| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.694760820045558, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02847380410022779, | |
| "grad_norm": 8.29788875579834, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 1.7669, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05694760820045558, | |
| "grad_norm": 5.953624725341797, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.347, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08542141230068337, | |
| "grad_norm": 6.313845157623291, | |
| "learning_rate": 1.5e-06, | |
| "loss": 1.0011, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11389521640091116, | |
| "grad_norm": 5.495311260223389, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.8525, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.14236902050113895, | |
| "grad_norm": 4.961582183837891, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.7632, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.17084282460136674, | |
| "grad_norm": 5.068117141723633, | |
| "learning_rate": 3e-06, | |
| "loss": 0.7467, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.19931662870159453, | |
| "grad_norm": 5.016373634338379, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.7178, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.22779043280182232, | |
| "grad_norm": 4.4371137619018555, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.6761, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.25626423690205014, | |
| "grad_norm": 5.090580940246582, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.6484, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.2847380410022779, | |
| "grad_norm": 4.779529094696045, | |
| "learning_rate": 5e-06, | |
| "loss": 0.677, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.3132118451025057, | |
| "grad_norm": 4.2773518562316895, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.6384, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3416856492027335, | |
| "grad_norm": 4.421960830688477, | |
| "learning_rate": 6e-06, | |
| "loss": 0.6291, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3701594533029613, | |
| "grad_norm": 4.473905563354492, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.6275, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.39863325740318906, | |
| "grad_norm": 5.127504825592041, | |
| "learning_rate": 7e-06, | |
| "loss": 0.5969, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.4271070615034169, | |
| "grad_norm": 4.176600933074951, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5706, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.45558086560364464, | |
| "grad_norm": 4.631809234619141, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.6083, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.48405466970387245, | |
| "grad_norm": 4.151340484619141, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.5777, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.5125284738041003, | |
| "grad_norm": 4.039161205291748, | |
| "learning_rate": 9e-06, | |
| "loss": 0.5529, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.541002277904328, | |
| "grad_norm": 4.89629602432251, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.5581, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5694760820045558, | |
| "grad_norm": 11.599489212036133, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5621, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5979498861047836, | |
| "grad_norm": 4.748031139373779, | |
| "learning_rate": 9.944444444444445e-06, | |
| "loss": 0.5785, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.6264236902050114, | |
| "grad_norm": 4.0705485343933105, | |
| "learning_rate": 9.88888888888889e-06, | |
| "loss": 0.5625, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6548974943052391, | |
| "grad_norm": 3.605722427368164, | |
| "learning_rate": 9.833333333333333e-06, | |
| "loss": 0.5613, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.683371298405467, | |
| "grad_norm": 3.7756550312042236, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 0.5521, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7118451025056948, | |
| "grad_norm": 3.743063449859619, | |
| "learning_rate": 9.722222222222223e-06, | |
| "loss": 0.5521, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.7403189066059226, | |
| "grad_norm": 3.786343812942505, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 0.5374, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7687927107061503, | |
| "grad_norm": 3.285397529602051, | |
| "learning_rate": 9.611111111111112e-06, | |
| "loss": 0.5155, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7972665148063781, | |
| "grad_norm": 4.082733154296875, | |
| "learning_rate": 9.555555555555556e-06, | |
| "loss": 0.5266, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8257403189066059, | |
| "grad_norm": 3.823092222213745, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.5022, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.8542141230068337, | |
| "grad_norm": 3.905947685241699, | |
| "learning_rate": 9.444444444444445e-06, | |
| "loss": 0.4924, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8826879271070615, | |
| "grad_norm": 3.707880735397339, | |
| "learning_rate": 9.38888888888889e-06, | |
| "loss": 0.496, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.9111617312072893, | |
| "grad_norm": 3.18892502784729, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.5012, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9396355353075171, | |
| "grad_norm": 3.8103792667388916, | |
| "learning_rate": 9.277777777777778e-06, | |
| "loss": 0.4767, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.9681093394077449, | |
| "grad_norm": 3.4749033451080322, | |
| "learning_rate": 9.222222222222224e-06, | |
| "loss": 0.5078, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9965831435079726, | |
| "grad_norm": 3.836411476135254, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.481, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.0250569476082005, | |
| "grad_norm": 2.716602325439453, | |
| "learning_rate": 9.111111111111112e-06, | |
| "loss": 0.3908, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.0535307517084282, | |
| "grad_norm": 3.388803482055664, | |
| "learning_rate": 9.055555555555556e-06, | |
| "loss": 0.3558, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.082004555808656, | |
| "grad_norm": 3.0076682567596436, | |
| "learning_rate": 9e-06, | |
| "loss": 0.3697, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.1104783599088839, | |
| "grad_norm": 3.24129056930542, | |
| "learning_rate": 8.944444444444446e-06, | |
| "loss": 0.3739, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "grad_norm": 2.9381356239318848, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.3674, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1389521640091116, | |
| "eval_loss": 0.4708998501300812, | |
| "eval_runtime": 2991.8311, | |
| "eval_samples_per_second": 2.347, | |
| "eval_steps_per_second": 0.147, | |
| "eval_wer": 0.34162802747479043, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1674259681093395, | |
| "grad_norm": 3.604189872741699, | |
| "learning_rate": 8.833333333333334e-06, | |
| "loss": 0.3498, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.1958997722095672, | |
| "grad_norm": 3.277430772781372, | |
| "learning_rate": 8.777777777777778e-06, | |
| "loss": 0.3628, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.224373576309795, | |
| "grad_norm": 3.1814355850219727, | |
| "learning_rate": 8.722222222222224e-06, | |
| "loss": 0.3506, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.2528473804100229, | |
| "grad_norm": 3.2664480209350586, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.3559, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.2813211845102506, | |
| "grad_norm": 2.933966875076294, | |
| "learning_rate": 8.611111111111112e-06, | |
| "loss": 0.3697, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.3097949886104785, | |
| "grad_norm": 2.8139538764953613, | |
| "learning_rate": 8.555555555555556e-06, | |
| "loss": 0.3674, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.3382687927107062, | |
| "grad_norm": 3.1578900814056396, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.3411, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.366742596810934, | |
| "grad_norm": 3.069774866104126, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 0.3655, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.3952164009111616, | |
| "grad_norm": 2.6920642852783203, | |
| "learning_rate": 8.38888888888889e-06, | |
| "loss": 0.3492, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.4236902050113895, | |
| "grad_norm": 2.7643768787384033, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.3534, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.4521640091116172, | |
| "grad_norm": 3.1626651287078857, | |
| "learning_rate": 8.277777777777778e-06, | |
| "loss": 0.3317, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.4806378132118452, | |
| "grad_norm": 3.417654514312744, | |
| "learning_rate": 8.222222222222222e-06, | |
| "loss": 0.3828, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.5091116173120729, | |
| "grad_norm": 3.388624668121338, | |
| "learning_rate": 8.166666666666668e-06, | |
| "loss": 0.3611, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.5375854214123006, | |
| "grad_norm": 2.685516834259033, | |
| "learning_rate": 8.111111111111112e-06, | |
| "loss": 0.3562, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.5660592255125285, | |
| "grad_norm": 3.0524353981018066, | |
| "learning_rate": 8.055555555555557e-06, | |
| "loss": 0.3633, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.5945330296127562, | |
| "grad_norm": 3.1776881217956543, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.353, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.6230068337129842, | |
| "grad_norm": 3.046795129776001, | |
| "learning_rate": 7.944444444444445e-06, | |
| "loss": 0.3556, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.6514806378132119, | |
| "grad_norm": 3.1991679668426514, | |
| "learning_rate": 7.88888888888889e-06, | |
| "loss": 0.3407, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.6799544419134396, | |
| "grad_norm": 2.918931007385254, | |
| "learning_rate": 7.833333333333333e-06, | |
| "loss": 0.3608, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.7084282460136673, | |
| "grad_norm": 3.0339889526367188, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 0.3539, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.7369020501138952, | |
| "grad_norm": 3.138218402862549, | |
| "learning_rate": 7.722222222222223e-06, | |
| "loss": 0.345, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.7653758542141231, | |
| "grad_norm": 3.70744252204895, | |
| "learning_rate": 7.666666666666667e-06, | |
| "loss": 0.347, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.7938496583143508, | |
| "grad_norm": 2.8581366539001465, | |
| "learning_rate": 7.611111111111111e-06, | |
| "loss": 0.3541, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.8223234624145785, | |
| "grad_norm": 2.9355995655059814, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 0.356, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.8507972665148062, | |
| "grad_norm": 3.1599183082580566, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.3677, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.8792710706150342, | |
| "grad_norm": 3.0059103965759277, | |
| "learning_rate": 7.444444444444445e-06, | |
| "loss": 0.3443, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.907744874715262, | |
| "grad_norm": 2.8699076175689697, | |
| "learning_rate": 7.38888888888889e-06, | |
| "loss": 0.3593, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.9362186788154898, | |
| "grad_norm": 3.279315233230591, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.3386, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.9646924829157175, | |
| "grad_norm": 3.5267879962921143, | |
| "learning_rate": 7.277777777777778e-06, | |
| "loss": 0.3486, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.9931662870159452, | |
| "grad_norm": 2.794961929321289, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 0.3372, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.021640091116173, | |
| "grad_norm": 2.9796462059020996, | |
| "learning_rate": 7.166666666666667e-06, | |
| "loss": 0.252, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 2.050113895216401, | |
| "grad_norm": 2.33910870552063, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 0.2207, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.078587699316629, | |
| "grad_norm": 2.6431703567504883, | |
| "learning_rate": 7.055555555555557e-06, | |
| "loss": 0.2158, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 2.1070615034168565, | |
| "grad_norm": 2.63135027885437, | |
| "learning_rate": 7e-06, | |
| "loss": 0.2172, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.135535307517084, | |
| "grad_norm": 2.7133092880249023, | |
| "learning_rate": 6.944444444444445e-06, | |
| "loss": 0.2095, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.164009111617312, | |
| "grad_norm": 2.2241263389587402, | |
| "learning_rate": 6.88888888888889e-06, | |
| "loss": 0.2118, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.19248291571754, | |
| "grad_norm": 2.509427070617676, | |
| "learning_rate": 6.833333333333334e-06, | |
| "loss": 0.217, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.2209567198177678, | |
| "grad_norm": 3.031168222427368, | |
| "learning_rate": 6.777777777777779e-06, | |
| "loss": 0.2239, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.2494305239179955, | |
| "grad_norm": 2.869800567626953, | |
| "learning_rate": 6.7222222222222235e-06, | |
| "loss": 0.2247, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "grad_norm": 2.5018348693847656, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.2228, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.277904328018223, | |
| "eval_loss": 0.4358457922935486, | |
| "eval_runtime": 3078.2421, | |
| "eval_samples_per_second": 2.281, | |
| "eval_steps_per_second": 0.143, | |
| "eval_wer": 0.3116688145185933, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.306378132118451, | |
| "grad_norm": 3.0430192947387695, | |
| "learning_rate": 6.6111111111111115e-06, | |
| "loss": 0.2138, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 2.334851936218679, | |
| "grad_norm": 2.651179075241089, | |
| "learning_rate": 6.555555555555556e-06, | |
| "loss": 0.233, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.3633257403189067, | |
| "grad_norm": 2.935361623764038, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.207, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.3917995444191344, | |
| "grad_norm": 2.7219200134277344, | |
| "learning_rate": 6.444444444444445e-06, | |
| "loss": 0.2291, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.420273348519362, | |
| "grad_norm": 2.854238510131836, | |
| "learning_rate": 6.3888888888888885e-06, | |
| "loss": 0.2372, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.44874715261959, | |
| "grad_norm": 2.149994373321533, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 0.2304, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.477220956719818, | |
| "grad_norm": 2.346008062362671, | |
| "learning_rate": 6.277777777777778e-06, | |
| "loss": 0.2181, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.5056947608200457, | |
| "grad_norm": 2.859161138534546, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 0.2235, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.5341685649202734, | |
| "grad_norm": 2.6515421867370605, | |
| "learning_rate": 6.166666666666667e-06, | |
| "loss": 0.2218, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.562642369020501, | |
| "grad_norm": 2.8457796573638916, | |
| "learning_rate": 6.111111111111112e-06, | |
| "loss": 0.2197, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.591116173120729, | |
| "grad_norm": 2.9670958518981934, | |
| "learning_rate": 6.055555555555555e-06, | |
| "loss": 0.2179, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.619589977220957, | |
| "grad_norm": 3.2194952964782715, | |
| "learning_rate": 6e-06, | |
| "loss": 0.2201, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.6480637813211843, | |
| "grad_norm": 2.424835205078125, | |
| "learning_rate": 5.944444444444445e-06, | |
| "loss": 0.2295, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.6765375854214124, | |
| "grad_norm": 2.3368935585021973, | |
| "learning_rate": 5.88888888888889e-06, | |
| "loss": 0.211, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.70501138952164, | |
| "grad_norm": 2.0860483646392822, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.2171, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.733485193621868, | |
| "grad_norm": 2.6217784881591797, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 0.2047, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.7619589977220955, | |
| "grad_norm": 3.0176894664764404, | |
| "learning_rate": 5.722222222222222e-06, | |
| "loss": 0.2248, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.7904328018223232, | |
| "grad_norm": 2.360590696334839, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 0.2101, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.8189066059225514, | |
| "grad_norm": 2.451885223388672, | |
| "learning_rate": 5.611111111111112e-06, | |
| "loss": 0.2275, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.847380410022779, | |
| "grad_norm": 2.3589582443237305, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 0.217, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.875854214123007, | |
| "grad_norm": 2.6570801734924316, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.2304, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.9043280182232345, | |
| "grad_norm": 2.6280150413513184, | |
| "learning_rate": 5.444444444444445e-06, | |
| "loss": 0.2105, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.932801822323462, | |
| "grad_norm": 2.4569592475891113, | |
| "learning_rate": 5.388888888888889e-06, | |
| "loss": 0.2027, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.9612756264236904, | |
| "grad_norm": 3.0375263690948486, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.2164, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.989749430523918, | |
| "grad_norm": 2.2923061847686768, | |
| "learning_rate": 5.2777777777777785e-06, | |
| "loss": 0.2194, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 3.0182232346241458, | |
| "grad_norm": 2.045480966567993, | |
| "learning_rate": 5.2222222222222226e-06, | |
| "loss": 0.1584, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 3.0466970387243735, | |
| "grad_norm": 2.0926032066345215, | |
| "learning_rate": 5.1666666666666675e-06, | |
| "loss": 0.1257, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 3.075170842824601, | |
| "grad_norm": 2.2969894409179688, | |
| "learning_rate": 5.1111111111111115e-06, | |
| "loss": 0.1211, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 3.1036446469248293, | |
| "grad_norm": 2.1000475883483887, | |
| "learning_rate": 5.0555555555555555e-06, | |
| "loss": 0.1225, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 3.132118451025057, | |
| "grad_norm": 2.593099594116211, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1289, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 3.1605922551252847, | |
| "grad_norm": 1.8630212545394897, | |
| "learning_rate": 4.944444444444445e-06, | |
| "loss": 0.129, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 3.1890660592255125, | |
| "grad_norm": 2.0034408569335938, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 0.1222, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.21753986332574, | |
| "grad_norm": 2.4184627532958984, | |
| "learning_rate": 4.833333333333333e-06, | |
| "loss": 0.1382, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 3.2460136674259683, | |
| "grad_norm": 2.2710211277008057, | |
| "learning_rate": 4.777777777777778e-06, | |
| "loss": 0.1185, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 3.274487471526196, | |
| "grad_norm": 2.108302116394043, | |
| "learning_rate": 4.722222222222222e-06, | |
| "loss": 0.1256, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 3.3029612756264237, | |
| "grad_norm": 2.66013503074646, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.126, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 3.3314350797266514, | |
| "grad_norm": 2.502146005630493, | |
| "learning_rate": 4.611111111111112e-06, | |
| "loss": 0.1188, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 3.359908883826879, | |
| "grad_norm": 2.9654130935668945, | |
| "learning_rate": 4.555555555555556e-06, | |
| "loss": 0.1265, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 3.3883826879271073, | |
| "grad_norm": 2.2393388748168945, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.1197, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 3.416856492027335, | |
| "grad_norm": 2.5502378940582275, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.1266, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.416856492027335, | |
| "eval_loss": 0.45878028869628906, | |
| "eval_runtime": 3136.0137, | |
| "eval_samples_per_second": 2.239, | |
| "eval_steps_per_second": 0.14, | |
| "eval_wer": 0.3070321114934435, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.4453302961275627, | |
| "grad_norm": 2.695366382598877, | |
| "learning_rate": 4.388888888888889e-06, | |
| "loss": 0.1261, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 3.4738041002277904, | |
| "grad_norm": 2.2533366680145264, | |
| "learning_rate": 4.333333333333334e-06, | |
| "loss": 0.1153, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 3.502277904328018, | |
| "grad_norm": 2.5580222606658936, | |
| "learning_rate": 4.277777777777778e-06, | |
| "loss": 0.1275, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 3.5307517084282463, | |
| "grad_norm": 2.1148972511291504, | |
| "learning_rate": 4.222222222222223e-06, | |
| "loss": 0.1254, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 3.559225512528474, | |
| "grad_norm": 2.1242175102233887, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.1265, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 3.5876993166287017, | |
| "grad_norm": 2.1025946140289307, | |
| "learning_rate": 4.111111111111111e-06, | |
| "loss": 0.1193, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 3.6161731207289294, | |
| "grad_norm": 1.925126314163208, | |
| "learning_rate": 4.055555555555556e-06, | |
| "loss": 0.1243, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 3.644646924829157, | |
| "grad_norm": 2.383075714111328, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.1192, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.6731207289293852, | |
| "grad_norm": 1.961838960647583, | |
| "learning_rate": 3.944444444444445e-06, | |
| "loss": 0.1192, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 3.7015945330296125, | |
| "grad_norm": 2.1979856491088867, | |
| "learning_rate": 3.88888888888889e-06, | |
| "loss": 0.1276, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 3.7300683371298406, | |
| "grad_norm": 1.8826966285705566, | |
| "learning_rate": 3.833333333333334e-06, | |
| "loss": 0.1184, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 3.7585421412300684, | |
| "grad_norm": 2.3589093685150146, | |
| "learning_rate": 3.777777777777778e-06, | |
| "loss": 0.1193, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.787015945330296, | |
| "grad_norm": 2.299064874649048, | |
| "learning_rate": 3.7222222222222225e-06, | |
| "loss": 0.1218, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 3.8154897494305238, | |
| "grad_norm": 2.088062047958374, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 0.1241, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 3.8439635535307515, | |
| "grad_norm": 2.695964813232422, | |
| "learning_rate": 3.6111111111111115e-06, | |
| "loss": 0.1297, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 3.8724373576309796, | |
| "grad_norm": 2.341107130050659, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 0.1287, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.9009111617312073, | |
| "grad_norm": 2.7177109718322754, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.1222, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 3.929384965831435, | |
| "grad_norm": 2.182626962661743, | |
| "learning_rate": 3.444444444444445e-06, | |
| "loss": 0.1257, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 3.9578587699316627, | |
| "grad_norm": 2.2234408855438232, | |
| "learning_rate": 3.3888888888888893e-06, | |
| "loss": 0.121, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 3.9863325740318905, | |
| "grad_norm": 2.3420960903167725, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.1232, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.014806378132119, | |
| "grad_norm": 1.4395020008087158, | |
| "learning_rate": 3.277777777777778e-06, | |
| "loss": 0.0916, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 4.043280182232346, | |
| "grad_norm": 1.6081222295761108, | |
| "learning_rate": 3.2222222222222227e-06, | |
| "loss": 0.0605, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 4.071753986332574, | |
| "grad_norm": 1.7407175302505493, | |
| "learning_rate": 3.1666666666666667e-06, | |
| "loss": 0.0585, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 4.100227790432802, | |
| "grad_norm": 1.4883263111114502, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 0.0605, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 4.128701594533029, | |
| "grad_norm": 1.864080786705017, | |
| "learning_rate": 3.055555555555556e-06, | |
| "loss": 0.0583, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 4.157175398633258, | |
| "grad_norm": 2.384763479232788, | |
| "learning_rate": 3e-06, | |
| "loss": 0.0608, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 4.185649202733485, | |
| "grad_norm": 1.896714448928833, | |
| "learning_rate": 2.944444444444445e-06, | |
| "loss": 0.0637, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 4.214123006833713, | |
| "grad_norm": 2.230041265487671, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 0.0651, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 4.242596810933941, | |
| "grad_norm": 1.9410207271575928, | |
| "learning_rate": 2.8333333333333335e-06, | |
| "loss": 0.0594, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 4.271070615034168, | |
| "grad_norm": 2.4352822303771973, | |
| "learning_rate": 2.7777777777777783e-06, | |
| "loss": 0.0581, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 4.2995444191343966, | |
| "grad_norm": 1.998757243156433, | |
| "learning_rate": 2.7222222222222224e-06, | |
| "loss": 0.0627, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 4.328018223234624, | |
| "grad_norm": 1.773343563079834, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.0639, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 4.356492027334852, | |
| "grad_norm": 2.35654616355896, | |
| "learning_rate": 2.6111111111111113e-06, | |
| "loss": 0.0578, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 4.38496583143508, | |
| "grad_norm": 2.2221155166625977, | |
| "learning_rate": 2.5555555555555557e-06, | |
| "loss": 0.0585, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 4.413439635535307, | |
| "grad_norm": 1.8902034759521484, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0598, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 4.4419134396355355, | |
| "grad_norm": 1.73625910282135, | |
| "learning_rate": 2.446666666666667e-06, | |
| "loss": 0.0616, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 4.470387243735763, | |
| "grad_norm": 1.6135858297348022, | |
| "learning_rate": 2.3911111111111113e-06, | |
| "loss": 0.0592, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 4.498861047835991, | |
| "grad_norm": 2.070269823074341, | |
| "learning_rate": 2.3355555555555557e-06, | |
| "loss": 0.0573, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 4.527334851936219, | |
| "grad_norm": 1.600358247756958, | |
| "learning_rate": 2.28e-06, | |
| "loss": 0.0607, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 4.555808656036446, | |
| "grad_norm": 2.013101577758789, | |
| "learning_rate": 2.2244444444444447e-06, | |
| "loss": 0.0611, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.555808656036446, | |
| "eval_loss": 0.5140041708946228, | |
| "eval_runtime": 2974.0961, | |
| "eval_samples_per_second": 2.361, | |
| "eval_steps_per_second": 0.148, | |
| "eval_wer": 0.30393654758267014, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.5842824601366745, | |
| "grad_norm": 2.0340895652770996, | |
| "learning_rate": 2.168888888888889e-06, | |
| "loss": 0.0579, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 4.612756264236902, | |
| "grad_norm": 1.8350932598114014, | |
| "learning_rate": 2.1133333333333336e-06, | |
| "loss": 0.0567, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 4.64123006833713, | |
| "grad_norm": 2.042941093444824, | |
| "learning_rate": 2.057777777777778e-06, | |
| "loss": 0.0688, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 4.669703872437358, | |
| "grad_norm": 2.5973660945892334, | |
| "learning_rate": 2.0022222222222225e-06, | |
| "loss": 0.057, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 4.698177676537585, | |
| "grad_norm": 1.658544659614563, | |
| "learning_rate": 1.9466666666666665e-06, | |
| "loss": 0.0573, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 4.7266514806378135, | |
| "grad_norm": 1.918231725692749, | |
| "learning_rate": 1.8911111111111114e-06, | |
| "loss": 0.0591, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 4.755125284738041, | |
| "grad_norm": 1.6860567331314087, | |
| "learning_rate": 1.8355555555555557e-06, | |
| "loss": 0.0598, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 4.783599088838269, | |
| "grad_norm": 1.7576823234558105, | |
| "learning_rate": 1.7800000000000001e-06, | |
| "loss": 0.0606, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 4.812072892938497, | |
| "grad_norm": 2.130852699279785, | |
| "learning_rate": 1.7244444444444448e-06, | |
| "loss": 0.0584, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 4.840546697038724, | |
| "grad_norm": 2.2588632106781006, | |
| "learning_rate": 1.668888888888889e-06, | |
| "loss": 0.0594, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 4.8690205011389525, | |
| "grad_norm": 1.6730296611785889, | |
| "learning_rate": 1.6133333333333335e-06, | |
| "loss": 0.0542, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 4.89749430523918, | |
| "grad_norm": 2.031381607055664, | |
| "learning_rate": 1.5577777777777777e-06, | |
| "loss": 0.0534, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 4.925968109339408, | |
| "grad_norm": 1.2967132329940796, | |
| "learning_rate": 1.5022222222222224e-06, | |
| "loss": 0.053, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 4.954441913439636, | |
| "grad_norm": 2.0789716243743896, | |
| "learning_rate": 1.4466666666666669e-06, | |
| "loss": 0.0546, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 4.982915717539863, | |
| "grad_norm": 1.8398990631103516, | |
| "learning_rate": 1.3911111111111111e-06, | |
| "loss": 0.0548, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 5.011389521640091, | |
| "grad_norm": 1.2099305391311646, | |
| "learning_rate": 1.3355555555555558e-06, | |
| "loss": 0.0464, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 5.039863325740319, | |
| "grad_norm": 0.9306023120880127, | |
| "learning_rate": 1.28e-06, | |
| "loss": 0.0286, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 5.068337129840547, | |
| "grad_norm": 1.1420648097991943, | |
| "learning_rate": 1.2244444444444445e-06, | |
| "loss": 0.0252, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 5.096810933940774, | |
| "grad_norm": 1.3366564512252808, | |
| "learning_rate": 1.168888888888889e-06, | |
| "loss": 0.0251, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 5.125284738041002, | |
| "grad_norm": 1.437525749206543, | |
| "learning_rate": 1.1133333333333334e-06, | |
| "loss": 0.0252, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 5.15375854214123, | |
| "grad_norm": 0.8598686456680298, | |
| "learning_rate": 1.0577777777777779e-06, | |
| "loss": 0.0249, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 5.182232346241458, | |
| "grad_norm": 1.447130560874939, | |
| "learning_rate": 1.0022222222222223e-06, | |
| "loss": 0.0243, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 5.210706150341686, | |
| "grad_norm": 1.1469416618347168, | |
| "learning_rate": 9.466666666666667e-07, | |
| "loss": 0.0276, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 5.239179954441913, | |
| "grad_norm": 1.7906535863876343, | |
| "learning_rate": 8.911111111111112e-07, | |
| "loss": 0.026, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 5.267653758542141, | |
| "grad_norm": 1.0731186866760254, | |
| "learning_rate": 8.355555555555556e-07, | |
| "loss": 0.0253, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 5.296127562642369, | |
| "grad_norm": 2.2479116916656494, | |
| "learning_rate": 7.8e-07, | |
| "loss": 0.0261, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 5.324601366742597, | |
| "grad_norm": 1.143470287322998, | |
| "learning_rate": 7.244444444444446e-07, | |
| "loss": 0.0249, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 5.353075170842825, | |
| "grad_norm": 1.1293048858642578, | |
| "learning_rate": 6.68888888888889e-07, | |
| "loss": 0.0236, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 5.381548974943052, | |
| "grad_norm": 1.7358245849609375, | |
| "learning_rate": 6.133333333333333e-07, | |
| "loss": 0.023, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 5.41002277904328, | |
| "grad_norm": 1.031139612197876, | |
| "learning_rate": 5.577777777777779e-07, | |
| "loss": 0.0234, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 5.438496583143508, | |
| "grad_norm": 1.8313310146331787, | |
| "learning_rate": 5.022222222222222e-07, | |
| "loss": 0.0261, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 5.466970387243736, | |
| "grad_norm": 1.6976441144943237, | |
| "learning_rate": 4.466666666666667e-07, | |
| "loss": 0.0237, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 5.495444191343964, | |
| "grad_norm": 1.3694329261779785, | |
| "learning_rate": 3.9111111111111115e-07, | |
| "loss": 0.0257, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 5.523917995444191, | |
| "grad_norm": 1.0080304145812988, | |
| "learning_rate": 3.3555555555555556e-07, | |
| "loss": 0.0246, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 5.552391799544419, | |
| "grad_norm": 0.9774746298789978, | |
| "learning_rate": 2.8e-07, | |
| "loss": 0.0254, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 5.5808656036446465, | |
| "grad_norm": 1.1044853925704956, | |
| "learning_rate": 2.2444444444444445e-07, | |
| "loss": 0.026, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 5.609339407744875, | |
| "grad_norm": 1.2919448614120483, | |
| "learning_rate": 1.6888888888888888e-07, | |
| "loss": 0.0236, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 5.637813211845103, | |
| "grad_norm": 1.2227028608322144, | |
| "learning_rate": 1.1333333333333336e-07, | |
| "loss": 0.0273, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 5.66628701594533, | |
| "grad_norm": 0.887917697429657, | |
| "learning_rate": 5.777777777777778e-08, | |
| "loss": 0.0249, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 5.694760820045558, | |
| "grad_norm": 0.8368929624557495, | |
| "learning_rate": 2.2222222222222225e-09, | |
| "loss": 0.0245, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.694760820045558, | |
| "eval_loss": 0.5775743126869202, | |
| "eval_runtime": 2967.3924, | |
| "eval_samples_per_second": 2.366, | |
| "eval_steps_per_second": 0.148, | |
| "eval_wer": 0.3064475414845421, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.694760820045558, | |
| "step": 5000, | |
| "total_flos": 5.43429854134272e+20, | |
| "train_loss": 0.25138797866106033, | |
| "train_runtime": 67213.2375, | |
| "train_samples_per_second": 2.38, | |
| "train_steps_per_second": 0.074 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "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": 5.43429854134272e+20, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |