| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 35.08849557522124, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4424778761061947, | |
| "grad_norm": 70.23552703857422, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 1.0554, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.8849557522123894, | |
| "grad_norm": 19.844127655029297, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.7061, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.3185840707964602, | |
| "grad_norm": 7.727048397064209, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.3636, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.7610619469026547, | |
| "grad_norm": 4.834627628326416, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.1976, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.1946902654867255, | |
| "grad_norm": 4.151981830596924, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1237, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.6371681415929205, | |
| "grad_norm": 3.742462635040283, | |
| "learning_rate": 3e-06, | |
| "loss": 0.1117, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.0707964601769913, | |
| "grad_norm": 4.222021102905273, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.0927, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.5132743362831858, | |
| "grad_norm": 3.9585046768188477, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.067, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.9557522123893807, | |
| "grad_norm": 2.4528427124023438, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.0605, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.389380530973451, | |
| "grad_norm": 3.324159622192383, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0598, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.831858407079646, | |
| "grad_norm": 3.3953816890716553, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.0543, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.265486725663717, | |
| "grad_norm": 4.006282329559326, | |
| "learning_rate": 6e-06, | |
| "loss": 0.0372, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.707964601769912, | |
| "grad_norm": 0.8992710113525391, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.0382, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.1415929203539825, | |
| "grad_norm": 1.6717145442962646, | |
| "learning_rate": 7e-06, | |
| "loss": 0.0379, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 6.584070796460177, | |
| "grad_norm": 1.683933973312378, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0354, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.017699115044247, | |
| "grad_norm": 4.067246437072754, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.0267, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.460176991150442, | |
| "grad_norm": 0.14960134029388428, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.0215, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 7.902654867256637, | |
| "grad_norm": 2.5055603981018066, | |
| "learning_rate": 9e-06, | |
| "loss": 0.0294, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 8.336283185840708, | |
| "grad_norm": 2.1897315979003906, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.0202, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 8.778761061946902, | |
| "grad_norm": 2.6399195194244385, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0195, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 8.778761061946902, | |
| "eval_loss": 3.6442341804504395, | |
| "eval_runtime": 98.5424, | |
| "eval_samples_per_second": 1.005, | |
| "eval_steps_per_second": 0.071, | |
| "eval_wer": 1.2195824334053276, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 9.212389380530974, | |
| "grad_norm": 1.9207507371902466, | |
| "learning_rate": 9.833333333333333e-06, | |
| "loss": 0.0136, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 9.654867256637168, | |
| "grad_norm": 1.3496884107589722, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 0.0184, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 10.08849557522124, | |
| "grad_norm": 1.9232096672058105, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.0194, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 10.530973451327434, | |
| "grad_norm": 2.705425977706909, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 0.0139, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 10.973451327433628, | |
| "grad_norm": 0.18757027387619019, | |
| "learning_rate": 9.166666666666666e-06, | |
| "loss": 0.0075, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 11.4070796460177, | |
| "grad_norm": 2.2813608646392822, | |
| "learning_rate": 9e-06, | |
| "loss": 0.0088, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 11.849557522123893, | |
| "grad_norm": 0.2083648443222046, | |
| "learning_rate": 8.833333333333334e-06, | |
| "loss": 0.0108, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 12.283185840707965, | |
| "grad_norm": 3.0644192695617676, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 0.0088, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 12.725663716814159, | |
| "grad_norm": 0.33310991525650024, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.007, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 13.15929203539823, | |
| "grad_norm": 0.6558440923690796, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 0.0071, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 13.601769911504425, | |
| "grad_norm": 0.04615064710378647, | |
| "learning_rate": 8.166666666666668e-06, | |
| "loss": 0.0066, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 14.035398230088495, | |
| "grad_norm": 2.2992634773254395, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.0051, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 14.47787610619469, | |
| "grad_norm": 1.2088443040847778, | |
| "learning_rate": 7.833333333333333e-06, | |
| "loss": 0.0035, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 14.920353982300885, | |
| "grad_norm": 1.7774810791015625, | |
| "learning_rate": 7.666666666666667e-06, | |
| "loss": 0.0059, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 15.353982300884956, | |
| "grad_norm": 0.07690909504890442, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0047, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 15.79646017699115, | |
| "grad_norm": 1.2860054969787598, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 0.0053, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 16.23008849557522, | |
| "grad_norm": 0.6857427358627319, | |
| "learning_rate": 7.166666666666667e-06, | |
| "loss": 0.0023, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 16.672566371681416, | |
| "grad_norm": 0.7444378733634949, | |
| "learning_rate": 7e-06, | |
| "loss": 0.0043, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 17.106194690265486, | |
| "grad_norm": 0.0285570677369833, | |
| "learning_rate": 6.833333333333334e-06, | |
| "loss": 0.0033, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "grad_norm": 1.755213737487793, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.0044, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.548672566371682, | |
| "eval_loss": 4.271320343017578, | |
| "eval_runtime": 87.4509, | |
| "eval_samples_per_second": 1.132, | |
| "eval_steps_per_second": 0.08, | |
| "eval_wer": 1.2203023758099352, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 17.991150442477878, | |
| "grad_norm": 3.525615930557251, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.004, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 18.424778761061948, | |
| "grad_norm": 1.9330180883407593, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 0.0026, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 18.86725663716814, | |
| "grad_norm": 0.3188052475452423, | |
| "learning_rate": 6.166666666666667e-06, | |
| "loss": 0.0046, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 19.300884955752213, | |
| "grad_norm": 1.5139490365982056, | |
| "learning_rate": 6e-06, | |
| "loss": 0.0042, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 19.743362831858406, | |
| "grad_norm": 5.1269354820251465, | |
| "learning_rate": 5.833333333333334e-06, | |
| "loss": 0.0032, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 20.17699115044248, | |
| "grad_norm": 0.024498550221323967, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 0.0029, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 20.61946902654867, | |
| "grad_norm": 0.017963914200663567, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.0011, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 21.053097345132745, | |
| "grad_norm": 0.10410400480031967, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.0008, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 21.495575221238937, | |
| "grad_norm": 0.05556880682706833, | |
| "learning_rate": 5.1666666666666675e-06, | |
| "loss": 0.0003, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 21.938053097345133, | |
| "grad_norm": 0.005813132505863905, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0004, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 22.371681415929203, | |
| "grad_norm": 0.008638879284262657, | |
| "learning_rate": 4.833333333333333e-06, | |
| "loss": 0.0003, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 22.8141592920354, | |
| "grad_norm": 0.033254146575927734, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 0.0003, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 23.24778761061947, | |
| "grad_norm": 0.018534550443291664, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.0002, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 23.690265486725664, | |
| "grad_norm": 0.00659082131460309, | |
| "learning_rate": 4.333333333333334e-06, | |
| "loss": 0.0003, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 24.123893805309734, | |
| "grad_norm": 0.009523949585855007, | |
| "learning_rate": 4.166666666666667e-06, | |
| "loss": 0.0002, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 24.56637168141593, | |
| "grad_norm": 0.028162825852632523, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.0002, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.012596463784575462, | |
| "learning_rate": 3.833333333333334e-06, | |
| "loss": 0.0001, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 25.442477876106196, | |
| "grad_norm": 0.006055805366486311, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 0.0002, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 25.884955752212388, | |
| "grad_norm": 0.01260603778064251, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.0002, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 26.31858407079646, | |
| "grad_norm": 0.02524421364068985, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.0002, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 26.31858407079646, | |
| "eval_loss": 4.261180877685547, | |
| "eval_runtime": 92.5387, | |
| "eval_samples_per_second": 1.07, | |
| "eval_steps_per_second": 0.076, | |
| "eval_wer": 1.2203023758099352, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 26.761061946902654, | |
| "grad_norm": 0.04416137561202049, | |
| "learning_rate": 3.1666666666666667e-06, | |
| "loss": 0.0002, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 27.194690265486727, | |
| "grad_norm": 0.002192295156419277, | |
| "learning_rate": 3e-06, | |
| "loss": 0.0002, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 27.63716814159292, | |
| "grad_norm": 0.005449674092233181, | |
| "learning_rate": 2.8333333333333335e-06, | |
| "loss": 0.0002, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 28.07079646017699, | |
| "grad_norm": 0.02437576837837696, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 0.0002, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 28.513274336283185, | |
| "grad_norm": 0.042830660939216614, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0002, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 28.95575221238938, | |
| "grad_norm": 0.0039032117929309607, | |
| "learning_rate": 2.3333333333333336e-06, | |
| "loss": 0.0001, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 29.38938053097345, | |
| "grad_norm": 0.008059768006205559, | |
| "learning_rate": 2.166666666666667e-06, | |
| "loss": 0.0002, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 29.831858407079647, | |
| "grad_norm": 0.012063885107636452, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.0002, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 30.265486725663717, | |
| "grad_norm": 0.04550086334347725, | |
| "learning_rate": 1.8333333333333333e-06, | |
| "loss": 0.0001, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 30.707964601769913, | |
| "grad_norm": 0.0046273404732346535, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 0.0001, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 31.141592920353983, | |
| "grad_norm": 0.00888855941593647, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.0002, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 31.58407079646018, | |
| "grad_norm": 0.011799165047705173, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 0.0002, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 32.017699115044245, | |
| "grad_norm": 0.033783987164497375, | |
| "learning_rate": 1.1666666666666668e-06, | |
| "loss": 0.0001, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 32.46017699115044, | |
| "grad_norm": 0.002975311130285263, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.0001, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 32.902654867256636, | |
| "grad_norm": 0.009084297344088554, | |
| "learning_rate": 8.333333333333333e-07, | |
| "loss": 0.0002, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 33.336283185840706, | |
| "grad_norm": 0.011445912532508373, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 0.0002, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 33.7787610619469, | |
| "grad_norm": 0.025164591148495674, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.0001, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 34.21238938053097, | |
| "grad_norm": 0.00487129669636488, | |
| "learning_rate": 3.3333333333333335e-07, | |
| "loss": 0.0001, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 34.65486725663717, | |
| "grad_norm": 0.014068867079913616, | |
| "learning_rate": 1.6666666666666668e-07, | |
| "loss": 0.0002, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 35.08849557522124, | |
| "grad_norm": 0.009848229587078094, | |
| "learning_rate": 0.0, | |
| "loss": 0.0001, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 35.08849557522124, | |
| "eval_loss": 4.253291606903076, | |
| "eval_runtime": 97.0898, | |
| "eval_samples_per_second": 1.02, | |
| "eval_steps_per_second": 0.072, | |
| "eval_wer": 1.2203023758099352, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 35.08849557522124, | |
| "step": 2000, | |
| "total_flos": 1.83078577963008e+19, | |
| "train_loss": 0.0418594888363732, | |
| "train_runtime": 32879.5132, | |
| "train_samples_per_second": 1.947, | |
| "train_steps_per_second": 0.061 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 36, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.83078577963008e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |