2806007862091064,
"eval_runtime": 64.3065,
"eval_samples_per_second": 16.297,
"eval_steps_per_second": 2.037,
"eval_wer": 0.6265202040015693,
"step": 2800
},
{
"epoch": 6.342494714587738,
"grad_norm": 0.605645477771759,
"learning_rate": 0.0002452593133674214,
"loss": 0.1334,
"step": 3000
},
{
"epoch": 6.342494714587738,
"eval_loss": 1.2484089136123657,
"eval_runtime": 65.2815,
"eval_samples_per_second": 16.054,
"eval_steps_per_second": 2.007,
"eval_wer": 0.6369164378187524,
"step": 3000
},
{
"epoch": 6.765327695560254,
"grad_norm": 0.6708455085754395,
"learning_rate": 0.00024087655222790357,
"loss": 0.1383,
"step": 3200
},
{
"epoch": 6.765327695560254,
"eval_loss": 1.2701318264007568,
"eval_runtime": 63.6121,
"eval_samples_per_second": 16.475,
"eval_steps_per_second": 2.059,
"eval_wer": 0.6479011377010593,
"step": 3200
},
{
"epoch": 7.188160676532769,
"grad_norm": 0.3204992711544037,
"learning_rate": 0.00023649379108838566,
"loss": 0.1281,
"step": 3400
},
{
"epoch": 7.188160676532769,
"eval_loss": 1.1926395893096924,
"eval_runtime": 65.1953,
"eval_samples_per_second": 16.075,
"eval_steps_per_second": 2.009,
"eval_wer": 0.6314240878775991,
"step": 3400
},
{
"epoch": 7.6109936575052854,
"grad_norm": 2.2028682231903076,
"learning_rate": 0.00023211102994886775,
"loss": 0.1232,
"step": 3600
},
{
"epoch": 7.6109936575052854,
"eval_loss": 1.2255371809005737,
"eval_runtime": 65.3961,
"eval_samples_per_second": 16.025,
"eval_steps_per_second": 2.003,
"eval_wer": 0.6186739897999215,
"step": 3600
},
{
"epoch": 8.033826638477802,
"grad_norm": 0.33132538199424744,
"learning_rate": 2e-05,
"loss": 0.0727,
"step": 3800
},
{
"epoch": 8.033826638477802,
"eval_loss": 1.23982572555542,
"eval_runtime": 66.6836,
"eval_samples_per_second": 15.716,
"eval_steps_per_second": 1.575,
"eval_wer": 0.6014123185562966,
"step": 3800
},
{
"epoch": 8.456659619450317,
"grad_norm": 0.3985452950000763,
"learning_rate": 4e-05,
"loss": 0.0749,
"step": 4000
},
{
"epoch": 8.456659619450317,
"eval_loss": 1.2319424152374268,
"eval_runtime": 67.0281,
"eval_samples_per_second": 15.635,
"eval_steps_per_second": 1.567,
"eval_wer": 0.595723813260102,
"step": 4000
},
{
"epoch": 8.879492600422832,
"grad_norm": 0.5187695622444153,
"learning_rate": 4.955849889624724e-05,
"loss": 0.0734,
"step": 4200
},
{
"epoch": 8.879492600422832,
"eval_loss": 1.2246508598327637,
"eval_runtime": 66.6194,
"eval_samples_per_second": 15.731,
"eval_steps_per_second": 1.576,
"eval_wer": 0.5878775990584543,
"step": 4200
},
{
"epoch": 9.30232558139535,
"grad_norm": 0.3908683955669403,
"learning_rate": 4.867549668874172e-05,
"loss": 0.0684,
"step": 4400
},
{
"epoch": 9.30232558139535,
"eval_loss": 1.347394585609436,
"eval_runtime": 65.574,
"eval_samples_per_second": 15.982,
"eval_steps_per_second": 1.601,
"eval_wer": 0.6135739505688506,
"step": 4400
},
{
"epoch": 9.725158562367865,
"grad_norm": 1.09392511844635,
"learning_rate": 4.779690949227373e-05,
"loss": 0.073,
"step": 4600
},
{
"epoch": 9.725158562367865,
"eval_loss": 1.2836501598358154,
"eval_runtime": 76.5173,
"eval_samples_per_second": 13.696,
"eval_steps_per_second": 1.372,
"eval_wer": 0.5935661043546488,
"step": 4600
},
{
"epoch": 10.14799154334038,
"grad_norm": 0.786432683467865,
"learning_rate": 4.691390728476822e-05,
"loss": 0.0728,
"step": 4800
},
{
"epoch": 10.14799154334038,
"eval_loss": 1.247693657875061,
"eval_runtime": 66.6059,
"eval_samples_per_second": 15.734,
"eval_steps_per_second": 1.576,
"eval_wer": 0.5910160847391134,
"step": 4800
},
{
"epoch": 10.570824524312897,
"grad_norm": 0.22806741297245026,
"learning_rate": 4.6030905077262693e-05,
"loss": 0.0718,
"step": 5000
},
{
"epoch": 10.570824524312897,
"eval_loss": 1.2471730709075928,
"eval_runtime": 67.4477,
"eval_samples_per_second": 15.538,
"eval_steps_per_second": 1.557,
"eval_wer": 0.5867006669282071,
"step": 5000
},
{
"epoch": 10.993657505285412,
"grad_norm": 1.541914463043213,
"learning_rate": 4.5147902869757175e-05,
"loss": 0.0685,
"step": 5200
},
{
"epoch": 10.993657505285412,
"eval_loss": 1.2693225145339966,
"eval_runtime": 65.4788,
"eval_samples_per_second": 16.005,
"eval_steps_per_second": 1.604,
"eval_wer": 0.5788544527265594,
"step": 5200
},
{
"epoch": 11.416490486257928,
"grad_norm": 0.4897485673427582,
"learning_rate": 4.4264900662251656e-05,
"loss": 0.0649,
"step": 5400
},
{
"epoch": 11.416490486257928,
"eval_loss": 1.2164980173110962,
"eval_runtime": 66.1024,
"eval_samples_per_second": 15.854,
"eval_steps_per_second": 1.588,
"eval_wer": 0.5786582973715182,
"step": 5400
},
{
"epoch": 11.839323467230443,
"grad_norm": 0.269406795501709,
"learning_rate": 4.338189845474614e-05,
"loss": 0.0632,
"step": 5600
},
{
"epoch": 11.839323467230443,
"eval_loss": 1.2446550130844116,
"eval_runtime": 69.6183,
"eval_samples_per_second": 15.054,
"eval_steps_per_second": 1.508,
"eval_wer": 0.5841506473126716,
"step": 5600
},
{
"epoch": 12.26215644820296,
"grad_norm": 0.6048020720481873,
"learning_rate": 4.249889624724062e-05,
"loss": 0.0625,
"step": 5800
},
{
"epoch": 12.26215644820296,
"eval_loss": 1.308754324913025,
"eval_runtime": 67.8481,
"eval_samples_per_second": 15.446,
"eval_steps_per_second": 1.548,
"eval_wer": 0.5806198509219301,
"step": 5800
},
{
"epoch": 12.684989429175475,
"grad_norm": 0.2880701720714569,
"learning_rate": 4.16158940397351e-05,
"loss": 0.061,
"step": 6000
},
{
"epoch": 12.684989429175475,
"eval_loss": 1.3398616313934326,
"eval_runtime": 67.8965,
"eval_samples_per_second": 15.435,
"eval_steps_per_second": 1.546,
"eval_wer": 0.5923891722244017,
"step": 6000
},
{
"epoch": 13.10782241014799,
"grad_norm": 0.419969767332077,
"learning_rate": 4.073289183222958e-05,
"loss": 0.0595,
"step": 6200
},
{
"epoch": 13.10782241014799,
"eval_loss": 1.3048608303070068,
"eval_runtime": 66.7124,
"eval_samples_per_second": 15.709,
"eval_steps_per_second": 1.574,
"eval_wer": 0.5768928991761475,
"step": 6200
},
{
"epoch": 13.530655391120508,
"grad_norm": 0.17939791083335876,
"learning_rate": 3.9849889624724064e-05,
"loss": 0.0608,
"step": 6400
},
{
"epoch": 13.530655391120508,
"eval_loss": 1.2737226486206055,
"eval_runtime": 67.084,
"eval_samples_per_second": 15.622,
"eval_steps_per_second": 1.565,
"eval_wer": 0.573362102785406,
"step": 6400
},
{
"epoch": 13.953488372093023,
"grad_norm": 0.32987338304519653,
"learning_rate": 3.8966887417218545e-05,
"loss": 0.0596,
"step": 6600
},
{
"epoch": 13.953488372093023,
"eval_loss": 1.228838324546814,
"eval_runtime": 66.3842,
"eval_samples_per_second": 15.787,
"eval_steps_per_second": 1.582,
"eval_wer": 0.5747351902706944,
"step": 6600
},
{
"epoch": 14.376321353065538,
"grad_norm": 1.2768590450286865,
"learning_rate": 3.808388520971303e-05,
"loss": 0.0565,
"step": 6800
},
{
"epoch": 14.376321353065538,
"eval_loss": 1.2599496841430664,
"eval_runtime": 66.2771,
"eval_samples_per_second": 15.812,
"eval_steps_per_second": 1.584,
"eval_wer": 0.5676735974892114,
"step": 6800
},
{
"epoch": 14.799154334038056,
"grad_norm": 0.410118043422699,
"learning_rate": 3.720088300220751e-05,
"loss": 0.0568,
"step": 7000
},
{
"epoch": 14.799154334038056,
"eval_loss": 1.2704553604125977,
"eval_runtime": 66.0979,
"eval_samples_per_second": 15.855,
"eval_steps_per_second": 1.589,
"eval_wer": 0.5621812475480581,
"step": 7000
},
{
"epoch": 15.221987315010571,
"grad_norm": 0.3582391142845154,
"learning_rate": 3.631788079470198e-05,
"loss": 0.0538,
"step": 7200
},
{
"epoch": 15.221987315010571,
"eval_loss": 1.3540122509002686,
"eval_runtime": 66.4126,
"eval_samples_per_second": 15.78,
"eval_steps_per_second": 1.581,
"eval_wer": 0.5837583366025892,
"step": 7200
},
{
"epoch": 15.644820295983086,
"grad_norm": 0.261311411857605,
"learning_rate": 3.5439293598234e-05,
"loss": 0.0585,
"step": 7400
},
{
"epoch": 15.644820295983086,
"eval_loss": 1.3333740234375,
"eval_runtime": 66.9559,
"eval_samples_per_second": 15.652,
"eval_steps_per_second": 1.568,
"eval_wer": 0.5798352295017654,
"step": 7400
},
{
"epoch": 16.067653276955603,
"grad_norm": 0.3650034964084625,
"learning_rate": 3.455629139072848e-05,
"loss": 0.0548,
"step": 7600
},
{
"epoch": 16.067653276955603,
"eval_loss": 1.3312934637069702,
"eval_runtime": 66.7147,
"eval_samples_per_second": 15.709,
"eval_steps_per_second": 1.574,
"eval_wer": 0.5723813260102001,
"step": 7600
},
{
"epoch": 16.49048625792812,
"grad_norm": 0.4263480305671692,
"learning_rate": 3.367328918322296e-05,
"loss": 0.0526,
"step": 7800
},
{
"epoch": 16.49048625792812,
"eval_loss": 1.3299189805984497,
"eval_runtime": 67.2527,
"eval_samples_per_second": 15.583,
"eval_steps_per_second": 1.561,
"eval_wer": 0.5719890153001177,
"step": 7800
},
{
"epoch": 16.913319238900634,
"grad_norm": 0.41519472002983093,
"learning_rate": 3.2790286975717444e-05,
"loss": 0.0577,
"step": 8000
},
{
"epoch": 16.913319238900634,
"eval_loss": 1.3206344842910767,
"eval_runtime": 67.3412,
"eval_samples_per_second": 15.563,
"eval_steps_per_second": 1.559,
"eval_wer": 0.5829737151824245,
"step": 8000
},
{
"epoch": 17.33615221987315,
"grad_norm": 0.9854322075843811,
"learning_rate": 3.190728476821192e-05,
"loss": 0.0513,
"step": 8200
},
{
"epoch": 17.33615221987315,
"eval_loss": 1.3500277996063232,
"eval_runtime": 66.908,
"eval_samples_per_second": 15.663,
"eval_steps_per_second": 1.569,
"eval_wer": 0.5786582973715182,
"step": 8200
},
{
"epoch": 17.758985200845665,
"grad_norm": 0.4509384334087372,
"learning_rate": 3.10242825607064e-05,
"loss": 0.0506,
"step": 8400
},
{
"epoch": 17.758985200845665,
"eval_loss": 1.3184752464294434,
"eval_runtime": 66.0674,
"eval_samples_per_second": 15.863,
"eval_steps_per_second": 1.589,
"eval_wer": 0.5698313063946646,
"step": 8400
},
{
"epoch": 18.181818181818183,
"grad_norm": 0.20881137251853943,
"learning_rate": 3.0145695364238412e-05,
"loss": 0.0498,
"step": 8600
},
{
"epoch": 18.181818181818183,
"eval_loss": 1.3655884265899658,
"eval_runtime": 65.7835,
"eval_samples_per_second": 15.931,
"eval_steps_per_second": 1.596,
"eval_wer": 0.5800313848568066,
"step": 8600
},
{
"epoch": 18.6046511627907,
"grad_norm": 8.674209594726562,
"learning_rate": 2.9262693156732894e-05,
"loss": 0.0515,
"step": 8800
},
{
"epoch": 18.6046511627907,
"eval_loss": 1.3253123760223389,
"eval_runtime": 65.9715,
"eval_samples_per_second": 15.886,
"eval_steps_per_second": 1.592,
"eval_wer": 0.5668889760690466,
"step": 8800
},
{
"epoch": 19.027484143763214,
"grad_norm": 0.2627590298652649,
"learning_rate": 2.83841059602649e-05,
"loss": 0.05,
"step": 9000
},
{
"epoch": 19.027484143763214,
"eval_loss": 1.3410530090332031,
"eval_runtime": 65.3408,
"eval_samples_per_second": 16.039,
"eval_steps_per_second": 1.607,
"eval_wer": 0.5810121616320125,
"step": 9000
},
{
"epoch": 19.45031712473573,
"grad_norm": 0.3467521667480469,
"learning_rate": 2.7501103752759385e-05,
"loss": 0.048,
"step": 9200
},
{
"epoch": 19.45031712473573,
"eval_loss": 1.3627517223358154,
"eval_runtime": 67.7618,
"eval_samples_per_second": 15.466,
"eval_steps_per_second": 1.55,
"eval_wer": 0.5729697920753236,
"step": 9200
},
{
"epoch": 19.873150105708245,
"grad_norm": 0.3944872319698334,
"learning_rate": 2.6618101545253866e-05,
"loss": 0.049,
"step": 9400
},
{
"epoch": 19.873150105708245,
"eval_loss": 1.3700072765350342,
"eval_runtime": 71.6559,
"eval_samples_per_second": 14.625,
"eval_steps_per_second": 1.465,
"eval_wer": 0.5729697920753236,
"step": 9400
},
{
"epoch": 20.29598308668076,
"grad_norm": 0.63824862241745,
"learning_rate": 2.5735099337748348e-05,
"loss": 0.0469,
"step": 9600
},
{
"epoch": 20.29598308668076,
"eval_loss": 1.364588975906372,
"eval_runtime": 75.7846,
"eval_samples_per_second": 13.829,
"eval_steps_per_second": 1.386,
"eval_wer": 0.5717928599450764,
"step": 9600
},
{
"epoch": 20.718816067653275,
"grad_norm": 0.29639732837677,
"learning_rate": 2.4852097130242826e-05,
"loss": 0.0474,
"step": 9800
},
{
"epoch": 20.718816067653275,
"eval_loss": 1.4190597534179688,
"eval_runtime": 72.4218,
"eval_samples_per_second": 14.471,
"eval_steps_per_second": 1.45,
"eval_wer": 0.5786582973715182,
"step": 9800
},
{
"epoch": 21.141649048625794,
"grad_norm": 0.3832600712776184,
"learning_rate": 2.396909492273731e-05,
"loss": 0.0488,
"step": 10000
},
{
"epoch": 21.141649048625794,
"eval_loss": 1.3449620008468628,
"eval_runtime": 76.5571,
"eval_samples_per_second": 13.689,
"eval_steps_per_second": 1.372,
"eval_wer": 0.575323656335818,
"step": 10000
},
{
"epoch": 21.56448202959831,
"grad_norm": 0.44069257378578186,
"learning_rate": 2.308609271523179e-05,
"loss": 0.0466,
"step": 10200
},
{
"epoch": 21.56448202959831,
"eval_loss": 1.2960785627365112,
"eval_runtime": 72.0544,
"eval_samples_per_second": 14.545,
"eval_steps_per_second": 1.457,
"eval_wer": 0.5612004707728521,
"step": 10200
},
{
"epoch": 21.987315010570825,
"grad_norm": 0.336251437664032,
"learning_rate": 2.220309050772627e-05,
"loss": 0.0462,
"step": 10400
},
{
"epoch": 21.987315010570825,
"eval_loss": 1.3378616571426392,
"eval_runtime": 66.2158,
"eval_samples_per_second": 15.827,
"eval_steps_per_second": 1.586,
"eval_wer": 0.5731659474303649,
"step": 10400
},
{
"epoch": 22.41014799154334,
"grad_norm": 0.24060240387916565,
"learning_rate": 2.1320088300220752e-05,
"loss": 0.0479,
"step": 10600
},
{
"epoch": 22.41014799154334,
"eval_loss": 1.364053726196289,
"eval_runtime": 65.0016,
"eval_samples_per_second": 16.123,
"eval_steps_per_second": 1.615,
"eval_wer": 0.5755198116908592,
"step": 10600
},
{
"epoch": 22.832980972515855,
"grad_norm": 0.28006625175476074,
"learning_rate": 2.0437086092715233e-05,
"loss": 0.0475,
"step": 10800
},
{
"epoch": 22.832980972515855,
"eval_loss": 1.3316149711608887,
"eval_runtime": 66.7406,
"eval_samples_per_second": 15.703,
"eval_steps_per_second": 1.573,
"eval_wer": 0.5751275009807768,
"step": 10800
},
{
"epoch": 23.25581395348837,
"grad_norm": 1.4817793369293213,
"learning_rate": 1.9554083885209715e-05,
"loss": 0.0461,
"step": 11000
},
{
"epoch": 23.25581395348837,
"eval_loss": 1.4020766019821167,
"eval_runtime": 65.2476,
"eval_samples_per_second": 16.062,
"eval_steps_per_second": 1.609,
"eval_wer": 0.5778736759513535,
"step": 11000
},
{
"epoch": 23.67864693446089,
"grad_norm": 0.13971921801567078,
"learning_rate": 1.8671081677704193e-05,
"loss": 0.0443,
"step": 11200
},
{
"epoch": 23.67864693446089,
"eval_loss": 1.3807997703552246,
"eval_runtime": 65.7311,
"eval_samples_per_second": 15.944,
"eval_steps_per_second": 1.597,
"eval_wer": 0.5766967438211064,
"step": 11200
},
{
"epoch": 24.101479915433405,
"grad_norm": 1.3055016994476318,
"learning_rate": 1.7788079470198678e-05,
"loss": 0.0448,
"step": 11400
},
{
"epoch": 24.101479915433405,
"eval_loss": 1.4156945943832397,
"eval_runtime": 65.6108,
"eval_samples_per_second": 15.973,
"eval_steps_per_second": 1.6,
"eval_wer": 0.5778736759513535,
"step": 11400
},
{
"epoch": 16.360902255639097,
"grad_norm": 0.8831915259361267,
"learning_rate": 2.8240928882438317e-05,
"loss": 0.1948,
"step": 11600
},
{
"epoch": 16.360902255639097,
"eval_loss": 0.8630273342132568,
"eval_runtime": 66.0147,
"eval_samples_per_second": 15.875,
"eval_steps_per_second": 1.591,
"eval_wer": 0.5619850921930168,
"step": 11600
},
{
"epoch": 16.642857142857142,
"grad_norm": 3.0019664764404297,
"learning_rate": 2.7663280116110306e-05,
"loss": 0.1658,
"step": 11800
},
{
"epoch": 16.642857142857142,
"eval_loss": 0.933027982711792,
"eval_runtime": 65.0734,
"eval_samples_per_second": 16.105,
"eval_steps_per_second": 1.614,
"eval_wer": 0.569242840329541,
"step": 11800
},
{
"epoch": 16.924812030075188,
"grad_norm": 1.1889337301254272,
"learning_rate": 2.7082728592162554e-05,
"loss": 0.1632,
"step": 12000
},
{
"epoch": 16.924812030075188,
"eval_loss": 0.8790073394775391,
"eval_runtime": 66.0534,
"eval_samples_per_second": 15.866,
"eval_steps_per_second": 1.59,
"eval_wer": 0.5517850137308749,
"step": 12000
},
{
"epoch": 17.206766917293233,
"grad_norm": 0.7651678919792175,
"learning_rate": 2.6502177068214808e-05,
"loss": 0.1373,
"step": 12200
},
{
"epoch": 17.206766917293233,
"eval_loss": 0.9278713464736938,
"eval_runtime": 66.2333,
"eval_samples_per_second": 15.823,
"eval_steps_per_second": 1.585,
"eval_wer": 0.5455080423695566,
"step": 12200
}
],
"logging_steps": 200,
"max_steps": 17725,
"num_input_tokens_seen": 0,
"num_train_epochs": 25,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.903488774188463e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}