| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.4596464258262873, | |
| "eval_steps": 300, | |
| "global_step": 21600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.034161755914253994, | |
| "grad_norm": 121.1594467163086, | |
| "learning_rate": 2.97e-05, | |
| "loss": 2.9674, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.034161755914253994, | |
| "eval_loss": 1.3304753303527832, | |
| "eval_runtime": 352.5943, | |
| "eval_samples_per_second": 10.488, | |
| "eval_steps_per_second": 10.488, | |
| "eval_wer": 0.7000701638532337, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.06832351182850799, | |
| "grad_norm": 23.02712631225586, | |
| "learning_rate": 4.986136783733826e-05, | |
| "loss": 1.2476, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.06832351182850799, | |
| "eval_loss": 1.1660360097885132, | |
| "eval_runtime": 306.2849, | |
| "eval_samples_per_second": 12.074, | |
| "eval_steps_per_second": 12.074, | |
| "eval_wer": 0.5879318172437987, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.10248526774276198, | |
| "grad_norm": 25.825851440429688, | |
| "learning_rate": 4.942814232902033e-05, | |
| "loss": 1.0692, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.10248526774276198, | |
| "eval_loss": 0.9110434055328369, | |
| "eval_runtime": 299.9179, | |
| "eval_samples_per_second": 12.33, | |
| "eval_steps_per_second": 12.33, | |
| "eval_wer": 0.4886293284906517, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.13664702365701598, | |
| "grad_norm": 34.84864044189453, | |
| "learning_rate": 4.899491682070241e-05, | |
| "loss": 0.9443, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.13664702365701598, | |
| "eval_loss": 0.7600523233413696, | |
| "eval_runtime": 300.4784, | |
| "eval_samples_per_second": 12.307, | |
| "eval_steps_per_second": 12.307, | |
| "eval_wer": 0.4726567336662677, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.17080877957126997, | |
| "grad_norm": 80.90399932861328, | |
| "learning_rate": 4.856169131238448e-05, | |
| "loss": 0.8235, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17080877957126997, | |
| "eval_loss": 0.7760590314865112, | |
| "eval_runtime": 301.4362, | |
| "eval_samples_per_second": 12.268, | |
| "eval_steps_per_second": 12.268, | |
| "eval_wer": 0.3972718642948533, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.20497053548552396, | |
| "grad_norm": 41.970706939697266, | |
| "learning_rate": 4.8128465804066546e-05, | |
| "loss": 0.8155, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.20497053548552396, | |
| "eval_loss": 0.7084236741065979, | |
| "eval_runtime": 300.2268, | |
| "eval_samples_per_second": 12.317, | |
| "eval_steps_per_second": 12.317, | |
| "eval_wer": 0.40218333402121426, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.23913229139977796, | |
| "grad_norm": 29.83930206298828, | |
| "learning_rate": 4.769668438077634e-05, | |
| "loss": 0.767, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.23913229139977796, | |
| "eval_loss": 0.6251249313354492, | |
| "eval_runtime": 298.3691, | |
| "eval_samples_per_second": 12.394, | |
| "eval_steps_per_second": 12.394, | |
| "eval_wer": 0.37560361550208426, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.27329404731403195, | |
| "grad_norm": 203.37936401367188, | |
| "learning_rate": 4.726345887245841e-05, | |
| "loss": 0.7517, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.27329404731403195, | |
| "eval_loss": 0.6125071048736572, | |
| "eval_runtime": 297.9477, | |
| "eval_samples_per_second": 12.412, | |
| "eval_steps_per_second": 12.412, | |
| "eval_wer": 0.36536794750092866, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.3074558032282859, | |
| "grad_norm": 15.38986873626709, | |
| "learning_rate": 4.683023336414049e-05, | |
| "loss": 0.687, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3074558032282859, | |
| "eval_loss": 0.5848016142845154, | |
| "eval_runtime": 299.5504, | |
| "eval_samples_per_second": 12.345, | |
| "eval_steps_per_second": 12.345, | |
| "eval_wer": 0.3439266994097982, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.34161755914253994, | |
| "grad_norm": 32.728450775146484, | |
| "learning_rate": 4.6397007855822556e-05, | |
| "loss": 0.6509, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.34161755914253994, | |
| "eval_loss": 0.564321756362915, | |
| "eval_runtime": 300.1274, | |
| "eval_samples_per_second": 12.321, | |
| "eval_steps_per_second": 12.321, | |
| "eval_wer": 0.328160468859631, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3757793150567939, | |
| "grad_norm": 30.201148986816406, | |
| "learning_rate": 4.5963782347504624e-05, | |
| "loss": 0.6632, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.3757793150567939, | |
| "eval_loss": 0.5509063005447388, | |
| "eval_runtime": 297.3553, | |
| "eval_samples_per_second": 12.436, | |
| "eval_steps_per_second": 12.436, | |
| "eval_wer": 0.3199265343183788, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.4099410709710479, | |
| "grad_norm": 105.19284057617188, | |
| "learning_rate": 4.553055683918669e-05, | |
| "loss": 0.6108, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.4099410709710479, | |
| "eval_loss": 0.5392822623252869, | |
| "eval_runtime": 300.6256, | |
| "eval_samples_per_second": 12.301, | |
| "eval_steps_per_second": 12.301, | |
| "eval_wer": 0.3341243963844979, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.4441028268853019, | |
| "grad_norm": 4.016674041748047, | |
| "learning_rate": 4.5097331330868766e-05, | |
| "loss": 0.5898, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.4441028268853019, | |
| "eval_loss": 0.5222619771957397, | |
| "eval_runtime": 298.4266, | |
| "eval_samples_per_second": 12.392, | |
| "eval_steps_per_second": 12.392, | |
| "eval_wer": 0.32772710388377563, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.4782645827995559, | |
| "grad_norm": 27.7945613861084, | |
| "learning_rate": 4.4664105822550834e-05, | |
| "loss": 0.595, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.4782645827995559, | |
| "eval_loss": 0.5199422836303711, | |
| "eval_runtime": 301.9933, | |
| "eval_samples_per_second": 12.245, | |
| "eval_steps_per_second": 12.245, | |
| "eval_wer": 0.31996780717322215, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.5124263387138099, | |
| "grad_norm": 37.06507873535156, | |
| "learning_rate": 4.42308803142329e-05, | |
| "loss": 0.5644, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5124263387138099, | |
| "eval_loss": 0.5507590174674988, | |
| "eval_runtime": 302.7901, | |
| "eval_samples_per_second": 12.213, | |
| "eval_steps_per_second": 12.213, | |
| "eval_wer": 0.29186099302488755, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5465880946280639, | |
| "grad_norm": 62.698890686035156, | |
| "learning_rate": 4.37990988909427e-05, | |
| "loss": 0.5787, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5465880946280639, | |
| "eval_loss": 0.4994393587112427, | |
| "eval_runtime": 301.2905, | |
| "eval_samples_per_second": 12.274, | |
| "eval_steps_per_second": 12.274, | |
| "eval_wer": 0.3059763093813199, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5807498505423179, | |
| "grad_norm": 18.494293212890625, | |
| "learning_rate": 4.336587338262477e-05, | |
| "loss": 0.5752, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5807498505423179, | |
| "eval_loss": 0.49656036496162415, | |
| "eval_runtime": 303.7733, | |
| "eval_samples_per_second": 12.174, | |
| "eval_steps_per_second": 12.174, | |
| "eval_wer": 0.29966156259028437, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.6149116064565718, | |
| "grad_norm": 38.18902587890625, | |
| "learning_rate": 4.2932647874306844e-05, | |
| "loss": 0.5353, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.6149116064565718, | |
| "eval_loss": 0.47309231758117676, | |
| "eval_runtime": 302.8595, | |
| "eval_samples_per_second": 12.21, | |
| "eval_steps_per_second": 12.21, | |
| "eval_wer": 0.3237030005365471, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.6490733623708259, | |
| "grad_norm": 15.917391777038574, | |
| "learning_rate": 4.249942236598891e-05, | |
| "loss": 0.5473, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6490733623708259, | |
| "eval_loss": 0.46649235486984253, | |
| "eval_runtime": 304.2045, | |
| "eval_samples_per_second": 12.156, | |
| "eval_steps_per_second": 12.156, | |
| "eval_wer": 0.3062033100829584, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6832351182850799, | |
| "grad_norm": 17.125015258789062, | |
| "learning_rate": 4.206619685767098e-05, | |
| "loss": 0.5498, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6832351182850799, | |
| "eval_loss": 0.48897644877433777, | |
| "eval_runtime": 305.5464, | |
| "eval_samples_per_second": 12.103, | |
| "eval_steps_per_second": 12.103, | |
| "eval_wer": 0.2875892525485988, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.7173968741993338, | |
| "grad_norm": 18.530088424682617, | |
| "learning_rate": 4.1632971349353054e-05, | |
| "loss": 0.5146, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.7173968741993338, | |
| "eval_loss": 0.4747436046600342, | |
| "eval_runtime": 304.4465, | |
| "eval_samples_per_second": 12.147, | |
| "eval_steps_per_second": 12.147, | |
| "eval_wer": 0.2926451772669115, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.7515586301135878, | |
| "grad_norm": 24.51532554626465, | |
| "learning_rate": 4.119974584103512e-05, | |
| "loss": 0.5398, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7515586301135878, | |
| "eval_loss": 0.45806387066841125, | |
| "eval_runtime": 303.5016, | |
| "eval_samples_per_second": 12.184, | |
| "eval_steps_per_second": 12.184, | |
| "eval_wer": 0.29072598951669487, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7857203860278418, | |
| "grad_norm": 18.871652603149414, | |
| "learning_rate": 4.076652033271719e-05, | |
| "loss": 0.5154, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.7857203860278418, | |
| "eval_loss": 0.4556685984134674, | |
| "eval_runtime": 301.5962, | |
| "eval_samples_per_second": 12.261, | |
| "eval_steps_per_second": 12.261, | |
| "eval_wer": 0.29951710759833255, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.8198821419420959, | |
| "grad_norm": 80.3409194946289, | |
| "learning_rate": 4.0333294824399264e-05, | |
| "loss": 0.5386, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.8198821419420959, | |
| "eval_loss": 0.45153045654296875, | |
| "eval_runtime": 302.4297, | |
| "eval_samples_per_second": 12.228, | |
| "eval_steps_per_second": 12.228, | |
| "eval_wer": 0.2947500928639234, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.8540438978563498, | |
| "grad_norm": 4.5673346519470215, | |
| "learning_rate": 3.990006931608133e-05, | |
| "loss": 0.5037, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8540438978563498, | |
| "eval_loss": 0.44558125734329224, | |
| "eval_runtime": 306.0244, | |
| "eval_samples_per_second": 12.084, | |
| "eval_steps_per_second": 12.084, | |
| "eval_wer": 0.2961120970737546, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8882056537706038, | |
| "grad_norm": 11.731285095214844, | |
| "learning_rate": 3.9466843807763406e-05, | |
| "loss": 0.5344, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.8882056537706038, | |
| "eval_loss": 0.45088666677474976, | |
| "eval_runtime": 303.7688, | |
| "eval_samples_per_second": 12.174, | |
| "eval_steps_per_second": 12.174, | |
| "eval_wer": 0.2987741962111519, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.9223674096848578, | |
| "grad_norm": 21.575855255126953, | |
| "learning_rate": 3.90350623844732e-05, | |
| "loss": 0.501, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.9223674096848578, | |
| "eval_loss": 0.44363677501678467, | |
| "eval_runtime": 305.2285, | |
| "eval_samples_per_second": 12.116, | |
| "eval_steps_per_second": 12.116, | |
| "eval_wer": 0.2710594741838293, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.9565291655991118, | |
| "grad_norm": 89.06358337402344, | |
| "learning_rate": 3.860183687615527e-05, | |
| "loss": 0.487, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.9565291655991118, | |
| "eval_loss": 0.42328470945358276, | |
| "eval_runtime": 303.3048, | |
| "eval_samples_per_second": 12.192, | |
| "eval_steps_per_second": 12.192, | |
| "eval_wer": 0.27491848611168435, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.9906909215133658, | |
| "grad_norm": 2.7736120223999023, | |
| "learning_rate": 3.816861136783734e-05, | |
| "loss": 0.4692, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.9906909215133658, | |
| "eval_loss": 0.46611106395721436, | |
| "eval_runtime": 303.8834, | |
| "eval_samples_per_second": 12.169, | |
| "eval_steps_per_second": 12.169, | |
| "eval_wer": 0.25322960089149366, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 1.0248526774276199, | |
| "grad_norm": 2.908313035964966, | |
| "learning_rate": 3.773538585951941e-05, | |
| "loss": 0.462, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0248526774276199, | |
| "eval_loss": 0.4197135269641876, | |
| "eval_runtime": 302.465, | |
| "eval_samples_per_second": 12.226, | |
| "eval_steps_per_second": 12.226, | |
| "eval_wer": 0.2722770234017087, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0590144333418738, | |
| "grad_norm": 1.9524333477020264, | |
| "learning_rate": 3.7302160351201484e-05, | |
| "loss": 0.4508, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.0590144333418738, | |
| "eval_loss": 0.4316381514072418, | |
| "eval_runtime": 304.5385, | |
| "eval_samples_per_second": 12.143, | |
| "eval_steps_per_second": 12.143, | |
| "eval_wer": 0.25838870774691486, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.0931761892561278, | |
| "grad_norm": 2.584221124649048, | |
| "learning_rate": 3.686893484288355e-05, | |
| "loss": 0.4702, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.0931761892561278, | |
| "eval_loss": 0.4147655963897705, | |
| "eval_runtime": 302.8309, | |
| "eval_samples_per_second": 12.211, | |
| "eval_steps_per_second": 12.211, | |
| "eval_wer": 0.2689132857319741, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.1273379451703818, | |
| "grad_norm": 1.6557509899139404, | |
| "learning_rate": 3.643570933456562e-05, | |
| "loss": 0.4517, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.1273379451703818, | |
| "eval_loss": 0.39497703313827515, | |
| "eval_runtime": 304.7729, | |
| "eval_samples_per_second": 12.134, | |
| "eval_steps_per_second": 12.134, | |
| "eval_wer": 0.2549217879400718, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.1614997010846357, | |
| "grad_norm": 1.9434764385223389, | |
| "learning_rate": 3.6002483826247694e-05, | |
| "loss": 0.4408, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.1614997010846357, | |
| "eval_loss": 0.43082132935523987, | |
| "eval_runtime": 303.9383, | |
| "eval_samples_per_second": 12.167, | |
| "eval_steps_per_second": 12.167, | |
| "eval_wer": 0.25514878864171037, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.1956614569988897, | |
| "grad_norm": 0.8506720066070557, | |
| "learning_rate": 3.556925831792976e-05, | |
| "loss": 0.4636, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.1956614569988897, | |
| "eval_loss": 0.4033319354057312, | |
| "eval_runtime": 302.9721, | |
| "eval_samples_per_second": 12.206, | |
| "eval_steps_per_second": 12.206, | |
| "eval_wer": 0.26996574353048003, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.2298232129131437, | |
| "grad_norm": 4.787822246551514, | |
| "learning_rate": 3.513747689463956e-05, | |
| "loss": 0.4583, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.2298232129131437, | |
| "eval_loss": 0.40962740778923035, | |
| "eval_runtime": 301.5186, | |
| "eval_samples_per_second": 12.265, | |
| "eval_steps_per_second": 12.265, | |
| "eval_wer": 0.2555821536175657, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.2639849688273976, | |
| "grad_norm": 6.576323986053467, | |
| "learning_rate": 3.470425138632163e-05, | |
| "loss": 0.4315, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.2639849688273976, | |
| "eval_loss": 0.3883008658885956, | |
| "eval_runtime": 299.5319, | |
| "eval_samples_per_second": 12.346, | |
| "eval_steps_per_second": 12.346, | |
| "eval_wer": 0.26812910148995006, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.2981467247416516, | |
| "grad_norm": 1.379813313484192, | |
| "learning_rate": 3.427246996303142e-05, | |
| "loss": 0.4172, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.2981467247416516, | |
| "eval_loss": 0.37370720505714417, | |
| "eval_runtime": 301.6675, | |
| "eval_samples_per_second": 12.259, | |
| "eval_steps_per_second": 12.259, | |
| "eval_wer": 0.2529200544801684, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.3323084806559058, | |
| "grad_norm": 2.169112205505371, | |
| "learning_rate": 3.3839244454713497e-05, | |
| "loss": 0.4177, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.3323084806559058, | |
| "eval_loss": 0.399240642786026, | |
| "eval_runtime": 301.556, | |
| "eval_samples_per_second": 12.263, | |
| "eval_steps_per_second": 12.263, | |
| "eval_wer": 0.24724503693920508, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.3664702365701598, | |
| "grad_norm": 1.855020523071289, | |
| "learning_rate": 3.3406018946395564e-05, | |
| "loss": 0.3975, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.3664702365701598, | |
| "eval_loss": 0.3716418743133545, | |
| "eval_runtime": 301.4037, | |
| "eval_samples_per_second": 12.269, | |
| "eval_steps_per_second": 12.269, | |
| "eval_wer": 0.24852449543934954, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.4006319924844137, | |
| "grad_norm": 3.5865867137908936, | |
| "learning_rate": 3.297279343807764e-05, | |
| "loss": 0.4044, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.4006319924844137, | |
| "eval_loss": 0.3852696120738983, | |
| "eval_runtime": 299.7735, | |
| "eval_samples_per_second": 12.336, | |
| "eval_steps_per_second": 12.336, | |
| "eval_wer": 0.25232159808493954, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.4347937483986677, | |
| "grad_norm": 7.861583232879639, | |
| "learning_rate": 3.2539567929759707e-05, | |
| "loss": 0.4497, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.4347937483986677, | |
| "eval_loss": 0.37975797057151794, | |
| "eval_runtime": 303.0367, | |
| "eval_samples_per_second": 12.203, | |
| "eval_steps_per_second": 12.203, | |
| "eval_wer": 0.24646085269718107, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.4689555043129217, | |
| "grad_norm": 2.6698834896087646, | |
| "learning_rate": 3.2106342421441774e-05, | |
| "loss": 0.4188, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 1.4689555043129217, | |
| "eval_loss": 0.38219255208969116, | |
| "eval_runtime": 301.4473, | |
| "eval_samples_per_second": 12.267, | |
| "eval_steps_per_second": 12.267, | |
| "eval_wer": 0.24937058896363862, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 1.5031172602271756, | |
| "grad_norm": 2.8027360439300537, | |
| "learning_rate": 3.167311691312385e-05, | |
| "loss": 0.4424, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 1.5031172602271756, | |
| "eval_loss": 0.3559640944004059, | |
| "eval_runtime": 300.0982, | |
| "eval_samples_per_second": 12.323, | |
| "eval_steps_per_second": 12.323, | |
| "eval_wer": 0.24485121135828966, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 1.5372790161414298, | |
| "grad_norm": 3.83882212638855, | |
| "learning_rate": 3.1239891404805917e-05, | |
| "loss": 0.4249, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.5372790161414298, | |
| "eval_loss": 0.3629530966281891, | |
| "eval_runtime": 303.9778, | |
| "eval_samples_per_second": 12.165, | |
| "eval_steps_per_second": 12.165, | |
| "eval_wer": 0.25139295885096374, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.5714407720556838, | |
| "grad_norm": 3.2446749210357666, | |
| "learning_rate": 3.0806665896487984e-05, | |
| "loss": 0.4287, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 1.5714407720556838, | |
| "eval_loss": 0.3662088215351105, | |
| "eval_runtime": 303.0939, | |
| "eval_samples_per_second": 12.201, | |
| "eval_steps_per_second": 12.201, | |
| "eval_wer": 0.24173511081761526, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 1.6056025279699377, | |
| "grad_norm": 1.8408727645874023, | |
| "learning_rate": 3.0373440388170055e-05, | |
| "loss": 0.3712, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 1.6056025279699377, | |
| "eval_loss": 0.3713747262954712, | |
| "eval_runtime": 301.435, | |
| "eval_samples_per_second": 12.268, | |
| "eval_steps_per_second": 12.268, | |
| "eval_wer": 0.2561806100127946, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 1.6397642838841917, | |
| "grad_norm": 1.50343656539917, | |
| "learning_rate": 2.994021487985213e-05, | |
| "loss": 0.3893, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 1.6397642838841917, | |
| "eval_loss": 0.3711312413215637, | |
| "eval_runtime": 300.5222, | |
| "eval_samples_per_second": 12.305, | |
| "eval_steps_per_second": 12.305, | |
| "eval_wer": 0.23325353914730282, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 1.6739260397984457, | |
| "grad_norm": 2.6718335151672363, | |
| "learning_rate": 2.95069893715342e-05, | |
| "loss": 0.3935, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 1.6739260397984457, | |
| "eval_loss": 0.3715294897556305, | |
| "eval_runtime": 301.8378, | |
| "eval_samples_per_second": 12.252, | |
| "eval_steps_per_second": 12.252, | |
| "eval_wer": 0.2412604729869165, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 1.7080877957126996, | |
| "grad_norm": 1.3738945722579956, | |
| "learning_rate": 2.907376386321627e-05, | |
| "loss": 0.3982, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.7080877957126996, | |
| "eval_loss": 0.3551383316516876, | |
| "eval_runtime": 302.7384, | |
| "eval_samples_per_second": 12.215, | |
| "eval_steps_per_second": 12.215, | |
| "eval_wer": 0.24821494902802427, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.7422495516269536, | |
| "grad_norm": 2.5263662338256836, | |
| "learning_rate": 2.864053835489834e-05, | |
| "loss": 0.4124, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 1.7422495516269536, | |
| "eval_loss": 0.3519354462623596, | |
| "eval_runtime": 303.939, | |
| "eval_samples_per_second": 12.167, | |
| "eval_steps_per_second": 12.167, | |
| "eval_wer": 0.24121920013207312, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 1.7764113075412076, | |
| "grad_norm": 1.8874987363815308, | |
| "learning_rate": 2.8207312846580408e-05, | |
| "loss": 0.3853, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 1.7764113075412076, | |
| "eval_loss": 0.3428538739681244, | |
| "eval_runtime": 303.4637, | |
| "eval_samples_per_second": 12.186, | |
| "eval_steps_per_second": 12.186, | |
| "eval_wer": 0.2417970200998803, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 1.8105730634554615, | |
| "grad_norm": 1.8015034198760986, | |
| "learning_rate": 2.777408733826248e-05, | |
| "loss": 0.4096, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 1.8105730634554615, | |
| "eval_loss": 0.34074389934539795, | |
| "eval_runtime": 301.0523, | |
| "eval_samples_per_second": 12.284, | |
| "eval_steps_per_second": 12.284, | |
| "eval_wer": 0.23938255809154318, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 1.8447348193697155, | |
| "grad_norm": 1.9667861461639404, | |
| "learning_rate": 2.7340861829944547e-05, | |
| "loss": 0.3816, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 1.8447348193697155, | |
| "eval_loss": 0.360734224319458, | |
| "eval_runtime": 301.5423, | |
| "eval_samples_per_second": 12.264, | |
| "eval_steps_per_second": 12.264, | |
| "eval_wer": 0.23700936893804944, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 1.8788965752839695, | |
| "grad_norm": 1.8239201307296753, | |
| "learning_rate": 2.6909080406654346e-05, | |
| "loss": 0.3769, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.8788965752839695, | |
| "eval_loss": 0.36014851927757263, | |
| "eval_runtime": 301.5685, | |
| "eval_samples_per_second": 12.263, | |
| "eval_steps_per_second": 12.263, | |
| "eval_wer": 0.22910561723554418, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.9130583311982234, | |
| "grad_norm": 1.5791462659835815, | |
| "learning_rate": 2.6475854898336417e-05, | |
| "loss": 0.3428, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 1.9130583311982234, | |
| "eval_loss": 0.3577890694141388, | |
| "eval_runtime": 299.2343, | |
| "eval_samples_per_second": 12.358, | |
| "eval_steps_per_second": 12.358, | |
| "eval_wer": 0.22830079656609847, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 1.9472200871124776, | |
| "grad_norm": 2.0596249103546143, | |
| "learning_rate": 2.6042629390018485e-05, | |
| "loss": 0.3636, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 1.9472200871124776, | |
| "eval_loss": 0.348527193069458, | |
| "eval_runtime": 301.456, | |
| "eval_samples_per_second": 12.267, | |
| "eval_steps_per_second": 12.267, | |
| "eval_wer": 0.2333979941392546, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 1.9813818430267316, | |
| "grad_norm": 1.8134913444519043, | |
| "learning_rate": 2.5609403881700556e-05, | |
| "loss": 0.3594, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 1.9813818430267316, | |
| "eval_loss": 0.3539000451564789, | |
| "eval_runtime": 299.8125, | |
| "eval_samples_per_second": 12.334, | |
| "eval_steps_per_second": 12.334, | |
| "eval_wer": 0.23412026909901357, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 2.015543598940986, | |
| "grad_norm": 4.158298969268799, | |
| "learning_rate": 2.5176178373382624e-05, | |
| "loss": 0.3692, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 2.015543598940986, | |
| "eval_loss": 0.33834952116012573, | |
| "eval_runtime": 300.4735, | |
| "eval_samples_per_second": 12.307, | |
| "eval_steps_per_second": 12.307, | |
| "eval_wer": 0.22817697800156836, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 2.0497053548552397, | |
| "grad_norm": 1.628749132156372, | |
| "learning_rate": 2.47429528650647e-05, | |
| "loss": 0.3295, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0497053548552397, | |
| "eval_loss": 0.33542123436927795, | |
| "eval_runtime": 299.4927, | |
| "eval_samples_per_second": 12.348, | |
| "eval_steps_per_second": 12.348, | |
| "eval_wer": 0.23740146105906146, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.0838671107694937, | |
| "grad_norm": 1.8399901390075684, | |
| "learning_rate": 2.4309727356746766e-05, | |
| "loss": 0.3442, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 2.0838671107694937, | |
| "eval_loss": 0.33928370475769043, | |
| "eval_runtime": 299.4918, | |
| "eval_samples_per_second": 12.348, | |
| "eval_steps_per_second": 12.348, | |
| "eval_wer": 0.23403772338932685, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 2.1180288666837477, | |
| "grad_norm": 2.220979928970337, | |
| "learning_rate": 2.3876501848428837e-05, | |
| "loss": 0.3306, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 2.1180288666837477, | |
| "eval_loss": 0.35673782229423523, | |
| "eval_runtime": 303.0121, | |
| "eval_samples_per_second": 12.204, | |
| "eval_steps_per_second": 12.204, | |
| "eval_wer": 0.2381650088736638, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 2.1521906225980016, | |
| "grad_norm": 2.257617235183716, | |
| "learning_rate": 2.3443276340110905e-05, | |
| "loss": 0.3243, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 2.1521906225980016, | |
| "eval_loss": 0.3409930169582367, | |
| "eval_runtime": 299.9306, | |
| "eval_samples_per_second": 12.33, | |
| "eval_steps_per_second": 12.33, | |
| "eval_wer": 0.2286722522596888, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 2.1863523785122556, | |
| "grad_norm": 4.057617664337158, | |
| "learning_rate": 2.3010050831792976e-05, | |
| "loss": 0.3426, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 2.1863523785122556, | |
| "eval_loss": 0.3244094252586365, | |
| "eval_runtime": 302.9284, | |
| "eval_samples_per_second": 12.208, | |
| "eval_steps_per_second": 12.208, | |
| "eval_wer": 0.23230426348590533, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 2.2205141344265096, | |
| "grad_norm": 4.836043357849121, | |
| "learning_rate": 2.2576825323475047e-05, | |
| "loss": 0.3552, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.2205141344265096, | |
| "eval_loss": 0.33564862608909607, | |
| "eval_runtime": 300.1908, | |
| "eval_samples_per_second": 12.319, | |
| "eval_steps_per_second": 12.319, | |
| "eval_wer": 0.23176771637294152, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.2546758903407635, | |
| "grad_norm": 3.1353209018707275, | |
| "learning_rate": 2.214359981515712e-05, | |
| "loss": 0.3558, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 2.2546758903407635, | |
| "eval_loss": 0.3686023950576782, | |
| "eval_runtime": 300.6653, | |
| "eval_samples_per_second": 12.299, | |
| "eval_steps_per_second": 12.299, | |
| "eval_wer": 0.22246068760576168, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 2.2888376462550175, | |
| "grad_norm": 1.1471089124679565, | |
| "learning_rate": 2.1710374306839186e-05, | |
| "loss": 0.3485, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 2.2888376462550175, | |
| "eval_loss": 0.34854745864868164, | |
| "eval_runtime": 301.8158, | |
| "eval_samples_per_second": 12.253, | |
| "eval_steps_per_second": 12.253, | |
| "eval_wer": 0.2229765982913038, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 2.3229994021692715, | |
| "grad_norm": 4.745022773742676, | |
| "learning_rate": 2.1277148798521257e-05, | |
| "loss": 0.3195, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 2.3229994021692715, | |
| "eval_loss": 0.31967541575431824, | |
| "eval_runtime": 303.1201, | |
| "eval_samples_per_second": 12.2, | |
| "eval_steps_per_second": 12.2, | |
| "eval_wer": 0.2229765982913038, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 2.3571611580835254, | |
| "grad_norm": 0.8967244625091553, | |
| "learning_rate": 2.084392329020333e-05, | |
| "loss": 0.3145, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 2.3571611580835254, | |
| "eval_loss": 0.33116424083709717, | |
| "eval_runtime": 298.6697, | |
| "eval_samples_per_second": 12.382, | |
| "eval_steps_per_second": 12.382, | |
| "eval_wer": 0.22941516364686945, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 2.3913229139977794, | |
| "grad_norm": 1.37799870967865, | |
| "learning_rate": 2.04106977818854e-05, | |
| "loss": 0.3238, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.3913229139977794, | |
| "eval_loss": 0.3331291675567627, | |
| "eval_runtime": 299.3013, | |
| "eval_samples_per_second": 12.355, | |
| "eval_steps_per_second": 12.355, | |
| "eval_wer": 0.22097486483140039, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.4254846699120334, | |
| "grad_norm": 3.7956511974334717, | |
| "learning_rate": 1.9978916358595196e-05, | |
| "loss": 0.3288, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 2.4254846699120334, | |
| "eval_loss": 0.31718209385871887, | |
| "eval_runtime": 302.9796, | |
| "eval_samples_per_second": 12.205, | |
| "eval_steps_per_second": 12.205, | |
| "eval_wer": 0.2272070659127492, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 2.4596464258262873, | |
| "grad_norm": 2.695967674255371, | |
| "learning_rate": 1.9545690850277264e-05, | |
| "loss": 0.3398, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 2.4596464258262873, | |
| "eval_loss": 0.3227500915527344, | |
| "eval_runtime": 303.1823, | |
| "eval_samples_per_second": 12.197, | |
| "eval_steps_per_second": 12.197, | |
| "eval_wer": 0.21818894712947295, | |
| "step": 21600 | |
| } | |
| ], | |
| "logging_steps": 300, | |
| "max_steps": 35124, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 600, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.982387127199302e+19, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |