| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.6690790241624835, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 38.9212532043457, | |
| "learning_rate": 4.6e-09, | |
| "loss": 1.2718, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 38.4885368347168, | |
| "learning_rate": 9.599999999999998e-09, | |
| "loss": 1.2556, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 41.782493591308594, | |
| "learning_rate": 1.46e-08, | |
| "loss": 1.2688, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 37.6844367980957, | |
| "learning_rate": 1.9599999999999997e-08, | |
| "loss": 1.2761, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 33.974815368652344, | |
| "learning_rate": 2.46e-08, | |
| "loss": 1.2864, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 35.566436767578125, | |
| "learning_rate": 2.96e-08, | |
| "loss": 1.2951, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 36.60067367553711, | |
| "learning_rate": 3.46e-08, | |
| "loss": 1.3226, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 34.72969055175781, | |
| "learning_rate": 3.9600000000000004e-08, | |
| "loss": 1.2698, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 34.141136169433594, | |
| "learning_rate": 4.4599999999999996e-08, | |
| "loss": 1.2497, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 32.15596389770508, | |
| "learning_rate": 4.9599999999999994e-08, | |
| "loss": 1.2155, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 31.284732818603516, | |
| "learning_rate": 5.46e-08, | |
| "loss": 1.2187, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 29.053382873535156, | |
| "learning_rate": 5.96e-08, | |
| "loss": 1.1431, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 28.53986167907715, | |
| "learning_rate": 6.46e-08, | |
| "loss": 1.1342, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 26.937896728515625, | |
| "learning_rate": 6.959999999999999e-08, | |
| "loss": 1.0694, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 28.588802337646484, | |
| "learning_rate": 7.459999999999999e-08, | |
| "loss": 0.9657, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 28.448694229125977, | |
| "learning_rate": 7.96e-08, | |
| "loss": 0.9004, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 27.52383804321289, | |
| "learning_rate": 8.459999999999999e-08, | |
| "loss": 0.8309, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 30.11163902282715, | |
| "learning_rate": 8.96e-08, | |
| "loss": 0.6978, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 23.116390228271484, | |
| "learning_rate": 9.46e-08, | |
| "loss": 0.6042, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 22.033464431762695, | |
| "learning_rate": 9.959999999999999e-08, | |
| "loss": 0.5676, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 21.632957458496094, | |
| "learning_rate": 1.046e-07, | |
| "loss": 0.4607, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 17.29313087463379, | |
| "learning_rate": 1.096e-07, | |
| "loss": 0.4817, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 19.6722412109375, | |
| "learning_rate": 1.146e-07, | |
| "loss": 0.4447, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 16.24555206298828, | |
| "learning_rate": 1.1959999999999999e-07, | |
| "loss": 0.4028, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 15.58441162109375, | |
| "learning_rate": 1.246e-07, | |
| "loss": 0.3872, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 14.177542686462402, | |
| "learning_rate": 1.296e-07, | |
| "loss": 0.3831, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 14.774901390075684, | |
| "learning_rate": 1.346e-07, | |
| "loss": 0.3505, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 13.562920570373535, | |
| "learning_rate": 1.396e-07, | |
| "loss": 0.3468, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 12.933964729309082, | |
| "learning_rate": 1.446e-07, | |
| "loss": 0.3402, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 12.395262718200684, | |
| "learning_rate": 1.4960000000000002e-07, | |
| "loss": 0.3317, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 14.720085144042969, | |
| "learning_rate": 1.5459999999999997e-07, | |
| "loss": 0.2923, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 14.832243919372559, | |
| "learning_rate": 1.5959999999999997e-07, | |
| "loss": 0.3061, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 11.47397518157959, | |
| "learning_rate": 1.6459999999999998e-07, | |
| "loss": 0.2994, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 17.4206485748291, | |
| "learning_rate": 1.6959999999999998e-07, | |
| "loss": 0.2873, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 12.757941246032715, | |
| "learning_rate": 1.746e-07, | |
| "loss": 0.263, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 15.563201904296875, | |
| "learning_rate": 1.796e-07, | |
| "loss": 0.2659, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 12.566412925720215, | |
| "learning_rate": 1.8459999999999997e-07, | |
| "loss": 0.269, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 11.327591896057129, | |
| "learning_rate": 1.8959999999999998e-07, | |
| "loss": 0.2678, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 11.70463752746582, | |
| "learning_rate": 1.9459999999999998e-07, | |
| "loss": 0.2567, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 11.723058700561523, | |
| "learning_rate": 1.996e-07, | |
| "loss": 0.2552, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_loss": 0.21999725699424744, | |
| "eval_runtime": 7789.979, | |
| "eval_samples_per_second": 1.208, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.12200871952204101, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 15.0565824508667, | |
| "learning_rate": 2.046e-07, | |
| "loss": 0.2629, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 12.149203300476074, | |
| "learning_rate": 2.096e-07, | |
| "loss": 0.2535, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.3349385261535645, | |
| "learning_rate": 2.146e-07, | |
| "loss": 0.2429, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 4.134186744689941, | |
| "learning_rate": 2.1959999999999998e-07, | |
| "loss": 0.2214, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 3.6862359046936035, | |
| "learning_rate": 2.2459999999999999e-07, | |
| "loss": 0.2307, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 4.983838081359863, | |
| "learning_rate": 2.296e-07, | |
| "loss": 0.2327, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 5.041070461273193, | |
| "learning_rate": 2.346e-07, | |
| "loss": 0.224, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 4.438192844390869, | |
| "learning_rate": 2.396e-07, | |
| "loss": 0.2376, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 5.2538161277771, | |
| "learning_rate": 2.446e-07, | |
| "loss": 0.2234, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 4.371279239654541, | |
| "learning_rate": 2.4959999999999996e-07, | |
| "loss": 0.2104, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 3.8350584506988525, | |
| "learning_rate": 2.5459999999999996e-07, | |
| "loss": 0.2221, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 4.0391845703125, | |
| "learning_rate": 2.5959999999999997e-07, | |
| "loss": 0.2147, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 4.656424045562744, | |
| "learning_rate": 2.646e-07, | |
| "loss": 0.2259, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 4.003029823303223, | |
| "learning_rate": 2.696e-07, | |
| "loss": 0.2073, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 3.689852476119995, | |
| "learning_rate": 2.746e-07, | |
| "loss": 0.1964, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 2.934403419494629, | |
| "learning_rate": 2.796e-07, | |
| "loss": 0.1997, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 4.054893493652344, | |
| "learning_rate": 2.846e-07, | |
| "loss": 0.1945, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 4.135422229766846, | |
| "learning_rate": 2.896e-07, | |
| "loss": 0.2014, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 4.4028706550598145, | |
| "learning_rate": 2.9459999999999995e-07, | |
| "loss": 0.2013, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 4.187674045562744, | |
| "learning_rate": 2.9959999999999996e-07, | |
| "loss": 0.2108, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 5.227431297302246, | |
| "learning_rate": 3.0459999999999996e-07, | |
| "loss": 0.21, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 3.673109769821167, | |
| "learning_rate": 3.0959999999999997e-07, | |
| "loss": 0.1994, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 4.424607753753662, | |
| "learning_rate": 3.1459999999999997e-07, | |
| "loss": 0.2143, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 4.143054485321045, | |
| "learning_rate": 3.196e-07, | |
| "loss": 0.1961, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 5.010011672973633, | |
| "learning_rate": 3.246e-07, | |
| "loss": 0.1935, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 4.047982215881348, | |
| "learning_rate": 3.296e-07, | |
| "loss": 0.1925, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 3.367611885070801, | |
| "learning_rate": 3.346e-07, | |
| "loss": 0.2028, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 3.4524459838867188, | |
| "learning_rate": 3.396e-07, | |
| "loss": 0.2027, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 4.717390537261963, | |
| "learning_rate": 3.446e-07, | |
| "loss": 0.2, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 4.299264430999756, | |
| "learning_rate": 3.496e-07, | |
| "loss": 0.2117, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 3.89416241645813, | |
| "learning_rate": 3.546e-07, | |
| "loss": 0.2055, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 4.462187767028809, | |
| "learning_rate": 3.5959999999999996e-07, | |
| "loss": 0.2061, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 4.449103355407715, | |
| "learning_rate": 3.6459999999999997e-07, | |
| "loss": 0.1999, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 2.920868158340454, | |
| "learning_rate": 3.696e-07, | |
| "loss": 0.2063, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 3.8898050785064697, | |
| "learning_rate": 3.746e-07, | |
| "loss": 0.1862, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 3.56378173828125, | |
| "learning_rate": 3.796e-07, | |
| "loss": 0.1974, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 3.6201984882354736, | |
| "learning_rate": 3.846e-07, | |
| "loss": 0.2001, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 4.066067218780518, | |
| "learning_rate": 3.896e-07, | |
| "loss": 0.1954, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 3.4573721885681152, | |
| "learning_rate": 3.946e-07, | |
| "loss": 0.1885, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 3.976249933242798, | |
| "learning_rate": 3.996e-07, | |
| "loss": 0.1928, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_loss": 0.1644940823316574, | |
| "eval_runtime": 8054.0522, | |
| "eval_samples_per_second": 1.169, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.10624899079606007, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 3.10294246673584, | |
| "learning_rate": 4.0439999999999994e-07, | |
| "loss": 0.1868, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 3.2514965534210205, | |
| "learning_rate": 4.0939999999999995e-07, | |
| "loss": 0.194, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 4.614846229553223, | |
| "learning_rate": 4.1439999999999995e-07, | |
| "loss": 0.1998, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 3.231562852859497, | |
| "learning_rate": 4.1939999999999996e-07, | |
| "loss": 0.1841, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 4.466641902923584, | |
| "learning_rate": 4.2439999999999996e-07, | |
| "loss": 0.175, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 3.2693140506744385, | |
| "learning_rate": 4.2939999999999997e-07, | |
| "loss": 0.1824, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "grad_norm": 3.7038068771362305, | |
| "learning_rate": 4.3439999999999997e-07, | |
| "loss": 0.1779, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 4.4231181144714355, | |
| "learning_rate": 4.394e-07, | |
| "loss": 0.1704, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 3.4143128395080566, | |
| "learning_rate": 4.444e-07, | |
| "loss": 0.1559, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 3.6965668201446533, | |
| "learning_rate": 4.494e-07, | |
| "loss": 0.1708, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 4.041949272155762, | |
| "learning_rate": 4.544e-07, | |
| "loss": 0.1625, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 3.2054636478424072, | |
| "learning_rate": 4.5939999999999994e-07, | |
| "loss": 0.1631, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "grad_norm": 4.389201641082764, | |
| "learning_rate": 4.6439999999999995e-07, | |
| "loss": 0.1833, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 3.396207094192505, | |
| "learning_rate": 4.6939999999999995e-07, | |
| "loss": 0.1629, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 3.7798011302948, | |
| "learning_rate": 4.7439999999999996e-07, | |
| "loss": 0.158, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 3.4497921466827393, | |
| "learning_rate": 4.794e-07, | |
| "loss": 0.1659, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 4.368722915649414, | |
| "learning_rate": 4.844e-07, | |
| "loss": 0.1624, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 3.625420570373535, | |
| "learning_rate": 4.894e-07, | |
| "loss": 0.1729, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "grad_norm": 4.549854278564453, | |
| "learning_rate": 4.944e-07, | |
| "loss": 0.1647, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "grad_norm": 3.4779980182647705, | |
| "learning_rate": 4.994e-07, | |
| "loss": 0.1638, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 6.931514739990234, | |
| "learning_rate": 5.043999999999999e-07, | |
| "loss": 0.1617, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 5.141421318054199, | |
| "learning_rate": 5.093999999999999e-07, | |
| "loss": 0.1682, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 3.208487033843994, | |
| "learning_rate": 5.143999999999999e-07, | |
| "loss": 0.1478, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 4.365899562835693, | |
| "learning_rate": 5.194e-07, | |
| "loss": 0.1542, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 3.7716119289398193, | |
| "learning_rate": 5.243999999999999e-07, | |
| "loss": 0.1668, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "grad_norm": 4.948827743530273, | |
| "learning_rate": 5.294e-07, | |
| "loss": 0.1583, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 4.176829814910889, | |
| "learning_rate": 5.343999999999999e-07, | |
| "loss": 0.1607, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 4.278001308441162, | |
| "learning_rate": 5.394e-07, | |
| "loss": 0.1558, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 3.226017951965332, | |
| "learning_rate": 5.443999999999999e-07, | |
| "loss": 0.1495, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "grad_norm": 3.3415420055389404, | |
| "learning_rate": 5.494e-07, | |
| "loss": 0.1762, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 3.7578542232513428, | |
| "learning_rate": 5.543999999999999e-07, | |
| "loss": 0.1684, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "grad_norm": 4.540279388427734, | |
| "learning_rate": 5.594e-07, | |
| "loss": 0.1629, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 2.953626871109009, | |
| "learning_rate": 5.643999999999999e-07, | |
| "loss": 0.1539, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 4.061501502990723, | |
| "learning_rate": 5.694e-07, | |
| "loss": 0.1428, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 3.305874824523926, | |
| "learning_rate": 5.744e-07, | |
| "loss": 0.1713, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 3.4820492267608643, | |
| "learning_rate": 5.794e-07, | |
| "loss": 0.1712, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 4.517523765563965, | |
| "learning_rate": 5.844e-07, | |
| "loss": 0.1626, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 5.121035099029541, | |
| "learning_rate": 5.894e-07, | |
| "loss": 0.1591, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 3.2896835803985596, | |
| "learning_rate": 5.944e-07, | |
| "loss": 0.1652, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 3.9433748722076416, | |
| "learning_rate": 5.994e-07, | |
| "loss": 0.1646, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_loss": 0.15075725317001343, | |
| "eval_runtime": 7921.0228, | |
| "eval_samples_per_second": 1.188, | |
| "eval_steps_per_second": 0.594, | |
| "eval_wer": 0.10159857904085258, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 6.672876834869385, | |
| "learning_rate": 6.044e-07, | |
| "loss": 0.1694, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 3.1442010402679443, | |
| "learning_rate": 6.094e-07, | |
| "loss": 0.1655, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 4.091799736022949, | |
| "learning_rate": 6.143999999999999e-07, | |
| "loss": 0.1568, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 3.820805311203003, | |
| "learning_rate": 6.194e-07, | |
| "loss": 0.1635, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 3.704472303390503, | |
| "learning_rate": 6.242e-07, | |
| "loss": 0.1587, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 3.817859649658203, | |
| "learning_rate": 6.291999999999999e-07, | |
| "loss": 0.1443, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 4.0161452293396, | |
| "learning_rate": 6.342e-07, | |
| "loss": 0.147, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 5.137659072875977, | |
| "learning_rate": 6.392e-07, | |
| "loss": 0.1574, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "grad_norm": 3.7066938877105713, | |
| "learning_rate": 6.442e-07, | |
| "loss": 0.142, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "grad_norm": 4.4287919998168945, | |
| "learning_rate": 6.492e-07, | |
| "loss": 0.1367, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "grad_norm": 4.2995381355285645, | |
| "learning_rate": 6.542e-07, | |
| "loss": 0.1226, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 3.6767024993896484, | |
| "learning_rate": 6.592e-07, | |
| "loss": 0.1533, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 4.893792152404785, | |
| "learning_rate": 6.642e-07, | |
| "loss": 0.1304, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 3.13, | |
| "grad_norm": 1.7213168144226074, | |
| "learning_rate": 6.692e-07, | |
| "loss": 0.143, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 4.113219738006592, | |
| "learning_rate": 6.742e-07, | |
| "loss": 0.1311, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 3.17, | |
| "grad_norm": 4.480813980102539, | |
| "learning_rate": 6.792e-07, | |
| "loss": 0.1429, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 4.102989196777344, | |
| "learning_rate": 6.842e-07, | |
| "loss": 0.1361, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "grad_norm": 4.409080982208252, | |
| "learning_rate": 6.892e-07, | |
| "loss": 0.1271, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 3.579719305038452, | |
| "learning_rate": 6.942e-07, | |
| "loss": 0.1303, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "grad_norm": 5.013184547424316, | |
| "learning_rate": 6.992e-07, | |
| "loss": 0.1359, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "grad_norm": 4.512373447418213, | |
| "learning_rate": 7.042e-07, | |
| "loss": 0.1359, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 4.140831470489502, | |
| "learning_rate": 7.092e-07, | |
| "loss": 0.1239, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 3.34, | |
| "grad_norm": 3.4033215045928955, | |
| "learning_rate": 7.141999999999999e-07, | |
| "loss": 0.133, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 3.3019678592681885, | |
| "learning_rate": 7.191999999999999e-07, | |
| "loss": 0.1302, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 3.88134765625, | |
| "learning_rate": 7.241999999999999e-07, | |
| "loss": 0.1437, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 3.41, | |
| "grad_norm": 4.493922710418701, | |
| "learning_rate": 7.291999999999999e-07, | |
| "loss": 0.1346, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 3.43, | |
| "grad_norm": 4.8288421630859375, | |
| "learning_rate": 7.341999999999999e-07, | |
| "loss": 0.1331, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "grad_norm": 4.28986120223999, | |
| "learning_rate": 7.392e-07, | |
| "loss": 0.1309, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 3.411266326904297, | |
| "learning_rate": 7.441999999999999e-07, | |
| "loss": 0.1549, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 3.317307949066162, | |
| "learning_rate": 7.492e-07, | |
| "loss": 0.1265, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 3.53, | |
| "grad_norm": 4.695957183837891, | |
| "learning_rate": 7.541999999999999e-07, | |
| "loss": 0.1422, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 3.6588873863220215, | |
| "learning_rate": 7.592e-07, | |
| "loss": 0.1437, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 3.57, | |
| "grad_norm": 5.599457740783691, | |
| "learning_rate": 7.641999999999999e-07, | |
| "loss": 0.1436, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 4.065281391143799, | |
| "learning_rate": 7.692e-07, | |
| "loss": 0.1281, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 3.62, | |
| "grad_norm": 4.9737935066223145, | |
| "learning_rate": 7.741999999999999e-07, | |
| "loss": 0.1293, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 3.3383023738861084, | |
| "learning_rate": 7.792e-07, | |
| "loss": 0.1329, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 3.67, | |
| "grad_norm": 3.6315879821777344, | |
| "learning_rate": 7.841999999999999e-07, | |
| "loss": 0.1273, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 3.69, | |
| "grad_norm": 4.887773036956787, | |
| "learning_rate": 7.892e-07, | |
| "loss": 0.136, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 4.8274455070495605, | |
| "learning_rate": 7.942e-07, | |
| "loss": 0.1438, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 4.641810417175293, | |
| "learning_rate": 7.992e-07, | |
| "loss": 0.1333, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "eval_loss": 0.14375422894954681, | |
| "eval_runtime": 7811.8504, | |
| "eval_samples_per_second": 1.205, | |
| "eval_steps_per_second": 0.603, | |
| "eval_wer": 0.09701275633780074, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 5.432655334472656, | |
| "learning_rate": 8.042e-07, | |
| "loss": 0.1238, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 3.78, | |
| "grad_norm": 4.146833419799805, | |
| "learning_rate": 8.092e-07, | |
| "loss": 0.1234, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 3.81, | |
| "grad_norm": 4.496214389801025, | |
| "learning_rate": 8.142e-07, | |
| "loss": 0.1385, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 3.83, | |
| "grad_norm": 4.550703525543213, | |
| "learning_rate": 8.192e-07, | |
| "loss": 0.1327, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 4.395373821258545, | |
| "learning_rate": 8.242e-07, | |
| "loss": 0.1336, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 3.292907238006592, | |
| "learning_rate": 8.292e-07, | |
| "loss": 0.1243, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 3.963186502456665, | |
| "learning_rate": 8.342e-07, | |
| "loss": 0.1287, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 3.482177972793579, | |
| "learning_rate": 8.391999999999999e-07, | |
| "loss": 0.1305, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 3.331468105316162, | |
| "learning_rate": 8.441999999999999e-07, | |
| "loss": 0.1308, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 3.97, | |
| "grad_norm": 5.990370750427246, | |
| "learning_rate": 8.492e-07, | |
| "loss": 0.1411, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 3.99, | |
| "grad_norm": 3.512251615524292, | |
| "learning_rate": 8.541999999999999e-07, | |
| "loss": 0.132, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 3.779296398162842, | |
| "learning_rate": 8.592e-07, | |
| "loss": 0.1277, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 4.04, | |
| "grad_norm": 4.156597137451172, | |
| "learning_rate": 8.641999999999999e-07, | |
| "loss": 0.1082, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 3.5820066928863525, | |
| "learning_rate": 8.692e-07, | |
| "loss": 0.117, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 4.09, | |
| "grad_norm": 3.3737404346466064, | |
| "learning_rate": 8.741999999999999e-07, | |
| "loss": 0.1068, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "grad_norm": 4.2788166999816895, | |
| "learning_rate": 8.792e-07, | |
| "loss": 0.1129, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 3.872753381729126, | |
| "learning_rate": 8.841999999999999e-07, | |
| "loss": 0.1124, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 5.087253570556641, | |
| "learning_rate": 8.892e-07, | |
| "loss": 0.1149, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 4.363770484924316, | |
| "learning_rate": 8.941999999999999e-07, | |
| "loss": 0.1105, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 3.211000680923462, | |
| "learning_rate": 8.992e-07, | |
| "loss": 0.1162, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.23, | |
| "grad_norm": 4.609809875488281, | |
| "learning_rate": 9.042e-07, | |
| "loss": 0.1165, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 3.679553508758545, | |
| "learning_rate": 9.092e-07, | |
| "loss": 0.111, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 3.6184656620025635, | |
| "learning_rate": 9.142e-07, | |
| "loss": 0.1243, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 4.844442367553711, | |
| "learning_rate": 9.192e-07, | |
| "loss": 0.1112, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 5.363363265991211, | |
| "learning_rate": 9.242e-07, | |
| "loss": 0.1193, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 2.0149080753326416, | |
| "learning_rate": 9.292e-07, | |
| "loss": 0.1143, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 5.45695161819458, | |
| "learning_rate": 9.342e-07, | |
| "loss": 0.1056, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "grad_norm": 5.325240612030029, | |
| "learning_rate": 9.391999999999999e-07, | |
| "loss": 0.114, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "grad_norm": 3.116534471511841, | |
| "learning_rate": 9.442e-07, | |
| "loss": 0.0995, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "grad_norm": 2.405822515487671, | |
| "learning_rate": 9.492e-07, | |
| "loss": 0.1165, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 4.46, | |
| "grad_norm": 4.517841339111328, | |
| "learning_rate": 9.542e-07, | |
| "loss": 0.1052, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 2.7329490184783936, | |
| "learning_rate": 9.592e-07, | |
| "loss": 0.102, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 4.51, | |
| "grad_norm": 3.871314287185669, | |
| "learning_rate": 9.641999999999998e-07, | |
| "loss": 0.1081, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 3.3892979621887207, | |
| "learning_rate": 9.692e-07, | |
| "loss": 0.1138, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 4.42065954208374, | |
| "learning_rate": 9.742e-07, | |
| "loss": 0.106, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 4.373811721801758, | |
| "learning_rate": 9.791999999999999e-07, | |
| "loss": 0.1043, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 5.577483177185059, | |
| "learning_rate": 9.841999999999998e-07, | |
| "loss": 0.0962, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "grad_norm": 3.65592360496521, | |
| "learning_rate": 9.892e-07, | |
| "loss": 0.1161, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 4.68386173248291, | |
| "learning_rate": 9.942e-07, | |
| "loss": 0.1071, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "grad_norm": 3.9556305408477783, | |
| "learning_rate": 9.992e-07, | |
| "loss": 0.1027, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "eval_loss": 0.14086556434631348, | |
| "eval_runtime": 7847.8302, | |
| "eval_samples_per_second": 1.2, | |
| "eval_steps_per_second": 0.6, | |
| "eval_wer": 0.09421927983206846, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "step": 5000, | |
| "total_flos": 1.632967868976988e+20, | |
| "train_loss": 0.2694411336898804, | |
| "train_runtime": 106369.5753, | |
| "train_samples_per_second": 1.504, | |
| "train_steps_per_second": 0.047 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 1000, | |
| "total_flos": 1.632967868976988e+20, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |