| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.32156408772268313, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 58.440006256103516, | |
| "learning_rate": 1e-08, | |
| "loss": 1.3813, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 59.934757232666016, | |
| "learning_rate": 2.2e-08, | |
| "loss": 1.469, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 31.011018753051758, | |
| "learning_rate": 3.4500000000000005e-08, | |
| "loss": 1.2226, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 59.818233489990234, | |
| "learning_rate": 4.7e-08, | |
| "loss": 1.2458, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 60.51572036743164, | |
| "learning_rate": 5.95e-08, | |
| "loss": 1.2781, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 51.360103607177734, | |
| "learning_rate": 7.2e-08, | |
| "loss": 1.4055, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 73.35002136230469, | |
| "learning_rate": 8.45e-08, | |
| "loss": 1.3354, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 69.32823944091797, | |
| "learning_rate": 9.7e-08, | |
| "loss": 1.2005, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 51.02174377441406, | |
| "learning_rate": 1.095e-07, | |
| "loss": 1.3853, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 72.20179748535156, | |
| "learning_rate": 1.2199999999999998e-07, | |
| "loss": 1.4476, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 108.30382537841797, | |
| "learning_rate": 1.345e-07, | |
| "loss": 1.2339, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 66.15994262695312, | |
| "learning_rate": 1.4699999999999998e-07, | |
| "loss": 1.379, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 47.82923126220703, | |
| "learning_rate": 1.595e-07, | |
| "loss": 1.1467, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 85.7218246459961, | |
| "learning_rate": 1.7199999999999998e-07, | |
| "loss": 1.1622, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 68.25504302978516, | |
| "learning_rate": 1.845e-07, | |
| "loss": 1.1413, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 106.06077575683594, | |
| "learning_rate": 1.97e-07, | |
| "loss": 1.0855, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 79.60690307617188, | |
| "learning_rate": 2.095e-07, | |
| "loss": 0.929, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 42.14814376831055, | |
| "learning_rate": 2.22e-07, | |
| "loss": 0.8728, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 37.4913444519043, | |
| "learning_rate": 2.3449999999999996e-07, | |
| "loss": 0.6651, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 41.89991760253906, | |
| "learning_rate": 2.47e-07, | |
| "loss": 0.5875, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 75.21453094482422, | |
| "learning_rate": 2.595e-07, | |
| "loss": 0.6868, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 21.09180450439453, | |
| "learning_rate": 2.72e-07, | |
| "loss": 0.741, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 44.54707336425781, | |
| "learning_rate": 2.845e-07, | |
| "loss": 0.3898, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 31.656843185424805, | |
| "learning_rate": 2.9699999999999997e-07, | |
| "loss": 0.422, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 56.28642654418945, | |
| "learning_rate": 3.0949999999999996e-07, | |
| "loss": 0.3803, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 38.66410827636719, | |
| "learning_rate": 3.22e-07, | |
| "loss": 0.5062, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 31.183727264404297, | |
| "learning_rate": 3.345e-07, | |
| "loss": 0.4075, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 23.618703842163086, | |
| "learning_rate": 3.4699999999999997e-07, | |
| "loss": 0.3627, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 70.09487915039062, | |
| "learning_rate": 3.5949999999999996e-07, | |
| "loss": 0.3087, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 74.42188262939453, | |
| "learning_rate": 3.72e-07, | |
| "loss": 0.4021, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 44.99939727783203, | |
| "learning_rate": 3.845e-07, | |
| "loss": 0.3203, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 42.77998352050781, | |
| "learning_rate": 3.97e-07, | |
| "loss": 0.3797, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 64.61412811279297, | |
| "learning_rate": 4.0949999999999995e-07, | |
| "loss": 0.3403, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 29.286806106567383, | |
| "learning_rate": 4.2199999999999994e-07, | |
| "loss": 0.2879, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 58.146263122558594, | |
| "learning_rate": 4.345e-07, | |
| "loss": 0.4017, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 44.624202728271484, | |
| "learning_rate": 4.4699999999999997e-07, | |
| "loss": 0.3698, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 47.91656494140625, | |
| "learning_rate": 4.595e-07, | |
| "loss": 0.4008, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 36.263668060302734, | |
| "learning_rate": 4.7199999999999994e-07, | |
| "loss": 0.2041, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 12.398943901062012, | |
| "learning_rate": 4.845e-07, | |
| "loss": 0.2978, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 4.42283821105957, | |
| "learning_rate": 4.97e-07, | |
| "loss": 0.2614, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 0.29864633083343506, | |
| "eval_runtime": 7667.7674, | |
| "eval_samples_per_second": 1.228, | |
| "eval_steps_per_second": 0.614, | |
| "eval_wer": 0.14664944291942517, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 59.681270599365234, | |
| "learning_rate": 5.095e-07, | |
| "loss": 0.2588, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 27.44911766052246, | |
| "learning_rate": 5.22e-07, | |
| "loss": 0.2839, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 56.26525115966797, | |
| "learning_rate": 5.344999999999999e-07, | |
| "loss": 0.2329, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 112.37168884277344, | |
| "learning_rate": 5.47e-07, | |
| "loss": 0.2746, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 92.82706451416016, | |
| "learning_rate": 5.595e-07, | |
| "loss": 0.2621, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 17.20562744140625, | |
| "learning_rate": 5.719999999999999e-07, | |
| "loss": 0.1826, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 31.119354248046875, | |
| "learning_rate": 5.845e-07, | |
| "loss": 0.2487, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 7.4088850021362305, | |
| "learning_rate": 5.97e-07, | |
| "loss": 0.297, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 72.84540557861328, | |
| "learning_rate": 6.095e-07, | |
| "loss": 0.3108, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 38.68337631225586, | |
| "learning_rate": 6.219999999999999e-07, | |
| "loss": 0.2819, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 5.215000152587891, | |
| "learning_rate": 6.344999999999999e-07, | |
| "loss": 0.1953, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 41.42685317993164, | |
| "learning_rate": 6.47e-07, | |
| "loss": 0.2333, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 6.224233150482178, | |
| "learning_rate": 6.595e-07, | |
| "loss": 0.2213, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 48.12126541137695, | |
| "learning_rate": 6.72e-07, | |
| "loss": 0.3121, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 23.151887893676758, | |
| "learning_rate": 6.845e-07, | |
| "loss": 0.2076, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 53.516395568847656, | |
| "learning_rate": 6.97e-07, | |
| "loss": 0.2835, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 41.62558364868164, | |
| "learning_rate": 7.094999999999999e-07, | |
| "loss": 0.2574, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 98.05493927001953, | |
| "learning_rate": 7.219999999999999e-07, | |
| "loss": 0.3034, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 86.28963470458984, | |
| "learning_rate": 7.345e-07, | |
| "loss": 0.2657, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.8914854526519775, | |
| "learning_rate": 7.47e-07, | |
| "loss": 0.2636, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 56.13273239135742, | |
| "learning_rate": 7.594999999999999e-07, | |
| "loss": 0.1522, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 12.941767692565918, | |
| "learning_rate": 7.72e-07, | |
| "loss": 0.2097, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 13.613518714904785, | |
| "learning_rate": 7.845e-07, | |
| "loss": 0.2303, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 23.761892318725586, | |
| "learning_rate": 7.970000000000001e-07, | |
| "loss": 0.2763, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 31.896230697631836, | |
| "learning_rate": 8.094999999999999e-07, | |
| "loss": 0.2722, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 51.43158721923828, | |
| "learning_rate": 8.219999999999999e-07, | |
| "loss": 0.2228, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 2.879077196121216, | |
| "learning_rate": 8.345e-07, | |
| "loss": 0.2694, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 98.36167907714844, | |
| "learning_rate": 8.469999999999999e-07, | |
| "loss": 0.3088, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 16.084274291992188, | |
| "learning_rate": 8.595e-07, | |
| "loss": 0.1828, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 5.136277675628662, | |
| "learning_rate": 8.72e-07, | |
| "loss": 0.1753, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 64.78803253173828, | |
| "learning_rate": 8.845e-07, | |
| "loss": 0.198, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 41.84619903564453, | |
| "learning_rate": 8.969999999999999e-07, | |
| "loss": 0.2894, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 45.18673324584961, | |
| "learning_rate": 9.094999999999999e-07, | |
| "loss": 0.1844, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 20.42123794555664, | |
| "learning_rate": 9.22e-07, | |
| "loss": 0.2576, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 8.751657485961914, | |
| "learning_rate": 9.344999999999999e-07, | |
| "loss": 0.2492, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 29.69828224182129, | |
| "learning_rate": 9.469999999999999e-07, | |
| "loss": 0.1479, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 22.91360855102539, | |
| "learning_rate": 9.594999999999999e-07, | |
| "loss": 0.2164, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 17.22205352783203, | |
| "learning_rate": 9.72e-07, | |
| "loss": 0.2276, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 83.4366683959961, | |
| "learning_rate": 9.845e-07, | |
| "loss": 0.2717, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 11.829337120056152, | |
| "learning_rate": 9.97e-07, | |
| "loss": 0.2632, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_loss": 0.22439254820346832, | |
| "eval_runtime": 7771.0673, | |
| "eval_samples_per_second": 1.211, | |
| "eval_steps_per_second": 0.606, | |
| "eval_wer": 0.13156789924107865, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 42.36271667480469, | |
| "learning_rate": 9.936666666666667e-07, | |
| "loss": 0.21, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 57.45354461669922, | |
| "learning_rate": 9.853333333333333e-07, | |
| "loss": 0.2244, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 17.302165985107422, | |
| "learning_rate": 9.773333333333333e-07, | |
| "loss": 0.2492, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 41.777069091796875, | |
| "learning_rate": 9.69e-07, | |
| "loss": 0.1598, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 37.14485549926758, | |
| "learning_rate": 9.606666666666666e-07, | |
| "loss": 0.2483, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 53.22433090209961, | |
| "learning_rate": 9.523333333333333e-07, | |
| "loss": 0.1913, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 78.79158782958984, | |
| "learning_rate": 9.439999999999999e-07, | |
| "loss": 0.3075, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 2.1396484375, | |
| "learning_rate": 9.356666666666666e-07, | |
| "loss": 0.2427, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 9.334494590759277, | |
| "learning_rate": 9.273333333333333e-07, | |
| "loss": 0.2201, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 8.948480606079102, | |
| "learning_rate": 9.19e-07, | |
| "loss": 0.2047, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.004768133163452, | |
| "learning_rate": 9.106666666666666e-07, | |
| "loss": 0.1928, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 3.458395481109619, | |
| "learning_rate": 9.023333333333333e-07, | |
| "loss": 0.1788, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 61.66895294189453, | |
| "learning_rate": 8.939999999999999e-07, | |
| "loss": 0.1959, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 45.452789306640625, | |
| "learning_rate": 8.856666666666666e-07, | |
| "loss": 0.211, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 18.07378578186035, | |
| "learning_rate": 8.773333333333332e-07, | |
| "loss": 0.2391, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 46.68052291870117, | |
| "learning_rate": 8.69e-07, | |
| "loss": 0.1782, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 5.451249599456787, | |
| "learning_rate": 8.606666666666667e-07, | |
| "loss": 0.1569, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.6330296993255615, | |
| "learning_rate": 8.523333333333334e-07, | |
| "loss": 0.1381, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 43.628761291503906, | |
| "learning_rate": 8.439999999999999e-07, | |
| "loss": 0.1943, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 42.83442687988281, | |
| "learning_rate": 8.356666666666666e-07, | |
| "loss": 0.1937, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 41.783485412597656, | |
| "learning_rate": 8.273333333333333e-07, | |
| "loss": 0.1738, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 43.905025482177734, | |
| "learning_rate": 8.189999999999999e-07, | |
| "loss": 0.2068, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 35.906982421875, | |
| "learning_rate": 8.106666666666666e-07, | |
| "loss": 0.3004, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 75.37654113769531, | |
| "learning_rate": 8.023333333333333e-07, | |
| "loss": 0.2073, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.4306688904762268, | |
| "learning_rate": 7.94e-07, | |
| "loss": 0.1722, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 48.76789093017578, | |
| "learning_rate": 7.856666666666665e-07, | |
| "loss": 0.2808, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 21.527475357055664, | |
| "learning_rate": 7.773333333333333e-07, | |
| "loss": 0.1416, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 6.267962455749512, | |
| "learning_rate": 7.69e-07, | |
| "loss": 0.1522, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 38.2205696105957, | |
| "learning_rate": 7.606666666666667e-07, | |
| "loss": 0.1988, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 2.1994435787200928, | |
| "learning_rate": 7.523333333333333e-07, | |
| "loss": 0.2384, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 21.002376556396484, | |
| "learning_rate": 7.44e-07, | |
| "loss": 0.198, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 66.96015167236328, | |
| "learning_rate": 7.356666666666667e-07, | |
| "loss": 0.2185, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 16.91470718383789, | |
| "learning_rate": 7.273333333333333e-07, | |
| "loss": 0.2149, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 12.189261436462402, | |
| "learning_rate": 7.189999999999999e-07, | |
| "loss": 0.1907, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 5.648806095123291, | |
| "learning_rate": 7.106666666666666e-07, | |
| "loss": 0.1634, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 6.627074241638184, | |
| "learning_rate": 7.023333333333333e-07, | |
| "loss": 0.2275, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 42.446556091308594, | |
| "learning_rate": 6.939999999999999e-07, | |
| "loss": 0.1888, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 19.29751968383789, | |
| "learning_rate": 6.856666666666667e-07, | |
| "loss": 0.1603, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 88.98928833007812, | |
| "learning_rate": 6.773333333333334e-07, | |
| "loss": 0.2295, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 13.686836242675781, | |
| "learning_rate": 6.69e-07, | |
| "loss": 0.1694, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 0.20859745144844055, | |
| "eval_runtime": 7797.8116, | |
| "eval_samples_per_second": 1.207, | |
| "eval_steps_per_second": 0.604, | |
| "eval_wer": 0.12344582593250444, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 44.50759506225586, | |
| "learning_rate": 6.606666666666666e-07, | |
| "loss": 0.2145, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 53.09928512573242, | |
| "learning_rate": 6.523333333333333e-07, | |
| "loss": 0.1721, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 53.15538024902344, | |
| "learning_rate": 6.44e-07, | |
| "loss": 0.2002, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 28.469669342041016, | |
| "learning_rate": 6.356666666666667e-07, | |
| "loss": 0.184, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 63.475502014160156, | |
| "learning_rate": 6.273333333333333e-07, | |
| "loss": 0.1787, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 38.70827865600586, | |
| "learning_rate": 6.189999999999999e-07, | |
| "loss": 0.1815, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 48.54985809326172, | |
| "learning_rate": 6.106666666666666e-07, | |
| "loss": 0.1589, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 25.11480140686035, | |
| "learning_rate": 6.023333333333333e-07, | |
| "loss": 0.1795, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 17.55191421508789, | |
| "learning_rate": 5.939999999999999e-07, | |
| "loss": 0.1464, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 53.10033416748047, | |
| "learning_rate": 5.856666666666667e-07, | |
| "loss": 0.2208, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.9264570474624634, | |
| "learning_rate": 5.773333333333334e-07, | |
| "loss": 0.1538, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 11.413477897644043, | |
| "learning_rate": 5.69e-07, | |
| "loss": 0.2038, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 53.29780578613281, | |
| "learning_rate": 5.606666666666666e-07, | |
| "loss": 0.1312, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 3.7282674312591553, | |
| "learning_rate": 5.523333333333333e-07, | |
| "loss": 0.2328, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.443333333333333e-07, | |
| "loss": 0.2432, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 22.73953628540039, | |
| "learning_rate": 5.36e-07, | |
| "loss": 0.2353, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 19.394702911376953, | |
| "learning_rate": 5.276666666666666e-07, | |
| "loss": 0.0976, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 1.5477691888809204, | |
| "learning_rate": 5.193333333333332e-07, | |
| "loss": 0.1921, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 33.865806579589844, | |
| "learning_rate": 5.11e-07, | |
| "loss": 0.1957, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 8.566771507263184, | |
| "learning_rate": 5.026666666666667e-07, | |
| "loss": 0.1601, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 22.204965591430664, | |
| "learning_rate": 4.943333333333333e-07, | |
| "loss": 0.1894, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 32.81788635253906, | |
| "learning_rate": 4.86e-07, | |
| "loss": 0.2235, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 60.057193756103516, | |
| "learning_rate": 4.776666666666667e-07, | |
| "loss": 0.1714, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 11.461939811706543, | |
| "learning_rate": 4.693333333333333e-07, | |
| "loss": 0.1757, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 29.48383331298828, | |
| "learning_rate": 4.61e-07, | |
| "loss": 0.2581, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 5.3872270584106445, | |
| "learning_rate": 4.526666666666666e-07, | |
| "loss": 0.1493, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 24.588903427124023, | |
| "learning_rate": 4.4433333333333333e-07, | |
| "loss": 0.2003, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 42.52607727050781, | |
| "learning_rate": 4.36e-07, | |
| "loss": 0.2212, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 17.575077056884766, | |
| "learning_rate": 4.2766666666666664e-07, | |
| "loss": 0.1639, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 77.39998626708984, | |
| "learning_rate": 4.193333333333333e-07, | |
| "loss": 0.2154, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 65.65005493164062, | |
| "learning_rate": 4.1099999999999996e-07, | |
| "loss": 0.2396, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 45.75455093383789, | |
| "learning_rate": 4.0266666666666667e-07, | |
| "loss": 0.159, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 44.56821060180664, | |
| "learning_rate": 3.943333333333333e-07, | |
| "loss": 0.1935, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 11.89593505859375, | |
| "learning_rate": 3.86e-07, | |
| "loss": 0.1649, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 7.169790267944336, | |
| "learning_rate": 3.7766666666666665e-07, | |
| "loss": 0.1925, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 141.48680114746094, | |
| "learning_rate": 3.693333333333333e-07, | |
| "loss": 0.1757, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.0640227794647217, | |
| "learning_rate": 3.6099999999999996e-07, | |
| "loss": 0.1733, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 6.541628360748291, | |
| "learning_rate": 3.526666666666667e-07, | |
| "loss": 0.1857, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 62.5667610168457, | |
| "learning_rate": 3.4433333333333333e-07, | |
| "loss": 0.1946, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 2.5611398220062256, | |
| "learning_rate": 3.36e-07, | |
| "loss": 0.1658, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_loss": 0.19874949753284454, | |
| "eval_runtime": 7779.7728, | |
| "eval_samples_per_second": 1.21, | |
| "eval_steps_per_second": 0.605, | |
| "eval_wer": 0.12049087679638301, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 17.723182678222656, | |
| "learning_rate": 3.2766666666666665e-07, | |
| "loss": 0.1771, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 126.36833953857422, | |
| "learning_rate": 3.1933333333333336e-07, | |
| "loss": 0.1923, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 37.344024658203125, | |
| "learning_rate": 3.1099999999999997e-07, | |
| "loss": 0.1657, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 44.135009765625, | |
| "learning_rate": 3.026666666666666e-07, | |
| "loss": 0.2138, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 35.51801300048828, | |
| "learning_rate": 2.9433333333333334e-07, | |
| "loss": 0.2213, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 2.6660304069519043, | |
| "learning_rate": 2.8599999999999994e-07, | |
| "loss": 0.1414, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 27.147336959838867, | |
| "learning_rate": 2.7766666666666665e-07, | |
| "loss": 0.2362, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 44.95083999633789, | |
| "learning_rate": 2.693333333333333e-07, | |
| "loss": 0.1714, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 58.55304718017578, | |
| "learning_rate": 2.61e-07, | |
| "loss": 0.2604, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 16.97317123413086, | |
| "learning_rate": 2.526666666666666e-07, | |
| "loss": 0.1999, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 53.99211502075195, | |
| "learning_rate": 2.4433333333333334e-07, | |
| "loss": 0.1567, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 2.541445016860962, | |
| "learning_rate": 2.3599999999999997e-07, | |
| "loss": 0.1516, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 10.404768943786621, | |
| "learning_rate": 2.2766666666666665e-07, | |
| "loss": 0.1302, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 49.73779296875, | |
| "learning_rate": 2.193333333333333e-07, | |
| "loss": 0.1937, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 28.959285736083984, | |
| "learning_rate": 2.1099999999999997e-07, | |
| "loss": 0.2463, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 32.72858428955078, | |
| "learning_rate": 2.0266666666666666e-07, | |
| "loss": 0.1744, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 27.81814193725586, | |
| "learning_rate": 1.9433333333333331e-07, | |
| "loss": 0.1937, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 6.327540874481201, | |
| "learning_rate": 1.86e-07, | |
| "loss": 0.159, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 16.851764678955078, | |
| "learning_rate": 1.7766666666666666e-07, | |
| "loss": 0.1587, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 81.7389907836914, | |
| "learning_rate": 1.6933333333333334e-07, | |
| "loss": 0.1913, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 114.85218811035156, | |
| "learning_rate": 1.61e-07, | |
| "loss": 0.2065, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 41.08122253417969, | |
| "learning_rate": 1.5266666666666666e-07, | |
| "loss": 0.1874, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 36.792579650878906, | |
| "learning_rate": 1.4433333333333334e-07, | |
| "loss": 0.2145, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 2.39276123046875, | |
| "learning_rate": 1.36e-07, | |
| "loss": 0.1627, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 29.183069229125977, | |
| "learning_rate": 1.2766666666666668e-07, | |
| "loss": 0.1903, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 58.75885009765625, | |
| "learning_rate": 1.1933333333333332e-07, | |
| "loss": 0.2567, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 27.792022705078125, | |
| "learning_rate": 1.11e-07, | |
| "loss": 0.1841, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 36.91849899291992, | |
| "learning_rate": 1.0266666666666666e-07, | |
| "loss": 0.1458, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 40.75591278076172, | |
| "learning_rate": 9.433333333333333e-08, | |
| "loss": 0.1731, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 63.75225830078125, | |
| "learning_rate": 8.599999999999999e-08, | |
| "loss": 0.1597, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 39.913543701171875, | |
| "learning_rate": 7.766666666666666e-08, | |
| "loss": 0.2853, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 9.823902130126953, | |
| "learning_rate": 6.933333333333333e-08, | |
| "loss": 0.2302, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 31.42165184020996, | |
| "learning_rate": 6.099999999999999e-08, | |
| "loss": 0.2028, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 3.4078125953674316, | |
| "learning_rate": 5.266666666666666e-08, | |
| "loss": 0.1645, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 10.65793514251709, | |
| "learning_rate": 4.433333333333333e-08, | |
| "loss": 0.2271, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 16.98055648803711, | |
| "learning_rate": 3.6e-08, | |
| "loss": 0.1374, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 5.180314064025879, | |
| "learning_rate": 2.7666666666666663e-08, | |
| "loss": 0.1134, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 34.10147476196289, | |
| "learning_rate": 1.9333333333333334e-08, | |
| "loss": 0.1612, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 39.326698303222656, | |
| "learning_rate": 1.1e-08, | |
| "loss": 0.1835, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 22.33516502380371, | |
| "learning_rate": 2.6666666666666666e-09, | |
| "loss": 0.1391, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 0.19753539562225342, | |
| "eval_runtime": 7730.2125, | |
| "eval_samples_per_second": 1.218, | |
| "eval_steps_per_second": 0.609, | |
| "eval_wer": 0.11905377038591959, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "step": 5000, | |
| "total_flos": 1.0206049181106176e+19, | |
| "train_loss": 0.32176169362068174, | |
| "train_runtime": 49101.8688, | |
| "train_samples_per_second": 0.204, | |
| "train_steps_per_second": 0.102 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "total_flos": 1.0206049181106176e+19, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |