| { | |
| "best_metric": 6.499429874572406, | |
| "best_model_checkpoint": "../whisper-medium-NST-uf-linlr/checkpoint-13000", | |
| "epoch": 2.1946, | |
| "global_step": 20000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.47e-06, | |
| "loss": 1.7524, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 4.970000000000001e-06, | |
| "loss": 0.3854, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 7.4700000000000005e-06, | |
| "loss": 0.2369, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 9.970000000000001e-06, | |
| "loss": 0.2046, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "eval_loss": 0.34258419275283813, | |
| "eval_runtime": 49.7358, | |
| "eval_samples_per_second": 2.011, | |
| "eval_steps_per_second": 0.141, | |
| "eval_wer": 15.279361459521096, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 9.87e-06, | |
| "loss": 0.1852, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 9.738421052631579e-06, | |
| "loss": 0.167, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.606842105263159e-06, | |
| "loss": 0.1547, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.475263157894737e-06, | |
| "loss": 0.148, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "eval_loss": 0.3284447491168976, | |
| "eval_runtime": 52.426, | |
| "eval_samples_per_second": 1.907, | |
| "eval_steps_per_second": 0.134, | |
| "eval_wer": 10.832383124287343, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.343684210526317e-06, | |
| "loss": 0.1401, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.212105263157895e-06, | |
| "loss": 0.1298, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.080526315789475e-06, | |
| "loss": 0.1252, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 8.948947368421053e-06, | |
| "loss": 0.121, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "eval_loss": 0.30924132466316223, | |
| "eval_runtime": 57.6047, | |
| "eval_samples_per_second": 1.736, | |
| "eval_steps_per_second": 0.122, | |
| "eval_wer": 12.884834663625996, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 8.817368421052631e-06, | |
| "loss": 0.1169, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 8.685789473684211e-06, | |
| "loss": 0.1143, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 8.55421052631579e-06, | |
| "loss": 0.1128, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 8.42263157894737e-06, | |
| "loss": 0.1089, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.2808048129081726, | |
| "eval_runtime": 46.0082, | |
| "eval_samples_per_second": 2.174, | |
| "eval_steps_per_second": 0.152, | |
| "eval_wer": 10.490307867730902, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 8.291052631578948e-06, | |
| "loss": 0.1041, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 8.159473684210528e-06, | |
| "loss": 0.1073, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 8.027894736842106e-06, | |
| "loss": 0.0996, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 7.896315789473686e-06, | |
| "loss": 0.0976, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.26165536046028137, | |
| "eval_runtime": 47.1016, | |
| "eval_samples_per_second": 2.123, | |
| "eval_steps_per_second": 0.149, | |
| "eval_wer": 9.92018244013683, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 7.764736842105264e-06, | |
| "loss": 0.099, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 7.633157894736842e-06, | |
| "loss": 0.0925, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 7.501578947368422e-06, | |
| "loss": 0.0918, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 7.370000000000001e-06, | |
| "loss": 0.0901, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "eval_loss": 0.26038578152656555, | |
| "eval_runtime": 72.3813, | |
| "eval_samples_per_second": 1.382, | |
| "eval_steps_per_second": 0.097, | |
| "eval_wer": 21.892816419612316, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 7.238421052631579e-06, | |
| "loss": 0.0881, | |
| "step": 6250 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.106842105263159e-06, | |
| "loss": 0.0908, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 6.975263157894737e-06, | |
| "loss": 0.0887, | |
| "step": 6750 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 6.843684210526317e-06, | |
| "loss": 0.0834, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "eval_loss": 0.2877318859100342, | |
| "eval_runtime": 47.1887, | |
| "eval_samples_per_second": 2.119, | |
| "eval_steps_per_second": 0.148, | |
| "eval_wer": 9.35005701254276, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 6.712105263157895e-06, | |
| "loss": 0.0828, | |
| "step": 7250 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 6.580526315789474e-06, | |
| "loss": 0.0825, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 6.448947368421053e-06, | |
| "loss": 0.085, | |
| "step": 7750 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 6.317368421052632e-06, | |
| "loss": 0.0825, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.2793583571910858, | |
| "eval_runtime": 46.2565, | |
| "eval_samples_per_second": 2.162, | |
| "eval_steps_per_second": 0.151, | |
| "eval_wer": 9.35005701254276, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 6.185789473684211e-06, | |
| "loss": 0.0783, | |
| "step": 8250 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 6.05421052631579e-06, | |
| "loss": 0.0708, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 5.922631578947369e-06, | |
| "loss": 0.0605, | |
| "step": 8750 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 5.791052631578948e-06, | |
| "loss": 0.0553, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "eval_loss": 0.2844734191894531, | |
| "eval_runtime": 54.3571, | |
| "eval_samples_per_second": 1.84, | |
| "eval_steps_per_second": 0.129, | |
| "eval_wer": 9.578107183580387, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 5.659473684210527e-06, | |
| "loss": 0.0505, | |
| "step": 9250 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 5.527894736842105e-06, | |
| "loss": 0.0475, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 5.396842105263158e-06, | |
| "loss": 0.0465, | |
| "step": 9750 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 5.265263157894738e-06, | |
| "loss": 0.0472, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_loss": 0.2814468741416931, | |
| "eval_runtime": 74.4128, | |
| "eval_samples_per_second": 1.344, | |
| "eval_steps_per_second": 0.094, | |
| "eval_wer": 24.173318129988598, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 5.133684210526316e-06, | |
| "loss": 0.0471, | |
| "step": 10250 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 5.002105263157895e-06, | |
| "loss": 0.0431, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 4.870526315789474e-06, | |
| "loss": 0.0396, | |
| "step": 10750 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 4.738947368421053e-06, | |
| "loss": 0.0409, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "eval_loss": 0.30840516090393066, | |
| "eval_runtime": 47.429, | |
| "eval_samples_per_second": 2.108, | |
| "eval_steps_per_second": 0.148, | |
| "eval_wer": 8.095781071835804, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 4.607368421052632e-06, | |
| "loss": 0.0401, | |
| "step": 11250 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 4.475789473684211e-06, | |
| "loss": 0.041, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 4.344210526315789e-06, | |
| "loss": 0.0394, | |
| "step": 11750 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 4.212631578947368e-06, | |
| "loss": 0.041, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.2865241765975952, | |
| "eval_runtime": 57.8167, | |
| "eval_samples_per_second": 1.73, | |
| "eval_steps_per_second": 0.121, | |
| "eval_wer": 9.236031927023944, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 4.0810526315789474e-06, | |
| "loss": 0.0383, | |
| "step": 12250 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 3.9494736842105265e-06, | |
| "loss": 0.0404, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 3.817894736842106e-06, | |
| "loss": 0.0374, | |
| "step": 12750 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 3.6863157894736847e-06, | |
| "loss": 0.0353, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "eval_loss": 0.2827776074409485, | |
| "eval_runtime": 49.2884, | |
| "eval_samples_per_second": 2.029, | |
| "eval_steps_per_second": 0.142, | |
| "eval_wer": 6.499429874572406, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 3.554736842105264e-06, | |
| "loss": 0.0387, | |
| "step": 13250 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 3.423157894736842e-06, | |
| "loss": 0.0353, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 3.292105263157895e-06, | |
| "loss": 0.0353, | |
| "step": 13750 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 3.160526315789474e-06, | |
| "loss": 0.0348, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "eval_loss": 0.27084633708000183, | |
| "eval_runtime": 56.2306, | |
| "eval_samples_per_second": 1.778, | |
| "eval_steps_per_second": 0.124, | |
| "eval_wer": 7.525655644241732, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 3.028947368421053e-06, | |
| "loss": 0.0367, | |
| "step": 14250 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 2.897368421052632e-06, | |
| "loss": 0.0369, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 2.7657894736842104e-06, | |
| "loss": 0.0354, | |
| "step": 14750 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 2.6342105263157895e-06, | |
| "loss": 0.0349, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "eval_loss": 0.28421100974082947, | |
| "eval_runtime": 69.8567, | |
| "eval_samples_per_second": 1.432, | |
| "eval_steps_per_second": 0.1, | |
| "eval_wer": 23.033067274800455, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 2.5026315789473686e-06, | |
| "loss": 0.033, | |
| "step": 15250 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 2.3710526315789477e-06, | |
| "loss": 0.0351, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 2.24e-06, | |
| "loss": 0.0355, | |
| "step": 15750 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2.1084210526315792e-06, | |
| "loss": 0.0361, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "eval_loss": 0.2769256830215454, | |
| "eval_runtime": 54.7414, | |
| "eval_samples_per_second": 1.827, | |
| "eval_steps_per_second": 0.128, | |
| "eval_wer": 10.148232611174459, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 1.9773684210526317e-06, | |
| "loss": 0.0338, | |
| "step": 16250 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 1.8457894736842108e-06, | |
| "loss": 0.03, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 1.7142105263157897e-06, | |
| "loss": 0.0263, | |
| "step": 16750 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.5826315789473687e-06, | |
| "loss": 0.0249, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "eval_loss": 0.2934916913509369, | |
| "eval_runtime": 60.1234, | |
| "eval_samples_per_second": 1.663, | |
| "eval_steps_per_second": 0.116, | |
| "eval_wer": 8.893956670467503, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 1.4510526315789474e-06, | |
| "loss": 0.0214, | |
| "step": 17250 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 1.3194736842105263e-06, | |
| "loss": 0.0197, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 1.1878947368421054e-06, | |
| "loss": 0.0208, | |
| "step": 17750 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 1.0563157894736843e-06, | |
| "loss": 0.0204, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "eval_loss": 0.28742682933807373, | |
| "eval_runtime": 70.388, | |
| "eval_samples_per_second": 1.421, | |
| "eval_steps_per_second": 0.099, | |
| "eval_wer": 12.428734321550742, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 9.247368421052633e-07, | |
| "loss": 0.0207, | |
| "step": 18250 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 7.931578947368422e-07, | |
| "loss": 0.0188, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 6.615789473684211e-07, | |
| "loss": 0.0167, | |
| "step": 18750 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 5.3e-07, | |
| "loss": 0.0175, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "eval_loss": 0.28821495175361633, | |
| "eval_runtime": 69.297, | |
| "eval_samples_per_second": 1.443, | |
| "eval_steps_per_second": 0.101, | |
| "eval_wer": 12.998859749144811, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 3.98421052631579e-07, | |
| "loss": 0.0184, | |
| "step": 19250 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 2.6684210526315793e-07, | |
| "loss": 0.0186, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 1.3526315789473686e-07, | |
| "loss": 0.0179, | |
| "step": 19750 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 3.6842105263157898e-09, | |
| "loss": 0.0197, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "eval_loss": 0.3006799817085266, | |
| "eval_runtime": 45.4742, | |
| "eval_samples_per_second": 2.199, | |
| "eval_steps_per_second": 0.154, | |
| "eval_wer": 9.122006841505131, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "step": 20000, | |
| "total_flos": 2.077757165223936e+20, | |
| "train_loss": 0.09227749185562134, | |
| "train_runtime": 214472.7404, | |
| "train_samples_per_second": 6.714, | |
| "train_steps_per_second": 0.093 | |
| } | |
| ], | |
| "max_steps": 20000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 2.077757165223936e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |