| { | |
| "best_metric": 0.6922819018363953, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/whisper-medium-bsbigcgen-male-model/checkpoint-800", | |
| "epoch": 4.487961476725522, | |
| "eval_steps": 200, | |
| "global_step": 1400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08025682182985554, | |
| "grad_norm": 128.6523895263672, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 12.3604, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16051364365971107, | |
| "grad_norm": 105.59561920166016, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 10.9708, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.24077046548956663, | |
| "grad_norm": 80.34696197509766, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 8.8039, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.32102728731942215, | |
| "grad_norm": 82.3702621459961, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 6.3185, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4012841091492777, | |
| "grad_norm": 69.2629623413086, | |
| "learning_rate": 2.38e-06, | |
| "loss": 5.5448, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.48154093097913325, | |
| "grad_norm": 57.369197845458984, | |
| "learning_rate": 2.88e-06, | |
| "loss": 4.7462, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5617977528089888, | |
| "grad_norm": 66.99210357666016, | |
| "learning_rate": 3.3800000000000007e-06, | |
| "loss": 4.7684, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.6420545746388443, | |
| "grad_norm": 68.48348236083984, | |
| "learning_rate": 3.88e-06, | |
| "loss": 4.039, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6420545746388443, | |
| "eval_loss": 0.9763619303703308, | |
| "eval_runtime": 251.081, | |
| "eval_samples_per_second": 1.713, | |
| "eval_steps_per_second": 0.856, | |
| "eval_wer": 0.694201564657156, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.7223113964686998, | |
| "grad_norm": 73.68655395507812, | |
| "learning_rate": 4.38e-06, | |
| "loss": 4.0338, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.8025682182985554, | |
| "grad_norm": 69.01504516601562, | |
| "learning_rate": 4.880000000000001e-06, | |
| "loss": 3.9039, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8828250401284109, | |
| "grad_norm": 60.503868103027344, | |
| "learning_rate": 5.380000000000001e-06, | |
| "loss": 3.761, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9630818619582665, | |
| "grad_norm": 53.41858673095703, | |
| "learning_rate": 5.8800000000000005e-06, | |
| "loss": 3.6084, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.0417335473515248, | |
| "grad_norm": 68.72418975830078, | |
| "learning_rate": 6.380000000000001e-06, | |
| "loss": 2.9504, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.1219903691813804, | |
| "grad_norm": 60.11503982543945, | |
| "learning_rate": 6.88e-06, | |
| "loss": 2.6441, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.202247191011236, | |
| "grad_norm": 52.51095962524414, | |
| "learning_rate": 7.3800000000000005e-06, | |
| "loss": 2.7013, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.2825040128410916, | |
| "grad_norm": 56.025001525878906, | |
| "learning_rate": 7.88e-06, | |
| "loss": 2.6473, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2825040128410916, | |
| "eval_loss": 0.7705269455909729, | |
| "eval_runtime": 246.2904, | |
| "eval_samples_per_second": 1.746, | |
| "eval_steps_per_second": 0.873, | |
| "eval_wer": 0.5812241141279337, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.362760834670947, | |
| "grad_norm": 56.739906311035156, | |
| "learning_rate": 8.380000000000001e-06, | |
| "loss": 2.6946, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.4430176565008026, | |
| "grad_norm": 40.198238372802734, | |
| "learning_rate": 8.880000000000001e-06, | |
| "loss": 2.5734, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.523274478330658, | |
| "grad_norm": 53.963035583496094, | |
| "learning_rate": 9.38e-06, | |
| "loss": 2.5631, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.6035313001605136, | |
| "grad_norm": 58.04458999633789, | |
| "learning_rate": 9.88e-06, | |
| "loss": 2.7175, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.6837881219903692, | |
| "grad_norm": 76.02864074707031, | |
| "learning_rate": 9.978482446206116e-06, | |
| "loss": 2.9013, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.7640449438202248, | |
| "grad_norm": 55.0551872253418, | |
| "learning_rate": 9.950169875424689e-06, | |
| "loss": 2.7567, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.8443017656500804, | |
| "grad_norm": 45.51343536376953, | |
| "learning_rate": 9.921857304643261e-06, | |
| "loss": 2.5122, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.9245585874799358, | |
| "grad_norm": 45.42640686035156, | |
| "learning_rate": 9.893544733861835e-06, | |
| "loss": 2.3913, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.9245585874799358, | |
| "eval_loss": 0.700962483882904, | |
| "eval_runtime": 243.9773, | |
| "eval_samples_per_second": 1.762, | |
| "eval_steps_per_second": 0.881, | |
| "eval_wer": 0.5269213069489186, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.0032102728731944, | |
| "grad_norm": 30.623844146728516, | |
| "learning_rate": 9.865232163080408e-06, | |
| "loss": 2.4991, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.0834670947030496, | |
| "grad_norm": 52.41933059692383, | |
| "learning_rate": 9.836919592298982e-06, | |
| "loss": 1.1913, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.163723916532905, | |
| "grad_norm": 40.88090896606445, | |
| "learning_rate": 9.808607021517554e-06, | |
| "loss": 1.4546, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.243980738362761, | |
| "grad_norm": 38.167205810546875, | |
| "learning_rate": 9.780294450736127e-06, | |
| "loss": 1.4931, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.3242375601926164, | |
| "grad_norm": 31.256465911865234, | |
| "learning_rate": 9.751981879954701e-06, | |
| "loss": 1.4755, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.404494382022472, | |
| "grad_norm": 46.689884185791016, | |
| "learning_rate": 9.723669309173273e-06, | |
| "loss": 1.4373, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.4847512038523276, | |
| "grad_norm": 50.231746673583984, | |
| "learning_rate": 9.695356738391848e-06, | |
| "loss": 1.5603, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.5650080256821832, | |
| "grad_norm": 49.86467742919922, | |
| "learning_rate": 9.66704416761042e-06, | |
| "loss": 1.6283, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.5650080256821832, | |
| "eval_loss": 0.6922819018363953, | |
| "eval_runtime": 248.6409, | |
| "eval_samples_per_second": 1.729, | |
| "eval_steps_per_second": 0.865, | |
| "eval_wer": 0.5177174413253567, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.6452648475120384, | |
| "grad_norm": 43.58526611328125, | |
| "learning_rate": 9.638731596828992e-06, | |
| "loss": 1.717, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.725521669341894, | |
| "grad_norm": 34.9052619934082, | |
| "learning_rate": 9.610419026047567e-06, | |
| "loss": 1.6415, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.8057784911717496, | |
| "grad_norm": 45.21881103515625, | |
| "learning_rate": 9.582106455266139e-06, | |
| "loss": 1.4356, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.886035313001605, | |
| "grad_norm": 57.4312629699707, | |
| "learning_rate": 9.553793884484713e-06, | |
| "loss": 1.536, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.966292134831461, | |
| "grad_norm": 42.61119079589844, | |
| "learning_rate": 9.525481313703286e-06, | |
| "loss": 1.5945, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.044943820224719, | |
| "grad_norm": 22.571409225463867, | |
| "learning_rate": 9.497168742921858e-06, | |
| "loss": 0.9745, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.125200642054575, | |
| "grad_norm": 26.83526611328125, | |
| "learning_rate": 9.468856172140432e-06, | |
| "loss": 0.7011, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.20545746388443, | |
| "grad_norm": 32.36039733886719, | |
| "learning_rate": 9.440543601359004e-06, | |
| "loss": 0.7783, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.20545746388443, | |
| "eval_loss": 0.7241026163101196, | |
| "eval_runtime": 248.287, | |
| "eval_samples_per_second": 1.732, | |
| "eval_steps_per_second": 0.866, | |
| "eval_wer": 0.5131155085135757, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.2857142857142856, | |
| "grad_norm": 42.650146484375, | |
| "learning_rate": 9.412231030577577e-06, | |
| "loss": 0.6489, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.365971107544141, | |
| "grad_norm": 29.87259864807129, | |
| "learning_rate": 9.383918459796151e-06, | |
| "loss": 0.9697, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.446227929373997, | |
| "grad_norm": 32.14972686767578, | |
| "learning_rate": 9.355605889014723e-06, | |
| "loss": 0.7128, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.5264847512038524, | |
| "grad_norm": 40.77785110473633, | |
| "learning_rate": 9.327293318233296e-06, | |
| "loss": 0.8084, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.606741573033708, | |
| "grad_norm": 39.83961486816406, | |
| "learning_rate": 9.298980747451868e-06, | |
| "loss": 0.8268, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.686998394863563, | |
| "grad_norm": 34.79006576538086, | |
| "learning_rate": 9.270668176670442e-06, | |
| "loss": 0.6832, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.767255216693419, | |
| "grad_norm": 24.103979110717773, | |
| "learning_rate": 9.242355605889015e-06, | |
| "loss": 0.7754, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.8475120385232744, | |
| "grad_norm": 42.03251647949219, | |
| "learning_rate": 9.214043035107589e-06, | |
| "loss": 0.8723, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.8475120385232744, | |
| "eval_loss": 0.720678985118866, | |
| "eval_runtime": 245.5357, | |
| "eval_samples_per_second": 1.751, | |
| "eval_steps_per_second": 0.876, | |
| "eval_wer": 0.5140358950759318, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.92776886035313, | |
| "grad_norm": 27.95992088317871, | |
| "learning_rate": 9.185730464326161e-06, | |
| "loss": 0.8405, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.006420545746389, | |
| "grad_norm": 15.731476783752441, | |
| "learning_rate": 9.157417893544734e-06, | |
| "loss": 0.6524, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.086677367576244, | |
| "grad_norm": 15.728172302246094, | |
| "learning_rate": 9.129105322763308e-06, | |
| "loss": 0.3319, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.166934189406099, | |
| "grad_norm": 30.88848876953125, | |
| "learning_rate": 9.10079275198188e-06, | |
| "loss": 0.4029, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.247191011235955, | |
| "grad_norm": 20.073183059692383, | |
| "learning_rate": 9.072480181200455e-06, | |
| "loss": 0.4016, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 4.32744783306581, | |
| "grad_norm": 21.41939353942871, | |
| "learning_rate": 9.044167610419027e-06, | |
| "loss": 0.3585, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.407704654895666, | |
| "grad_norm": 25.81800651550293, | |
| "learning_rate": 9.0158550396376e-06, | |
| "loss": 0.3965, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.487961476725522, | |
| "grad_norm": 21.839908599853516, | |
| "learning_rate": 8.987542468856174e-06, | |
| "loss": 0.3633, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.487961476725522, | |
| "eval_loss": 0.7589691281318665, | |
| "eval_runtime": 247.0866, | |
| "eval_samples_per_second": 1.74, | |
| "eval_steps_per_second": 0.87, | |
| "eval_wer": 0.49608835710998617, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.487961476725522, | |
| "step": 1400, | |
| "total_flos": 1.14103628660736e+19, | |
| "train_loss": 2.465633350099836, | |
| "train_runtime": 4460.9347, | |
| "train_samples_per_second": 16.752, | |
| "train_steps_per_second": 2.091 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 9330, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.14103628660736e+19, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |