| { | |
| "best_metric": 30.100143061516455, | |
| "best_model_checkpoint": "./Malasar_50_latest/checkpoint-1000", | |
| "epoch": 2.846299810246679, | |
| "global_step": 1500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.4065, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 8.8e-06, | |
| "loss": 0.6768, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "eval_loss": 0.15913546085357666, | |
| "eval_runtime": 1024.1019, | |
| "eval_samples_per_second": 1.829, | |
| "eval_steps_per_second": 0.115, | |
| "eval_wer": 63.91988555078684, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.86896551724138e-06, | |
| "loss": 0.1782, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 9.696551724137932e-06, | |
| "loss": 0.1711, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 0.11521261185407639, | |
| "eval_runtime": 1068.1467, | |
| "eval_samples_per_second": 1.754, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 52.989985693848354, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 9.524137931034484e-06, | |
| "loss": 0.151, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 9.351724137931034e-06, | |
| "loss": 0.1392, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "eval_loss": 0.1063603013753891, | |
| "eval_runtime": 1067.453, | |
| "eval_samples_per_second": 1.755, | |
| "eval_steps_per_second": 0.111, | |
| "eval_wer": 42.00286123032904, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 9.179310344827587e-06, | |
| "loss": 0.1149, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 9.006896551724139e-06, | |
| "loss": 0.1131, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_loss": 0.09218524396419525, | |
| "eval_runtime": 1059.1591, | |
| "eval_samples_per_second": 1.768, | |
| "eval_steps_per_second": 0.111, | |
| "eval_wer": 40.2002861230329, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 8.83448275862069e-06, | |
| "loss": 0.1052, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 8.662068965517241e-06, | |
| "loss": 0.1208, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "eval_loss": 0.08814000338315964, | |
| "eval_runtime": 1037.3449, | |
| "eval_samples_per_second": 1.806, | |
| "eval_steps_per_second": 0.114, | |
| "eval_wer": 46.49499284692418, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 8.489655172413795e-06, | |
| "loss": 0.1107, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 8.317241379310345e-06, | |
| "loss": 0.1312, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_loss": 0.07998558133840561, | |
| "eval_runtime": 1053.0015, | |
| "eval_samples_per_second": 1.779, | |
| "eval_steps_per_second": 0.112, | |
| "eval_wer": 37.51072961373391, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 8.144827586206897e-06, | |
| "loss": 0.0947, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 7.972413793103448e-06, | |
| "loss": 0.1112, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "eval_loss": 0.07743828743696213, | |
| "eval_runtime": 1073.0726, | |
| "eval_samples_per_second": 1.745, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 37.56795422031474, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 0.118, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 7.627586206896552e-06, | |
| "loss": 0.1, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 0.0833013579249382, | |
| "eval_runtime": 1078.2864, | |
| "eval_samples_per_second": 1.737, | |
| "eval_steps_per_second": 0.109, | |
| "eval_wer": 33.59084406294707, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 7.455172413793104e-06, | |
| "loss": 0.1023, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 7.282758620689656e-06, | |
| "loss": 0.094, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "eval_loss": 0.07670725882053375, | |
| "eval_runtime": 1078.233, | |
| "eval_samples_per_second": 1.737, | |
| "eval_steps_per_second": 0.109, | |
| "eval_wer": 31.702432045779684, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 7.110344827586207e-06, | |
| "loss": 0.0787, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 6.937931034482759e-06, | |
| "loss": 0.0758, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "eval_loss": 0.07279360294342041, | |
| "eval_runtime": 1081.1361, | |
| "eval_samples_per_second": 1.732, | |
| "eval_steps_per_second": 0.109, | |
| "eval_wer": 32.33190271816881, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 6.7655172413793116e-06, | |
| "loss": 0.1144, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 6.593103448275863e-06, | |
| "loss": 0.0967, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "eval_loss": 0.07088593393564224, | |
| "eval_runtime": 1067.9119, | |
| "eval_samples_per_second": 1.754, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 34.39198855507868, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 6.420689655172414e-06, | |
| "loss": 0.0497, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 6.248275862068966e-06, | |
| "loss": 0.0423, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "eval_loss": 0.07051743566989899, | |
| "eval_runtime": 1075.828, | |
| "eval_samples_per_second": 1.741, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 31.04434907010014, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 6.075862068965518e-06, | |
| "loss": 0.0686, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 5.9034482758620695e-06, | |
| "loss": 0.0669, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_loss": 0.07203543931245804, | |
| "eval_runtime": 1076.3331, | |
| "eval_samples_per_second": 1.74, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 31.587982832618028, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 5.731034482758621e-06, | |
| "loss": 0.0615, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 5.558620689655173e-06, | |
| "loss": 0.0686, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "eval_loss": 0.06905751675367355, | |
| "eval_runtime": 1084.2893, | |
| "eval_samples_per_second": 1.727, | |
| "eval_steps_per_second": 0.109, | |
| "eval_wer": 31.21602288984263, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 5.386206896551725e-06, | |
| "loss": 0.0704, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 5.213793103448276e-06, | |
| "loss": 0.0798, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "eval_loss": 0.06694240123033524, | |
| "eval_runtime": 1087.4811, | |
| "eval_samples_per_second": 1.722, | |
| "eval_steps_per_second": 0.109, | |
| "eval_wer": 30.643776824034337, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 5.041379310344828e-06, | |
| "loss": 0.0756, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 4.8689655172413795e-06, | |
| "loss": 0.0818, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "eval_loss": 0.06710417568683624, | |
| "eval_runtime": 1073.6009, | |
| "eval_samples_per_second": 1.745, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 32.18884120171674, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 4.6965517241379315e-06, | |
| "loss": 0.0571, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 4.524137931034483e-06, | |
| "loss": 0.0841, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "eval_loss": 0.06191212683916092, | |
| "eval_runtime": 1076.9574, | |
| "eval_samples_per_second": 1.739, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 28.898426323319025, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 4.351724137931035e-06, | |
| "loss": 0.0927, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 4.179310344827587e-06, | |
| "loss": 0.0566, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_loss": 0.06160915642976761, | |
| "eval_runtime": 1078.2038, | |
| "eval_samples_per_second": 1.737, | |
| "eval_steps_per_second": 0.109, | |
| "eval_wer": 29.127324749642348, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 4.006896551724138e-06, | |
| "loss": 0.0466, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 3.83448275862069e-06, | |
| "loss": 0.0542, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "eval_loss": 0.06113772094249725, | |
| "eval_runtime": 1072.2754, | |
| "eval_samples_per_second": 1.747, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 28.841201716738198, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 3.6620689655172415e-06, | |
| "loss": 0.0536, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 3.489655172413793e-06, | |
| "loss": 0.0638, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "eval_loss": 0.06121109798550606, | |
| "eval_runtime": 1073.2525, | |
| "eval_samples_per_second": 1.745, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 30.100143061516455, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 3.3172413793103453e-06, | |
| "loss": 0.0559, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 3.1448275862068965e-06, | |
| "loss": 0.0503, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "eval_loss": 0.06099317967891693, | |
| "eval_runtime": 1074.3991, | |
| "eval_samples_per_second": 1.743, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 29.32761087267525, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 2.9724137931034486e-06, | |
| "loss": 0.0627, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.0369, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "eval_loss": 0.06123083084821701, | |
| "eval_runtime": 1071.2265, | |
| "eval_samples_per_second": 1.748, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 29.32761087267525, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 2.627586206896552e-06, | |
| "loss": 0.0395, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 2.4551724137931036e-06, | |
| "loss": 0.0273, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "eval_loss": 0.06379574537277222, | |
| "eval_runtime": 1064.3917, | |
| "eval_samples_per_second": 1.76, | |
| "eval_steps_per_second": 0.111, | |
| "eval_wer": 32.58941344778255, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 2.2827586206896553e-06, | |
| "loss": 0.0345, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 2.110344827586207e-06, | |
| "loss": 0.0269, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "eval_loss": 0.06369271874427795, | |
| "eval_runtime": 1147.8053, | |
| "eval_samples_per_second": 1.632, | |
| "eval_steps_per_second": 0.103, | |
| "eval_wer": 31.673819742489272, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 1.9379310344827586e-06, | |
| "loss": 0.0446, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 1.7655172413793103e-06, | |
| "loss": 0.0384, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "eval_loss": 0.06297700107097626, | |
| "eval_runtime": 1155.6941, | |
| "eval_samples_per_second": 1.621, | |
| "eval_steps_per_second": 0.102, | |
| "eval_wer": 32.04577968526466, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 1.5931034482758622e-06, | |
| "loss": 0.0254, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 1.4206896551724138e-06, | |
| "loss": 0.0574, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "eval_loss": 0.062077928334474564, | |
| "eval_runtime": 1073.2284, | |
| "eval_samples_per_second": 1.745, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 30.014306151645208, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 1.2482758620689655e-06, | |
| "loss": 0.0447, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 1.0758620689655174e-06, | |
| "loss": 0.053, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_loss": 0.06072333827614784, | |
| "eval_runtime": 1068.6925, | |
| "eval_samples_per_second": 1.753, | |
| "eval_steps_per_second": 0.11, | |
| "eval_wer": 30.529327610872674, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 9.034482758620689e-07, | |
| "loss": 0.0303, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 7.310344827586207e-07, | |
| "loss": 0.0217, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "eval_loss": 0.060529597103595734, | |
| "eval_runtime": 1230.1339, | |
| "eval_samples_per_second": 1.523, | |
| "eval_steps_per_second": 0.096, | |
| "eval_wer": 31.359084406294706, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 5.586206896551725e-07, | |
| "loss": 0.0394, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 3.862068965517242e-07, | |
| "loss": 0.0327, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "eval_loss": 0.060149554163217545, | |
| "eval_runtime": 1225.2391, | |
| "eval_samples_per_second": 1.529, | |
| "eval_steps_per_second": 0.096, | |
| "eval_wer": 31.301859799713878, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 2.1379310344827587e-07, | |
| "loss": 0.0326, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 4.137931034482759e-08, | |
| "loss": 0.0235, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "eval_loss": 0.05990791320800781, | |
| "eval_runtime": 1212.8468, | |
| "eval_samples_per_second": 1.544, | |
| "eval_steps_per_second": 0.097, | |
| "eval_wer": 30.90128755364807, | |
| "step": 1500 | |
| } | |
| ], | |
| "max_steps": 1500, | |
| "num_train_epochs": 3, | |
| "total_flos": 1.018527602098176e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |