| { | |
| "best_metric": 8.634474343167287, | |
| "best_model_checkpoint": "./whisper-small-mix-es/checkpoint-5000", | |
| "epoch": 1.0, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005, | |
| "grad_norm": 8.203697204589844, | |
| "learning_rate": 4.6000000000000004e-07, | |
| "loss": 0.9882, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 4.3468732833862305, | |
| "learning_rate": 9.600000000000001e-07, | |
| "loss": 0.9068, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015, | |
| "grad_norm": 3.7499096393585205, | |
| "learning_rate": 1.46e-06, | |
| "loss": 0.7332, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 3.43481183052063, | |
| "learning_rate": 1.9600000000000003e-06, | |
| "loss": 0.628, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.025, | |
| "grad_norm": 2.902010440826416, | |
| "learning_rate": 2.46e-06, | |
| "loss": 0.4929, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 2.6406397819519043, | |
| "learning_rate": 2.96e-06, | |
| "loss": 0.4353, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.035, | |
| "grad_norm": 2.446483850479126, | |
| "learning_rate": 3.46e-06, | |
| "loss": 0.3881, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 2.8479089736938477, | |
| "learning_rate": 3.96e-06, | |
| "loss": 0.3941, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.045, | |
| "grad_norm": 2.583815813064575, | |
| "learning_rate": 4.4600000000000005e-06, | |
| "loss": 0.431, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.4748358726501465, | |
| "learning_rate": 4.960000000000001e-06, | |
| "loss": 0.4073, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.055, | |
| "grad_norm": 2.5640480518341064, | |
| "learning_rate": 5.460000000000001e-06, | |
| "loss": 0.4266, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 2.221160650253296, | |
| "learning_rate": 5.9600000000000005e-06, | |
| "loss": 0.4025, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.065, | |
| "grad_norm": 2.4895880222320557, | |
| "learning_rate": 6.460000000000001e-06, | |
| "loss": 0.3504, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 2.318375825881958, | |
| "learning_rate": 6.96e-06, | |
| "loss": 0.311, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.075, | |
| "grad_norm": 3.063900947570801, | |
| "learning_rate": 7.4600000000000006e-06, | |
| "loss": 0.3309, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.898184061050415, | |
| "learning_rate": 7.960000000000002e-06, | |
| "loss": 0.3044, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.085, | |
| "grad_norm": 2.2967689037323, | |
| "learning_rate": 8.46e-06, | |
| "loss": 0.3071, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 2.990691900253296, | |
| "learning_rate": 8.96e-06, | |
| "loss": 0.3319, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.095, | |
| "grad_norm": 2.604433536529541, | |
| "learning_rate": 9.460000000000001e-06, | |
| "loss": 0.3606, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.8391056060791016, | |
| "learning_rate": 9.960000000000001e-06, | |
| "loss": 0.3517, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.105, | |
| "grad_norm": 2.3469793796539307, | |
| "learning_rate": 9.94888888888889e-06, | |
| "loss": 0.3338, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 1.7184339761734009, | |
| "learning_rate": 9.893333333333334e-06, | |
| "loss": 0.3195, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.115, | |
| "grad_norm": 2.248863697052002, | |
| "learning_rate": 9.837777777777778e-06, | |
| "loss": 0.2926, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.592674970626831, | |
| "learning_rate": 9.782222222222222e-06, | |
| "loss": 0.3141, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.125, | |
| "grad_norm": 1.9537230730056763, | |
| "learning_rate": 9.726666666666668e-06, | |
| "loss": 0.2778, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 2.4650800228118896, | |
| "learning_rate": 9.671111111111112e-06, | |
| "loss": 0.2952, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.135, | |
| "grad_norm": 2.378822088241577, | |
| "learning_rate": 9.615555555555558e-06, | |
| "loss": 0.3339, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.6371264457702637, | |
| "learning_rate": 9.56e-06, | |
| "loss": 0.3843, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.145, | |
| "grad_norm": 2.7936806678771973, | |
| "learning_rate": 9.504444444444446e-06, | |
| "loss": 0.3703, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 2.70261287689209, | |
| "learning_rate": 9.44888888888889e-06, | |
| "loss": 0.359, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.155, | |
| "grad_norm": 2.0972862243652344, | |
| "learning_rate": 9.393333333333334e-06, | |
| "loss": 0.2747, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 2.8365142345428467, | |
| "learning_rate": 9.33777777777778e-06, | |
| "loss": 0.2849, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.165, | |
| "grad_norm": 2.2393555641174316, | |
| "learning_rate": 9.282222222222222e-06, | |
| "loss": 0.2273, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 1.7303996086120605, | |
| "learning_rate": 9.226666666666668e-06, | |
| "loss": 0.2102, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.175, | |
| "grad_norm": 2.153916597366333, | |
| "learning_rate": 9.171111111111112e-06, | |
| "loss": 0.3088, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.3928215503692627, | |
| "learning_rate": 9.115555555555556e-06, | |
| "loss": 0.2912, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.185, | |
| "grad_norm": 2.6662864685058594, | |
| "learning_rate": 9.060000000000001e-06, | |
| "loss": 0.2926, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 2.274172067642212, | |
| "learning_rate": 9.004444444444445e-06, | |
| "loss": 0.2777, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.195, | |
| "grad_norm": 2.453396797180176, | |
| "learning_rate": 8.94888888888889e-06, | |
| "loss": 0.2869, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 2.171252489089966, | |
| "learning_rate": 8.893333333333333e-06, | |
| "loss": 0.247, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "eval_loss": 0.21598759293556213, | |
| "eval_runtime": 708.6287, | |
| "eval_samples_per_second": 22.377, | |
| "eval_steps_per_second": 2.798, | |
| "eval_wer": 10.397468000457591, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.205, | |
| "grad_norm": 2.494718313217163, | |
| "learning_rate": 8.83777777777778e-06, | |
| "loss": 0.2519, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.5458056926727295, | |
| "learning_rate": 8.782222222222223e-06, | |
| "loss": 0.2523, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.215, | |
| "grad_norm": 1.856636643409729, | |
| "learning_rate": 8.726666666666667e-06, | |
| "loss": 0.2481, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 2.324716091156006, | |
| "learning_rate": 8.671111111111113e-06, | |
| "loss": 0.2453, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.225, | |
| "grad_norm": 1.7981597185134888, | |
| "learning_rate": 8.615555555555555e-06, | |
| "loss": 0.2297, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 2.4602575302124023, | |
| "learning_rate": 8.560000000000001e-06, | |
| "loss": 0.2492, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.235, | |
| "grad_norm": 2.964137077331543, | |
| "learning_rate": 8.504444444444445e-06, | |
| "loss": 0.2793, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 2.031019687652588, | |
| "learning_rate": 8.448888888888889e-06, | |
| "loss": 0.3005, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.245, | |
| "grad_norm": 2.0715112686157227, | |
| "learning_rate": 8.393333333333335e-06, | |
| "loss": 0.2397, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 2.2880711555480957, | |
| "learning_rate": 8.337777777777777e-06, | |
| "loss": 0.251, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.255, | |
| "grad_norm": 2.0177316665649414, | |
| "learning_rate": 8.282222222222223e-06, | |
| "loss": 0.2234, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.3861050605773926, | |
| "learning_rate": 8.226666666666667e-06, | |
| "loss": 0.2855, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.265, | |
| "grad_norm": 2.8440380096435547, | |
| "learning_rate": 8.171111111111113e-06, | |
| "loss": 0.2608, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 2.476083993911743, | |
| "learning_rate": 8.115555555555557e-06, | |
| "loss": 0.255, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.275, | |
| "grad_norm": 2.3754682540893555, | |
| "learning_rate": 8.06e-06, | |
| "loss": 0.2479, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.507122755050659, | |
| "learning_rate": 8.004444444444445e-06, | |
| "loss": 0.2426, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.285, | |
| "grad_norm": 2.9778430461883545, | |
| "learning_rate": 7.948888888888889e-06, | |
| "loss": 0.2644, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 2.6550915241241455, | |
| "learning_rate": 7.893333333333335e-06, | |
| "loss": 0.2544, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.295, | |
| "grad_norm": 2.50137996673584, | |
| "learning_rate": 7.837777777777779e-06, | |
| "loss": 0.3037, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.3021984100341797, | |
| "learning_rate": 7.782222222222223e-06, | |
| "loss": 0.2831, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.305, | |
| "grad_norm": 2.2827396392822266, | |
| "learning_rate": 7.726666666666667e-06, | |
| "loss": 0.2387, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.9767779111862183, | |
| "learning_rate": 7.67111111111111e-06, | |
| "loss": 0.2293, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.315, | |
| "grad_norm": 2.132072687149048, | |
| "learning_rate": 7.6155555555555564e-06, | |
| "loss": 0.1855, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.581895112991333, | |
| "learning_rate": 7.5600000000000005e-06, | |
| "loss": 0.208, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.325, | |
| "grad_norm": 2.436007261276245, | |
| "learning_rate": 7.504444444444445e-06, | |
| "loss": 0.2112, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 2.5022451877593994, | |
| "learning_rate": 7.44888888888889e-06, | |
| "loss": 0.2001, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.335, | |
| "grad_norm": 2.7856762409210205, | |
| "learning_rate": 7.393333333333333e-06, | |
| "loss": 0.2304, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 3.307318925857544, | |
| "learning_rate": 7.337777777777778e-06, | |
| "loss": 0.2489, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.345, | |
| "grad_norm": 2.104020118713379, | |
| "learning_rate": 7.282222222222222e-06, | |
| "loss": 0.204, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 2.099266529083252, | |
| "learning_rate": 7.226666666666667e-06, | |
| "loss": 0.1787, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.355, | |
| "grad_norm": 2.0971200466156006, | |
| "learning_rate": 7.171111111111112e-06, | |
| "loss": 0.1672, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 1.7668758630752563, | |
| "learning_rate": 7.115555555555557e-06, | |
| "loss": 0.1537, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.365, | |
| "grad_norm": 2.1447267532348633, | |
| "learning_rate": 7.06e-06, | |
| "loss": 0.1568, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 2.092658758163452, | |
| "learning_rate": 7.004444444444445e-06, | |
| "loss": 0.1509, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.375, | |
| "grad_norm": 2.127138614654541, | |
| "learning_rate": 6.948888888888889e-06, | |
| "loss": 0.1447, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.7736611366271973, | |
| "learning_rate": 6.893333333333334e-06, | |
| "loss": 0.1479, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.385, | |
| "grad_norm": 1.845276951789856, | |
| "learning_rate": 6.837777777777779e-06, | |
| "loss": 0.1402, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 2.130812883377075, | |
| "learning_rate": 6.782222222222222e-06, | |
| "loss": 0.1394, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.395, | |
| "grad_norm": 1.9191032648086548, | |
| "learning_rate": 6.726666666666667e-06, | |
| "loss": 0.1491, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.805740237236023, | |
| "learning_rate": 6.671111111111112e-06, | |
| "loss": 0.1337, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.20104137063026428, | |
| "eval_runtime": 702.6623, | |
| "eval_samples_per_second": 22.567, | |
| "eval_steps_per_second": 2.822, | |
| "eval_wer": 9.674856685266864, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.405, | |
| "grad_norm": 2.0551412105560303, | |
| "learning_rate": 6.615555555555556e-06, | |
| "loss": 0.1269, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 2.093939781188965, | |
| "learning_rate": 6.560000000000001e-06, | |
| "loss": 0.1412, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.415, | |
| "grad_norm": 2.695679187774658, | |
| "learning_rate": 6.504444444444446e-06, | |
| "loss": 0.1537, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 2.2882978916168213, | |
| "learning_rate": 6.448888888888889e-06, | |
| "loss": 0.1669, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.425, | |
| "grad_norm": 1.3621838092803955, | |
| "learning_rate": 6.393333333333334e-06, | |
| "loss": 0.1607, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 2.160801410675049, | |
| "learning_rate": 6.3377777777777786e-06, | |
| "loss": 0.1482, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.435, | |
| "grad_norm": 2.119140863418579, | |
| "learning_rate": 6.282222222222223e-06, | |
| "loss": 0.1552, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 1.6484750509262085, | |
| "learning_rate": 6.2266666666666675e-06, | |
| "loss": 0.1592, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.445, | |
| "grad_norm": 1.9934930801391602, | |
| "learning_rate": 6.171111111111112e-06, | |
| "loss": 0.1527, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 1.8050808906555176, | |
| "learning_rate": 6.1155555555555555e-06, | |
| "loss": 0.1487, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.455, | |
| "grad_norm": 2.1619839668273926, | |
| "learning_rate": 6.0600000000000004e-06, | |
| "loss": 0.1462, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 2.169475555419922, | |
| "learning_rate": 6.004444444444445e-06, | |
| "loss": 0.1566, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.465, | |
| "grad_norm": 2.3247809410095215, | |
| "learning_rate": 5.948888888888889e-06, | |
| "loss": 0.1527, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.519680380821228, | |
| "learning_rate": 5.893333333333334e-06, | |
| "loss": 0.1368, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.475, | |
| "grad_norm": 1.9055320024490356, | |
| "learning_rate": 5.837777777777777e-06, | |
| "loss": 0.13, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.674070954322815, | |
| "learning_rate": 5.782222222222222e-06, | |
| "loss": 0.1236, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.485, | |
| "grad_norm": 2.0586557388305664, | |
| "learning_rate": 5.726666666666667e-06, | |
| "loss": 0.1353, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 2.190614700317383, | |
| "learning_rate": 5.671111111111112e-06, | |
| "loss": 0.1294, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.495, | |
| "grad_norm": 1.7952390909194946, | |
| "learning_rate": 5.615555555555556e-06, | |
| "loss": 0.1371, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 1.9759557247161865, | |
| "learning_rate": 5.560000000000001e-06, | |
| "loss": 0.1525, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.505, | |
| "grad_norm": 1.5587725639343262, | |
| "learning_rate": 5.504444444444444e-06, | |
| "loss": 0.1334, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 1.9309734106063843, | |
| "learning_rate": 5.448888888888889e-06, | |
| "loss": 0.1277, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.515, | |
| "grad_norm": 1.6689786911010742, | |
| "learning_rate": 5.393333333333334e-06, | |
| "loss": 0.1312, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 1.5617754459381104, | |
| "learning_rate": 5.337777777777779e-06, | |
| "loss": 0.1421, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.525, | |
| "grad_norm": 2.1239089965820312, | |
| "learning_rate": 5.282222222222223e-06, | |
| "loss": 0.1338, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 1.6004153490066528, | |
| "learning_rate": 5.226666666666667e-06, | |
| "loss": 0.1363, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.535, | |
| "grad_norm": 1.5653680562973022, | |
| "learning_rate": 5.171111111111111e-06, | |
| "loss": 0.1309, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 1.5964562892913818, | |
| "learning_rate": 5.115555555555556e-06, | |
| "loss": 0.133, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.545, | |
| "grad_norm": 1.9546246528625488, | |
| "learning_rate": 5.060000000000001e-06, | |
| "loss": 0.1281, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 1.6310886144638062, | |
| "learning_rate": 5.004444444444445e-06, | |
| "loss": 0.1361, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.555, | |
| "grad_norm": 1.7071685791015625, | |
| "learning_rate": 4.94888888888889e-06, | |
| "loss": 0.1382, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.6160541772842407, | |
| "learning_rate": 4.893333333333334e-06, | |
| "loss": 0.1394, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.565, | |
| "grad_norm": 1.4931234121322632, | |
| "learning_rate": 4.837777777777778e-06, | |
| "loss": 0.1221, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 1.4717397689819336, | |
| "learning_rate": 4.7822222222222226e-06, | |
| "loss": 0.1106, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.575, | |
| "grad_norm": 1.334092140197754, | |
| "learning_rate": 4.7266666666666674e-06, | |
| "loss": 0.1299, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.9642376899719238, | |
| "learning_rate": 4.6711111111111115e-06, | |
| "loss": 0.1509, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.585, | |
| "grad_norm": 1.5685168504714966, | |
| "learning_rate": 4.6155555555555555e-06, | |
| "loss": 0.1367, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 1.2653120756149292, | |
| "learning_rate": 4.56e-06, | |
| "loss": 0.1228, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.595, | |
| "grad_norm": 1.6056393384933472, | |
| "learning_rate": 4.504444444444444e-06, | |
| "loss": 0.1081, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 2.4478392601013184, | |
| "learning_rate": 4.448888888888889e-06, | |
| "loss": 0.1401, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "eval_loss": 0.19045059382915497, | |
| "eval_runtime": 711.5811, | |
| "eval_samples_per_second": 22.284, | |
| "eval_steps_per_second": 2.787, | |
| "eval_wer": 9.094606790131303, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.605, | |
| "grad_norm": 1.8749899864196777, | |
| "learning_rate": 4.393333333333334e-06, | |
| "loss": 0.1864, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 1.5327539443969727, | |
| "learning_rate": 4.337777777777778e-06, | |
| "loss": 0.184, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.615, | |
| "grad_norm": 2.0878913402557373, | |
| "learning_rate": 4.282222222222222e-06, | |
| "loss": 0.1969, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 2.027475118637085, | |
| "learning_rate": 4.226666666666667e-06, | |
| "loss": 0.1979, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.625, | |
| "grad_norm": 1.4763484001159668, | |
| "learning_rate": 4.171111111111111e-06, | |
| "loss": 0.1807, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 2.0090372562408447, | |
| "learning_rate": 4.115555555555556e-06, | |
| "loss": 0.1498, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.635, | |
| "grad_norm": 1.4188919067382812, | |
| "learning_rate": 4.060000000000001e-06, | |
| "loss": 0.1308, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.6923434734344482, | |
| "learning_rate": 4.004444444444445e-06, | |
| "loss": 0.1422, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.645, | |
| "grad_norm": 1.8720096349716187, | |
| "learning_rate": 3.948888888888889e-06, | |
| "loss": 0.133, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 1.367676854133606, | |
| "learning_rate": 3.893333333333333e-06, | |
| "loss": 0.1265, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.655, | |
| "grad_norm": 1.9834328889846802, | |
| "learning_rate": 3.837777777777778e-06, | |
| "loss": 0.1323, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 2.014779567718506, | |
| "learning_rate": 3.782222222222223e-06, | |
| "loss": 0.141, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.665, | |
| "grad_norm": 1.3820104598999023, | |
| "learning_rate": 3.726666666666667e-06, | |
| "loss": 0.1308, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 1.9776164293289185, | |
| "learning_rate": 3.6711111111111113e-06, | |
| "loss": 0.1331, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.675, | |
| "grad_norm": 1.625608205795288, | |
| "learning_rate": 3.615555555555556e-06, | |
| "loss": 0.116, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 1.7291717529296875, | |
| "learning_rate": 3.5600000000000002e-06, | |
| "loss": 0.1068, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.685, | |
| "grad_norm": 1.7692865133285522, | |
| "learning_rate": 3.5044444444444447e-06, | |
| "loss": 0.1134, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.9377959966659546, | |
| "learning_rate": 3.4488888888888896e-06, | |
| "loss": 0.1939, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.695, | |
| "grad_norm": 2.1209535598754883, | |
| "learning_rate": 3.3933333333333336e-06, | |
| "loss": 0.2623, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.0836637020111084, | |
| "learning_rate": 3.337777777777778e-06, | |
| "loss": 0.2272, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.705, | |
| "grad_norm": 1.8429718017578125, | |
| "learning_rate": 3.282222222222223e-06, | |
| "loss": 0.2041, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 1.7450307607650757, | |
| "learning_rate": 3.226666666666667e-06, | |
| "loss": 0.224, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.715, | |
| "grad_norm": 2.0255823135375977, | |
| "learning_rate": 3.1711111111111114e-06, | |
| "loss": 0.2067, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 2.1434662342071533, | |
| "learning_rate": 3.1155555555555555e-06, | |
| "loss": 0.182, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.725, | |
| "grad_norm": 1.3078871965408325, | |
| "learning_rate": 3.0600000000000003e-06, | |
| "loss": 0.1625, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 1.7805925607681274, | |
| "learning_rate": 3.004444444444445e-06, | |
| "loss": 0.1701, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.735, | |
| "grad_norm": 1.5756222009658813, | |
| "learning_rate": 2.948888888888889e-06, | |
| "loss": 0.1473, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 1.343863844871521, | |
| "learning_rate": 2.8933333333333337e-06, | |
| "loss": 0.1385, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.745, | |
| "grad_norm": 1.7856868505477905, | |
| "learning_rate": 2.837777777777778e-06, | |
| "loss": 0.156, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.6348956823349, | |
| "learning_rate": 2.7822222222222222e-06, | |
| "loss": 0.1477, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.755, | |
| "grad_norm": 1.698569655418396, | |
| "learning_rate": 2.726666666666667e-06, | |
| "loss": 0.1488, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 2.1124930381774902, | |
| "learning_rate": 2.6711111111111116e-06, | |
| "loss": 0.1365, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.765, | |
| "grad_norm": 1.5699976682662964, | |
| "learning_rate": 2.6155555555555556e-06, | |
| "loss": 0.1438, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 1.302807092666626, | |
| "learning_rate": 2.56e-06, | |
| "loss": 0.1238, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.775, | |
| "grad_norm": 1.5388458967208862, | |
| "learning_rate": 2.504444444444445e-06, | |
| "loss": 0.1264, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 1.3941725492477417, | |
| "learning_rate": 2.448888888888889e-06, | |
| "loss": 0.1182, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.785, | |
| "grad_norm": 1.8302580118179321, | |
| "learning_rate": 2.3933333333333334e-06, | |
| "loss": 0.1201, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 1.6596505641937256, | |
| "learning_rate": 2.337777777777778e-06, | |
| "loss": 0.114, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.795, | |
| "grad_norm": 1.5720983743667603, | |
| "learning_rate": 2.2822222222222223e-06, | |
| "loss": 0.1149, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.3143093585968018, | |
| "learning_rate": 2.226666666666667e-06, | |
| "loss": 0.1714, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.18485930562019348, | |
| "eval_runtime": 702.3906, | |
| "eval_samples_per_second": 22.576, | |
| "eval_steps_per_second": 2.823, | |
| "eval_wer": 8.85500743584203, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.805, | |
| "grad_norm": 1.8013464212417603, | |
| "learning_rate": 2.1711111111111113e-06, | |
| "loss": 0.1782, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 1.6786607503890991, | |
| "learning_rate": 2.1155555555555557e-06, | |
| "loss": 0.1262, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.815, | |
| "grad_norm": 1.412192702293396, | |
| "learning_rate": 2.06e-06, | |
| "loss": 0.1257, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 1.5031375885009766, | |
| "learning_rate": 2.0044444444444446e-06, | |
| "loss": 0.1164, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.825, | |
| "grad_norm": 1.3566547632217407, | |
| "learning_rate": 1.948888888888889e-06, | |
| "loss": 0.1238, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.5606483221054077, | |
| "learning_rate": 1.8933333333333333e-06, | |
| "loss": 0.1029, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.835, | |
| "grad_norm": 2.154038429260254, | |
| "learning_rate": 1.837777777777778e-06, | |
| "loss": 0.1045, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 1.4622442722320557, | |
| "learning_rate": 1.7822222222222225e-06, | |
| "loss": 0.1091, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.845, | |
| "grad_norm": 1.7122653722763062, | |
| "learning_rate": 1.7266666666666667e-06, | |
| "loss": 0.1158, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 2.029569625854492, | |
| "learning_rate": 1.6711111111111112e-06, | |
| "loss": 0.1267, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.855, | |
| "grad_norm": 1.5599546432495117, | |
| "learning_rate": 1.6155555555555559e-06, | |
| "loss": 0.1084, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.735775351524353, | |
| "learning_rate": 1.56e-06, | |
| "loss": 0.1028, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.865, | |
| "grad_norm": 1.6177542209625244, | |
| "learning_rate": 1.5044444444444446e-06, | |
| "loss": 0.1174, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 1.7929385900497437, | |
| "learning_rate": 1.4488888888888892e-06, | |
| "loss": 0.1193, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.875, | |
| "grad_norm": 1.6142146587371826, | |
| "learning_rate": 1.3933333333333335e-06, | |
| "loss": 0.1243, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.6547541618347168, | |
| "learning_rate": 1.337777777777778e-06, | |
| "loss": 0.1413, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.885, | |
| "grad_norm": 1.3043129444122314, | |
| "learning_rate": 1.2822222222222222e-06, | |
| "loss": 0.1394, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.324398398399353, | |
| "learning_rate": 1.2266666666666666e-06, | |
| "loss": 0.1262, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.895, | |
| "grad_norm": 1.5393497943878174, | |
| "learning_rate": 1.171111111111111e-06, | |
| "loss": 0.0991, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 1.1875627040863037, | |
| "learning_rate": 1.1155555555555558e-06, | |
| "loss": 0.0824, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.905, | |
| "grad_norm": 1.2755403518676758, | |
| "learning_rate": 1.06e-06, | |
| "loss": 0.0862, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 1.2878875732421875, | |
| "learning_rate": 1.0044444444444445e-06, | |
| "loss": 0.0856, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.915, | |
| "grad_norm": 1.5324125289916992, | |
| "learning_rate": 9.488888888888889e-07, | |
| "loss": 0.0896, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 1.5242027044296265, | |
| "learning_rate": 8.933333333333334e-07, | |
| "loss": 0.0996, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.925, | |
| "grad_norm": 1.3442718982696533, | |
| "learning_rate": 8.37777777777778e-07, | |
| "loss": 0.0914, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 1.601271629333496, | |
| "learning_rate": 7.822222222222223e-07, | |
| "loss": 0.0891, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.935, | |
| "grad_norm": 1.5180890560150146, | |
| "learning_rate": 7.266666666666668e-07, | |
| "loss": 0.0912, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.3631670475006104, | |
| "learning_rate": 6.711111111111111e-07, | |
| "loss": 0.0952, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.945, | |
| "grad_norm": 1.251734733581543, | |
| "learning_rate": 6.155555555555556e-07, | |
| "loss": 0.0924, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 1.3705124855041504, | |
| "learning_rate": 5.6e-07, | |
| "loss": 0.085, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.955, | |
| "grad_norm": 1.46409273147583, | |
| "learning_rate": 5.044444444444445e-07, | |
| "loss": 0.0955, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.6144453287124634, | |
| "learning_rate": 4.488888888888889e-07, | |
| "loss": 0.0852, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.965, | |
| "grad_norm": 1.9570475816726685, | |
| "learning_rate": 3.9333333333333336e-07, | |
| "loss": 0.0979, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 1.666892409324646, | |
| "learning_rate": 3.3777777777777777e-07, | |
| "loss": 0.0954, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.975, | |
| "grad_norm": 1.4862993955612183, | |
| "learning_rate": 2.822222222222222e-07, | |
| "loss": 0.0869, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 1.2195048332214355, | |
| "learning_rate": 2.266666666666667e-07, | |
| "loss": 0.094, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.985, | |
| "grad_norm": 1.4709336757659912, | |
| "learning_rate": 1.7111111111111114e-07, | |
| "loss": 0.0929, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 1.3563309907913208, | |
| "learning_rate": 1.1555555555555556e-07, | |
| "loss": 0.0915, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.995, | |
| "grad_norm": 1.3233599662780762, | |
| "learning_rate": 6.000000000000001e-08, | |
| "loss": 0.0937, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.5326294898986816, | |
| "learning_rate": 4.444444444444445e-09, | |
| "loss": 0.1046, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.18091823160648346, | |
| "eval_runtime": 696.7327, | |
| "eval_samples_per_second": 22.759, | |
| "eval_steps_per_second": 2.846, | |
| "eval_wer": 8.634474343167287, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 5000, | |
| "total_flos": 9.23473281024e+19, | |
| "train_loss": 0.20034765486717224, | |
| "train_runtime": 18318.4818, | |
| "train_samples_per_second": 17.469, | |
| "train_steps_per_second": 0.273 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9223372036854775807, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.23473281024e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |