| { | |
| "best_metric": 27.557846255071826, | |
| "best_model_checkpoint": "./whisper-small-vietnamese\\checkpoint-2000", | |
| "epoch": 22.988505747126435, | |
| "eval_steps": 1000, | |
| "global_step": 4000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 118.77095794677734, | |
| "learning_rate": 4.2000000000000006e-07, | |
| "loss": 5.0409, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 29.974971771240234, | |
| "learning_rate": 9.200000000000001e-07, | |
| "loss": 4.0132, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 25.654932022094727, | |
| "learning_rate": 1.42e-06, | |
| "loss": 2.9199, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 16.862260818481445, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 1.7066, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 13.715407371520996, | |
| "learning_rate": 2.42e-06, | |
| "loss": 1.2867, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 16.711305618286133, | |
| "learning_rate": 2.92e-06, | |
| "loss": 1.1578, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 14.38102912902832, | |
| "learning_rate": 3.4200000000000007e-06, | |
| "loss": 1.0035, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 15.285762786865234, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 0.8052, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 8.91514778137207, | |
| "learning_rate": 4.42e-06, | |
| "loss": 0.286, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 7.789880275726318, | |
| "learning_rate": 4.92e-06, | |
| "loss": 0.2305, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 8.093725204467773, | |
| "learning_rate": 5.420000000000001e-06, | |
| "loss": 0.2184, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 4.7622785568237305, | |
| "learning_rate": 5.92e-06, | |
| "loss": 0.2219, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 7.469940662384033, | |
| "learning_rate": 6.42e-06, | |
| "loss": 0.2067, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 4.024804592132568, | |
| "learning_rate": 6.92e-06, | |
| "loss": 0.2128, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 4.569382667541504, | |
| "learning_rate": 7.420000000000001e-06, | |
| "loss": 0.1294, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 5.233477592468262, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.1217, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 6.401954174041748, | |
| "learning_rate": 8.42e-06, | |
| "loss": 0.1272, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 4.058222770690918, | |
| "learning_rate": 8.920000000000001e-06, | |
| "loss": 0.1188, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "grad_norm": 4.228504657745361, | |
| "learning_rate": 9.42e-06, | |
| "loss": 0.1294, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "grad_norm": 5.160878658294678, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.1241, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 3.9591662883758545, | |
| "learning_rate": 9.940000000000001e-06, | |
| "loss": 0.1264, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 4.170319080352783, | |
| "learning_rate": 9.86857142857143e-06, | |
| "loss": 0.0793, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 3.4333674907684326, | |
| "learning_rate": 9.797142857142858e-06, | |
| "loss": 0.0643, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 4.089048862457275, | |
| "learning_rate": 9.725714285714287e-06, | |
| "loss": 0.0668, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.59, | |
| "grad_norm": 3.6274781227111816, | |
| "learning_rate": 9.654285714285716e-06, | |
| "loss": 0.069, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "grad_norm": 3.2134461402893066, | |
| "learning_rate": 9.582857142857143e-06, | |
| "loss": 0.0759, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 3.4892189502716064, | |
| "learning_rate": 9.511428571428572e-06, | |
| "loss": 0.0698, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 2.3172402381896973, | |
| "learning_rate": 9.440000000000001e-06, | |
| "loss": 0.0651, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 2.6511449813842773, | |
| "learning_rate": 9.368571428571428e-06, | |
| "loss": 0.0373, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.31, | |
| "grad_norm": 3.1017539501190186, | |
| "learning_rate": 9.297142857142857e-06, | |
| "loss": 0.0415, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 3.3337509632110596, | |
| "learning_rate": 9.225714285714286e-06, | |
| "loss": 0.0388, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 2.3944547176361084, | |
| "learning_rate": 9.154285714285715e-06, | |
| "loss": 0.0411, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.74, | |
| "grad_norm": 3.1757237911224365, | |
| "learning_rate": 9.082857142857143e-06, | |
| "loss": 0.0381, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 4.307478427886963, | |
| "learning_rate": 9.011428571428572e-06, | |
| "loss": 0.0434, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.03, | |
| "grad_norm": 1.4969714879989624, | |
| "learning_rate": 8.94e-06, | |
| "loss": 0.0384, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.17, | |
| "grad_norm": 1.416281819343567, | |
| "learning_rate": 8.86857142857143e-06, | |
| "loss": 0.0207, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.32, | |
| "grad_norm": 4.1161322593688965, | |
| "learning_rate": 8.797142857142857e-06, | |
| "loss": 0.0222, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 5.46, | |
| "grad_norm": 0.9343141317367554, | |
| "learning_rate": 8.725714285714286e-06, | |
| "loss": 0.0213, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 3.1798384189605713, | |
| "learning_rate": 8.654285714285715e-06, | |
| "loss": 0.0239, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "grad_norm": 1.7415159940719604, | |
| "learning_rate": 8.582857142857144e-06, | |
| "loss": 0.0243, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "eval_loss": 0.6310169696807861, | |
| "eval_runtime": 482.8872, | |
| "eval_samples_per_second": 2.562, | |
| "eval_steps_per_second": 0.321, | |
| "eval_wer": 29.213729575611364, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "grad_norm": 2.3477253913879395, | |
| "learning_rate": 8.511428571428571e-06, | |
| "loss": 0.0215, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "grad_norm": 0.7934111952781677, | |
| "learning_rate": 8.44e-06, | |
| "loss": 0.019, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.18, | |
| "grad_norm": 1.1651448011398315, | |
| "learning_rate": 8.36857142857143e-06, | |
| "loss": 0.0096, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 6.32, | |
| "grad_norm": 1.3079229593276978, | |
| "learning_rate": 8.297142857142859e-06, | |
| "loss": 0.0117, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "grad_norm": 2.6112120151519775, | |
| "learning_rate": 8.225714285714288e-06, | |
| "loss": 0.0097, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 6.61, | |
| "grad_norm": 0.8934247493743896, | |
| "learning_rate": 8.154285714285715e-06, | |
| "loss": 0.0085, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 6.75, | |
| "grad_norm": 1.6696081161499023, | |
| "learning_rate": 8.082857142857144e-06, | |
| "loss": 0.011, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 1.9043184518814087, | |
| "learning_rate": 8.011428571428573e-06, | |
| "loss": 0.0159, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "grad_norm": 0.7788882255554199, | |
| "learning_rate": 7.94e-06, | |
| "loss": 0.0081, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 7.18, | |
| "grad_norm": 0.44921088218688965, | |
| "learning_rate": 7.86857142857143e-06, | |
| "loss": 0.0075, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 7.33, | |
| "grad_norm": 1.6392821073532104, | |
| "learning_rate": 7.797142857142858e-06, | |
| "loss": 0.0055, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 7.47, | |
| "grad_norm": 2.472712278366089, | |
| "learning_rate": 7.725714285714286e-06, | |
| "loss": 0.0059, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 7.61, | |
| "grad_norm": 0.16271165013313293, | |
| "learning_rate": 7.654285714285715e-06, | |
| "loss": 0.0046, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "grad_norm": 0.6483273506164551, | |
| "learning_rate": 7.5828571428571444e-06, | |
| "loss": 0.0044, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 1.0499032735824585, | |
| "learning_rate": 7.511428571428572e-06, | |
| "loss": 0.0046, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 8.05, | |
| "grad_norm": 0.6188054084777832, | |
| "learning_rate": 7.440000000000001e-06, | |
| "loss": 0.0039, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 8.19, | |
| "grad_norm": 0.11255456507205963, | |
| "learning_rate": 7.36857142857143e-06, | |
| "loss": 0.0035, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 0.21570441126823425, | |
| "learning_rate": 7.297142857142858e-06, | |
| "loss": 0.0044, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 8.48, | |
| "grad_norm": 0.2393558919429779, | |
| "learning_rate": 7.225714285714286e-06, | |
| "loss": 0.0025, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 8.62, | |
| "grad_norm": 0.8370733857154846, | |
| "learning_rate": 7.154285714285715e-06, | |
| "loss": 0.0022, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 8.76, | |
| "grad_norm": 0.25669318437576294, | |
| "learning_rate": 7.082857142857143e-06, | |
| "loss": 0.002, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 8.91, | |
| "grad_norm": 0.2188798040151596, | |
| "learning_rate": 7.011428571428572e-06, | |
| "loss": 0.002, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 9.05, | |
| "grad_norm": 0.06543828547000885, | |
| "learning_rate": 6.9400000000000005e-06, | |
| "loss": 0.0035, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.05953472480177879, | |
| "learning_rate": 6.868571428571429e-06, | |
| "loss": 0.0012, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 9.34, | |
| "grad_norm": 0.05562310665845871, | |
| "learning_rate": 6.797142857142858e-06, | |
| "loss": 0.0012, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 9.48, | |
| "grad_norm": 0.06505300104618073, | |
| "learning_rate": 6.725714285714287e-06, | |
| "loss": 0.0011, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 9.63, | |
| "grad_norm": 0.05420365184545517, | |
| "learning_rate": 6.654285714285716e-06, | |
| "loss": 0.001, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 9.77, | |
| "grad_norm": 0.0459221787750721, | |
| "learning_rate": 6.582857142857143e-06, | |
| "loss": 0.0009, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "grad_norm": 0.05463138595223427, | |
| "learning_rate": 6.511428571428572e-06, | |
| "loss": 0.0009, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 10.06, | |
| "grad_norm": 0.2885764539241791, | |
| "learning_rate": 6.440000000000001e-06, | |
| "loss": 0.0009, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.03790697455406189, | |
| "learning_rate": 6.368571428571429e-06, | |
| "loss": 0.0007, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 10.34, | |
| "grad_norm": 0.04002631828188896, | |
| "learning_rate": 6.297142857142857e-06, | |
| "loss": 0.0007, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 10.49, | |
| "grad_norm": 0.04201885685324669, | |
| "learning_rate": 6.225714285714286e-06, | |
| "loss": 0.0006, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 10.63, | |
| "grad_norm": 0.042038705199956894, | |
| "learning_rate": 6.1542857142857145e-06, | |
| "loss": 0.0007, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 10.78, | |
| "grad_norm": 0.037008922547101974, | |
| "learning_rate": 6.0828571428571435e-06, | |
| "loss": 0.0006, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 10.92, | |
| "grad_norm": 0.03480172157287598, | |
| "learning_rate": 6.011428571428572e-06, | |
| "loss": 0.0006, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 11.06, | |
| "grad_norm": 0.0292174331843853, | |
| "learning_rate": 5.94e-06, | |
| "loss": 0.0006, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 11.21, | |
| "grad_norm": 0.03173697739839554, | |
| "learning_rate": 5.868571428571429e-06, | |
| "loss": 0.0005, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 11.35, | |
| "grad_norm": 0.026889758184552193, | |
| "learning_rate": 5.797142857142858e-06, | |
| "loss": 0.0005, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "grad_norm": 0.02639606036245823, | |
| "learning_rate": 5.725714285714287e-06, | |
| "loss": 0.0005, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "eval_loss": 0.6665691137313843, | |
| "eval_runtime": 480.6597, | |
| "eval_samples_per_second": 2.574, | |
| "eval_steps_per_second": 0.322, | |
| "eval_wer": 27.557846255071826, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 11.64, | |
| "grad_norm": 0.029212338849902153, | |
| "learning_rate": 5.654285714285714e-06, | |
| "loss": 0.0005, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 11.78, | |
| "grad_norm": 0.025965554639697075, | |
| "learning_rate": 5.582857142857143e-06, | |
| "loss": 0.0005, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 11.93, | |
| "grad_norm": 0.0334208607673645, | |
| "learning_rate": 5.511428571428572e-06, | |
| "loss": 0.0005, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 12.07, | |
| "grad_norm": 0.022900965064764023, | |
| "learning_rate": 5.4400000000000004e-06, | |
| "loss": 0.0005, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 12.21, | |
| "grad_norm": 0.024590345099568367, | |
| "learning_rate": 5.368571428571429e-06, | |
| "loss": 0.0004, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 12.36, | |
| "grad_norm": 0.02628406509757042, | |
| "learning_rate": 5.297142857142858e-06, | |
| "loss": 0.0005, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.037630677223205566, | |
| "learning_rate": 5.225714285714286e-06, | |
| "loss": 0.0004, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 12.64, | |
| "grad_norm": 0.020544828847050667, | |
| "learning_rate": 5.154285714285715e-06, | |
| "loss": 0.0004, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 12.79, | |
| "grad_norm": 0.02015620656311512, | |
| "learning_rate": 5.082857142857144e-06, | |
| "loss": 0.0004, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 12.93, | |
| "grad_norm": 0.027246667072176933, | |
| "learning_rate": 5.011428571428571e-06, | |
| "loss": 0.0005, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 13.07, | |
| "grad_norm": 0.023458700627088547, | |
| "learning_rate": 4.94e-06, | |
| "loss": 0.0004, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 13.22, | |
| "grad_norm": 0.020197845995426178, | |
| "learning_rate": 4.868571428571429e-06, | |
| "loss": 0.0004, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 13.36, | |
| "grad_norm": 0.020962951704859734, | |
| "learning_rate": 4.797142857142857e-06, | |
| "loss": 0.0004, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 13.51, | |
| "grad_norm": 0.024561889469623566, | |
| "learning_rate": 4.725714285714286e-06, | |
| "loss": 0.0004, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 13.65, | |
| "grad_norm": 0.021130310371518135, | |
| "learning_rate": 4.6542857142857145e-06, | |
| "loss": 0.0004, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 13.79, | |
| "grad_norm": 0.01879502646625042, | |
| "learning_rate": 4.5828571428571435e-06, | |
| "loss": 0.0004, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 13.94, | |
| "grad_norm": 0.020792394876480103, | |
| "learning_rate": 4.511428571428572e-06, | |
| "loss": 0.0004, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 14.08, | |
| "grad_norm": 0.015595002099871635, | |
| "learning_rate": 4.440000000000001e-06, | |
| "loss": 0.0004, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 14.22, | |
| "grad_norm": 0.017891034483909607, | |
| "learning_rate": 4.368571428571429e-06, | |
| "loss": 0.0003, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 14.37, | |
| "grad_norm": 0.02275862917304039, | |
| "learning_rate": 4.297142857142858e-06, | |
| "loss": 0.0004, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 14.51, | |
| "grad_norm": 0.019310912117362022, | |
| "learning_rate": 4.225714285714286e-06, | |
| "loss": 0.0004, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 14.66, | |
| "grad_norm": 0.018501998856663704, | |
| "learning_rate": 4.154285714285714e-06, | |
| "loss": 0.0003, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "grad_norm": 0.024563327431678772, | |
| "learning_rate": 4.082857142857143e-06, | |
| "loss": 0.0004, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 14.94, | |
| "grad_norm": 0.015636250376701355, | |
| "learning_rate": 4.011428571428571e-06, | |
| "loss": 0.0003, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 15.09, | |
| "grad_norm": 0.017548007890582085, | |
| "learning_rate": 3.94e-06, | |
| "loss": 0.0003, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 15.23, | |
| "grad_norm": 0.016372185200452805, | |
| "learning_rate": 3.8685714285714286e-06, | |
| "loss": 0.0003, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 15.37, | |
| "grad_norm": 0.0156166460365057, | |
| "learning_rate": 3.7971428571428576e-06, | |
| "loss": 0.0003, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 15.52, | |
| "grad_norm": 0.017448540776968002, | |
| "learning_rate": 3.7257142857142857e-06, | |
| "loss": 0.0003, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 15.66, | |
| "grad_norm": 0.01800905540585518, | |
| "learning_rate": 3.6542857142857148e-06, | |
| "loss": 0.0003, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "grad_norm": 0.01949344575405121, | |
| "learning_rate": 3.582857142857143e-06, | |
| "loss": 0.0003, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 15.95, | |
| "grad_norm": 0.01464366540312767, | |
| "learning_rate": 3.511428571428572e-06, | |
| "loss": 0.0003, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 16.09, | |
| "grad_norm": 0.017990361899137497, | |
| "learning_rate": 3.44e-06, | |
| "loss": 0.0003, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 16.24, | |
| "grad_norm": 0.013816111721098423, | |
| "learning_rate": 3.3685714285714287e-06, | |
| "loss": 0.0003, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 16.38, | |
| "grad_norm": 0.014646317809820175, | |
| "learning_rate": 3.2971428571428577e-06, | |
| "loss": 0.0003, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 16.52, | |
| "grad_norm": 0.014701823703944683, | |
| "learning_rate": 3.225714285714286e-06, | |
| "loss": 0.0003, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "grad_norm": 0.016701674088835716, | |
| "learning_rate": 3.154285714285715e-06, | |
| "loss": 0.0003, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 16.81, | |
| "grad_norm": 0.015475871041417122, | |
| "learning_rate": 3.082857142857143e-06, | |
| "loss": 0.0003, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 16.95, | |
| "grad_norm": 0.013779123313724995, | |
| "learning_rate": 3.0114285714285716e-06, | |
| "loss": 0.0003, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 17.1, | |
| "grad_norm": 0.01210096850991249, | |
| "learning_rate": 2.9400000000000002e-06, | |
| "loss": 0.0003, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 17.24, | |
| "grad_norm": 0.01389345247298479, | |
| "learning_rate": 2.868571428571429e-06, | |
| "loss": 0.0003, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 17.24, | |
| "eval_loss": 0.6930021047592163, | |
| "eval_runtime": 480.9102, | |
| "eval_samples_per_second": 2.572, | |
| "eval_steps_per_second": 0.322, | |
| "eval_wer": 27.601710713894068, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 17.39, | |
| "grad_norm": 0.013713260181248188, | |
| "learning_rate": 2.797142857142857e-06, | |
| "loss": 0.0003, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 17.53, | |
| "grad_norm": 0.012472675181925297, | |
| "learning_rate": 2.725714285714286e-06, | |
| "loss": 0.0003, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 17.67, | |
| "grad_norm": 0.013508176431059837, | |
| "learning_rate": 2.654285714285714e-06, | |
| "loss": 0.0003, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 17.82, | |
| "grad_norm": 0.01630580611526966, | |
| "learning_rate": 2.582857142857143e-06, | |
| "loss": 0.0003, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 17.96, | |
| "grad_norm": 0.014032123610377312, | |
| "learning_rate": 2.5114285714285718e-06, | |
| "loss": 0.0003, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 18.1, | |
| "grad_norm": 0.014417425729334354, | |
| "learning_rate": 2.4400000000000004e-06, | |
| "loss": 0.0003, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 18.25, | |
| "grad_norm": 0.013058402575552464, | |
| "learning_rate": 2.3685714285714285e-06, | |
| "loss": 0.0003, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 18.39, | |
| "grad_norm": 0.011255180463194847, | |
| "learning_rate": 2.297142857142857e-06, | |
| "loss": 0.0003, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 18.53, | |
| "grad_norm": 0.011813076213002205, | |
| "learning_rate": 2.2257142857142857e-06, | |
| "loss": 0.0003, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 18.68, | |
| "grad_norm": 0.016077186912298203, | |
| "learning_rate": 2.1542857142857147e-06, | |
| "loss": 0.0003, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 18.82, | |
| "grad_norm": 0.012286370620131493, | |
| "learning_rate": 2.0828571428571433e-06, | |
| "loss": 0.0002, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 18.97, | |
| "grad_norm": 0.010851682163774967, | |
| "learning_rate": 2.0114285714285715e-06, | |
| "loss": 0.0003, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 19.11, | |
| "grad_norm": 0.012027677148580551, | |
| "learning_rate": 1.94e-06, | |
| "loss": 0.0002, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 19.25, | |
| "grad_norm": 0.012901864014565945, | |
| "learning_rate": 1.8685714285714289e-06, | |
| "loss": 0.0002, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 19.4, | |
| "grad_norm": 0.010309391655027866, | |
| "learning_rate": 1.7971428571428572e-06, | |
| "loss": 0.0002, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 19.54, | |
| "grad_norm": 0.012053144164383411, | |
| "learning_rate": 1.7257142857142858e-06, | |
| "loss": 0.0002, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 19.68, | |
| "grad_norm": 0.01089125033468008, | |
| "learning_rate": 1.6542857142857144e-06, | |
| "loss": 0.0002, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 19.83, | |
| "grad_norm": 0.011995966546237469, | |
| "learning_rate": 1.582857142857143e-06, | |
| "loss": 0.0002, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 19.97, | |
| "grad_norm": 0.01478640642017126, | |
| "learning_rate": 1.5114285714285714e-06, | |
| "loss": 0.0003, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 20.11, | |
| "grad_norm": 0.010389993898570538, | |
| "learning_rate": 1.44e-06, | |
| "loss": 0.0003, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 20.26, | |
| "grad_norm": 0.01525922492146492, | |
| "learning_rate": 1.3685714285714286e-06, | |
| "loss": 0.0002, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 20.4, | |
| "grad_norm": 0.01428759004920721, | |
| "learning_rate": 1.2971428571428574e-06, | |
| "loss": 0.0002, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 20.55, | |
| "grad_norm": 0.012515067122876644, | |
| "learning_rate": 1.2257142857142857e-06, | |
| "loss": 0.0002, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 20.69, | |
| "grad_norm": 0.01102293748408556, | |
| "learning_rate": 1.1542857142857143e-06, | |
| "loss": 0.0002, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 20.83, | |
| "grad_norm": 0.012940412387251854, | |
| "learning_rate": 1.082857142857143e-06, | |
| "loss": 0.0002, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 20.98, | |
| "grad_norm": 0.009889173321425915, | |
| "learning_rate": 1.0114285714285715e-06, | |
| "loss": 0.0002, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 21.12, | |
| "grad_norm": 0.011683104559779167, | |
| "learning_rate": 9.400000000000001e-07, | |
| "loss": 0.0002, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 21.26, | |
| "grad_norm": 0.012568894773721695, | |
| "learning_rate": 8.685714285714286e-07, | |
| "loss": 0.0002, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 21.41, | |
| "grad_norm": 0.01185997761785984, | |
| "learning_rate": 7.971428571428572e-07, | |
| "loss": 0.0002, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 21.55, | |
| "grad_norm": 0.01229412853717804, | |
| "learning_rate": 7.257142857142857e-07, | |
| "loss": 0.0002, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 21.7, | |
| "grad_norm": 0.013192495331168175, | |
| "learning_rate": 6.542857142857144e-07, | |
| "loss": 0.0002, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 21.84, | |
| "grad_norm": 0.010738629847764969, | |
| "learning_rate": 5.82857142857143e-07, | |
| "loss": 0.0002, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 21.98, | |
| "grad_norm": 0.010789092630147934, | |
| "learning_rate": 5.114285714285714e-07, | |
| "loss": 0.0002, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 22.13, | |
| "grad_norm": 0.011501124128699303, | |
| "learning_rate": 4.4e-07, | |
| "loss": 0.0002, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 22.27, | |
| "grad_norm": 0.01108501572161913, | |
| "learning_rate": 3.685714285714286e-07, | |
| "loss": 0.0002, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 22.41, | |
| "grad_norm": 0.0128969457000494, | |
| "learning_rate": 2.9714285714285715e-07, | |
| "loss": 0.0002, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 22.56, | |
| "grad_norm": 0.014468998648226261, | |
| "learning_rate": 2.2571428571428574e-07, | |
| "loss": 0.0002, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 22.7, | |
| "grad_norm": 0.013158345595002174, | |
| "learning_rate": 1.542857142857143e-07, | |
| "loss": 0.0002, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 22.84, | |
| "grad_norm": 0.010379764251410961, | |
| "learning_rate": 8.285714285714285e-08, | |
| "loss": 0.0002, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 22.99, | |
| "grad_norm": 0.008165850304067135, | |
| "learning_rate": 1.142857142857143e-08, | |
| "loss": 0.0002, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 22.99, | |
| "eval_loss": 0.7029648423194885, | |
| "eval_runtime": 485.2218, | |
| "eval_samples_per_second": 2.549, | |
| "eval_steps_per_second": 0.319, | |
| "eval_wer": 27.579778484482947, | |
| "step": 4000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 4000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 23, | |
| "save_steps": 1000, | |
| "total_flos": 1.839962795360256e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |