| { | |
| "best_metric": 2.9224686264397457, | |
| "best_model_checkpoint": "./whisper-small-final-v2/checkpoint-5000", | |
| "epoch": 16.181229773462782, | |
| "eval_steps": 500, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0023542253766208887, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.0001, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.003053858410567045, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.0001, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.007697729859501123, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.0001, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.01041096169501543, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.0023387186229228973, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0036297033075243235, | |
| "learning_rate": 3e-06, | |
| "loss": 0.0001, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0016004816861823201, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.0035209895577281713, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.37100252509117126, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.0015, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.2593393325805664, | |
| "learning_rate": 5e-06, | |
| "loss": 0.003, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 1.1273244619369507, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.003, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.11678028106689453, | |
| "learning_rate": 6e-06, | |
| "loss": 0.0014, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 2.602248191833496, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.0025, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 1.053270697593689, | |
| "learning_rate": 7e-06, | |
| "loss": 0.0021, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 0.25527653098106384, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.0037, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 0.208439901471138, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.002, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.14851850271224976, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.0034, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.6750054359436035, | |
| "learning_rate": 9e-06, | |
| "loss": 0.0046, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 4.503875255584717, | |
| "learning_rate": 9.48e-06, | |
| "loss": 0.0093, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.6791089177131653, | |
| "learning_rate": 9.980000000000001e-06, | |
| "loss": 0.0073, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "eval_loss": 0.024947550147771835, | |
| "eval_runtime": 794.9124, | |
| "eval_samples_per_second": 1.551, | |
| "eval_steps_per_second": 0.195, | |
| "eval_wer": 5.002578648788035, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.7141709923744202, | |
| "learning_rate": 9.931428571428571e-06, | |
| "loss": 0.0022, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 1.1480615139007568, | |
| "learning_rate": 9.86e-06, | |
| "loss": 0.0036, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 1.9711138010025024, | |
| "learning_rate": 9.78857142857143e-06, | |
| "loss": 0.0043, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 1.0525217056274414, | |
| "learning_rate": 9.717142857142858e-06, | |
| "loss": 0.0051, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.15985222160816193, | |
| "learning_rate": 9.645714285714286e-06, | |
| "loss": 0.0057, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.8261871933937073, | |
| "learning_rate": 9.574285714285715e-06, | |
| "loss": 0.0023, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.04620926454663277, | |
| "learning_rate": 9.502857142857144e-06, | |
| "loss": 0.0019, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.3023536503314972, | |
| "learning_rate": 9.431428571428573e-06, | |
| "loss": 0.0012, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.11112522333860397, | |
| "learning_rate": 9.360000000000002e-06, | |
| "loss": 0.0024, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 1.6971714496612549, | |
| "learning_rate": 9.28857142857143e-06, | |
| "loss": 0.0031, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 0.22176721692085266, | |
| "learning_rate": 9.217142857142858e-06, | |
| "loss": 0.0026, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.2839481830596924, | |
| "learning_rate": 9.145714285714287e-06, | |
| "loss": 0.0016, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.6732001900672913, | |
| "learning_rate": 9.074285714285716e-06, | |
| "loss": 0.0044, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 2.15289044380188, | |
| "learning_rate": 9.002857142857144e-06, | |
| "loss": 0.0028, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "grad_norm": 0.4701692759990692, | |
| "learning_rate": 8.931428571428573e-06, | |
| "loss": 0.0009, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 0.38268008828163147, | |
| "learning_rate": 8.860000000000002e-06, | |
| "loss": 0.0012, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "grad_norm": 0.057314854115247726, | |
| "learning_rate": 8.788571428571429e-06, | |
| "loss": 0.001, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "grad_norm": 0.048205479979515076, | |
| "learning_rate": 8.717142857142858e-06, | |
| "loss": 0.0022, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 0.008282452821731567, | |
| "learning_rate": 8.645714285714287e-06, | |
| "loss": 0.0006, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "grad_norm": 1.3108314275741577, | |
| "learning_rate": 8.574285714285714e-06, | |
| "loss": 0.0014, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.24, | |
| "eval_loss": 0.02143564447760582, | |
| "eval_runtime": 802.0669, | |
| "eval_samples_per_second": 1.537, | |
| "eval_steps_per_second": 0.193, | |
| "eval_wer": 4.108647068935878, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.32, | |
| "grad_norm": 0.888531506061554, | |
| "learning_rate": 8.502857142857143e-06, | |
| "loss": 0.0025, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.013467811048030853, | |
| "learning_rate": 8.431428571428572e-06, | |
| "loss": 0.0009, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "grad_norm": 1.272592306137085, | |
| "learning_rate": 8.36e-06, | |
| "loss": 0.001, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 0.022731788456439972, | |
| "learning_rate": 8.288571428571429e-06, | |
| "loss": 0.0002, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.64, | |
| "grad_norm": 0.49200737476348877, | |
| "learning_rate": 8.217142857142858e-06, | |
| "loss": 0.0011, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 3.72, | |
| "grad_norm": 0.011444230563938618, | |
| "learning_rate": 8.145714285714287e-06, | |
| "loss": 0.0005, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.02571061998605728, | |
| "learning_rate": 8.074285714285714e-06, | |
| "loss": 0.0002, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "grad_norm": 0.02670286037027836, | |
| "learning_rate": 8.002857142857143e-06, | |
| "loss": 0.0005, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 0.04544826224446297, | |
| "learning_rate": 7.931428571428572e-06, | |
| "loss": 0.0007, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 0.028520189225673676, | |
| "learning_rate": 7.860000000000001e-06, | |
| "loss": 0.0003, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 4.13, | |
| "grad_norm": 0.017675643786787987, | |
| "learning_rate": 7.788571428571428e-06, | |
| "loss": 0.0006, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 4.21, | |
| "grad_norm": 0.16999584436416626, | |
| "learning_rate": 7.717142857142857e-06, | |
| "loss": 0.0005, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.29, | |
| "grad_norm": 0.22153036296367645, | |
| "learning_rate": 7.645714285714286e-06, | |
| "loss": 0.0004, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 4.37, | |
| "grad_norm": 0.0049127149395644665, | |
| "learning_rate": 7.574285714285715e-06, | |
| "loss": 0.0001, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 0.02039247751235962, | |
| "learning_rate": 7.502857142857144e-06, | |
| "loss": 0.0003, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 4.53, | |
| "grad_norm": 0.009562624618411064, | |
| "learning_rate": 7.431428571428572e-06, | |
| "loss": 0.0002, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.61, | |
| "grad_norm": 0.019309034571051598, | |
| "learning_rate": 7.360000000000001e-06, | |
| "loss": 0.0001, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 0.004860865883529186, | |
| "learning_rate": 7.28857142857143e-06, | |
| "loss": 0.0008, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 4.77, | |
| "grad_norm": 0.005680860485881567, | |
| "learning_rate": 7.217142857142858e-06, | |
| "loss": 0.0004, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 0.02605484426021576, | |
| "learning_rate": 7.145714285714286e-06, | |
| "loss": 0.0008, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "eval_loss": 0.022091694176197052, | |
| "eval_runtime": 791.3534, | |
| "eval_samples_per_second": 1.558, | |
| "eval_steps_per_second": 0.196, | |
| "eval_wer": 3.988310125494241, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.94, | |
| "grad_norm": 1.7420192956924438, | |
| "learning_rate": 7.074285714285715e-06, | |
| "loss": 0.0005, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "grad_norm": 0.15281762182712555, | |
| "learning_rate": 7.002857142857143e-06, | |
| "loss": 0.0004, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.013328706845641136, | |
| "learning_rate": 6.931428571428572e-06, | |
| "loss": 0.0002, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 5.18, | |
| "grad_norm": 0.23917707800865173, | |
| "learning_rate": 6.860000000000001e-06, | |
| "loss": 0.0019, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "grad_norm": 0.02083628997206688, | |
| "learning_rate": 6.7885714285714286e-06, | |
| "loss": 0.0003, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "grad_norm": 0.7623525857925415, | |
| "learning_rate": 6.7171428571428576e-06, | |
| "loss": 0.0006, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "grad_norm": 1.761312484741211, | |
| "learning_rate": 6.645714285714287e-06, | |
| "loss": 0.0005, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.02908872812986374, | |
| "learning_rate": 6.574285714285716e-06, | |
| "loss": 0.0003, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.58, | |
| "grad_norm": 0.0062005845829844475, | |
| "learning_rate": 6.502857142857143e-06, | |
| "loss": 0.0003, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "grad_norm": 0.009553766809403896, | |
| "learning_rate": 6.431428571428572e-06, | |
| "loss": 0.0001, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 5.74, | |
| "grad_norm": 0.006461838725954294, | |
| "learning_rate": 6.360000000000001e-06, | |
| "loss": 0.0001, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 0.006093865260481834, | |
| "learning_rate": 6.288571428571429e-06, | |
| "loss": 0.0003, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.91, | |
| "grad_norm": 0.006265025120228529, | |
| "learning_rate": 6.217142857142857e-06, | |
| "loss": 0.0005, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 5.99, | |
| "grad_norm": 0.08442416042089462, | |
| "learning_rate": 6.145714285714286e-06, | |
| "loss": 0.0002, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "grad_norm": 0.005628472194075584, | |
| "learning_rate": 6.0742857142857145e-06, | |
| "loss": 0.0001, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "grad_norm": 0.0012941138120368123, | |
| "learning_rate": 6.0028571428571435e-06, | |
| "loss": 0.0001, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "grad_norm": 0.003825143212452531, | |
| "learning_rate": 5.9314285714285725e-06, | |
| "loss": 0.0001, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "grad_norm": 0.002939299214631319, | |
| "learning_rate": 5.86e-06, | |
| "loss": 0.0, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 6.39, | |
| "grad_norm": 0.0030452311038970947, | |
| "learning_rate": 5.788571428571429e-06, | |
| "loss": 0.0001, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "grad_norm": 0.017602646723389626, | |
| "learning_rate": 5.717142857142858e-06, | |
| "loss": 0.0, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "eval_loss": 0.018006248399615288, | |
| "eval_runtime": 791.4436, | |
| "eval_samples_per_second": 1.558, | |
| "eval_steps_per_second": 0.196, | |
| "eval_wer": 2.974041602200447, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.55, | |
| "grad_norm": 0.09079013019800186, | |
| "learning_rate": 5.645714285714287e-06, | |
| "loss": 0.0001, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 6.63, | |
| "grad_norm": 0.011517921462655067, | |
| "learning_rate": 5.574285714285714e-06, | |
| "loss": 0.0001, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "grad_norm": 0.0036222347989678383, | |
| "learning_rate": 5.502857142857143e-06, | |
| "loss": 0.0, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.09321127831935883, | |
| "learning_rate": 5.431428571428572e-06, | |
| "loss": 0.0, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 0.019542288035154343, | |
| "learning_rate": 5.36e-06, | |
| "loss": 0.0, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "grad_norm": 0.010700986720621586, | |
| "learning_rate": 5.2885714285714285e-06, | |
| "loss": 0.0001, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "grad_norm": 0.005260708276182413, | |
| "learning_rate": 5.2171428571428575e-06, | |
| "loss": 0.0, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "grad_norm": 0.0018403552239760756, | |
| "learning_rate": 5.145714285714286e-06, | |
| "loss": 0.0, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.00291392276994884, | |
| "learning_rate": 5.074285714285715e-06, | |
| "loss": 0.0, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "grad_norm": 0.001351356622762978, | |
| "learning_rate": 5.002857142857144e-06, | |
| "loss": 0.0, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "grad_norm": 0.002212280174717307, | |
| "learning_rate": 4.931428571428572e-06, | |
| "loss": 0.0, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 7.44, | |
| "grad_norm": 0.0013704759767279029, | |
| "learning_rate": 4.86e-06, | |
| "loss": 0.0, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.52, | |
| "grad_norm": 0.004209516104310751, | |
| "learning_rate": 4.788571428571429e-06, | |
| "loss": 0.0, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 7.61, | |
| "grad_norm": 0.004072306212037802, | |
| "learning_rate": 4.717142857142857e-06, | |
| "loss": 0.0, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "grad_norm": 0.0258013978600502, | |
| "learning_rate": 4.645714285714286e-06, | |
| "loss": 0.0, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 7.77, | |
| "grad_norm": 0.0018390474142506719, | |
| "learning_rate": 4.574285714285714e-06, | |
| "loss": 0.0, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.85, | |
| "grad_norm": 0.002616502810269594, | |
| "learning_rate": 4.5028571428571434e-06, | |
| "loss": 0.0, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 7.93, | |
| "grad_norm": 0.0016982157249003649, | |
| "learning_rate": 4.431428571428572e-06, | |
| "loss": 0.0, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 8.01, | |
| "grad_norm": 0.0023189985658973455, | |
| "learning_rate": 4.360000000000001e-06, | |
| "loss": 0.0, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 8.09, | |
| "grad_norm": 0.003353666979819536, | |
| "learning_rate": 4.288571428571429e-06, | |
| "loss": 0.0, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.09, | |
| "eval_loss": 0.017685044556856155, | |
| "eval_runtime": 798.3523, | |
| "eval_samples_per_second": 1.544, | |
| "eval_steps_per_second": 0.194, | |
| "eval_wer": 3.094378545642084, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 8.17, | |
| "grad_norm": 0.0012944688787683845, | |
| "learning_rate": 4.217142857142858e-06, | |
| "loss": 0.0, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "grad_norm": 0.002697590971365571, | |
| "learning_rate": 4.145714285714286e-06, | |
| "loss": 0.0, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 0.0015665855025872588, | |
| "learning_rate": 4.074285714285714e-06, | |
| "loss": 0.0, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "grad_norm": 0.0023723498452454805, | |
| "learning_rate": 4.002857142857143e-06, | |
| "loss": 0.0, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.0024561272002756596, | |
| "learning_rate": 3.931428571428571e-06, | |
| "loss": 0.0, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 8.58, | |
| "grad_norm": 0.0012809446780011058, | |
| "learning_rate": 3.86e-06, | |
| "loss": 0.0, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 8.66, | |
| "grad_norm": 0.002741657430306077, | |
| "learning_rate": 3.7885714285714285e-06, | |
| "loss": 0.0, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 8.74, | |
| "grad_norm": 0.0022798278369009495, | |
| "learning_rate": 3.7171428571428575e-06, | |
| "loss": 0.0, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 8.82, | |
| "grad_norm": 0.0008324767695739865, | |
| "learning_rate": 3.6457142857142857e-06, | |
| "loss": 0.0, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.0021829677280038595, | |
| "learning_rate": 3.5742857142857147e-06, | |
| "loss": 0.0, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 8.98, | |
| "grad_norm": 0.0010186375584453344, | |
| "learning_rate": 3.5028571428571433e-06, | |
| "loss": 0.0, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 9.06, | |
| "grad_norm": 0.0021447737235575914, | |
| "learning_rate": 3.431428571428572e-06, | |
| "loss": 0.0, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 9.14, | |
| "grad_norm": 0.0014905674615874887, | |
| "learning_rate": 3.3600000000000004e-06, | |
| "loss": 0.0, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 9.22, | |
| "grad_norm": 0.0023917625658214092, | |
| "learning_rate": 3.2885714285714286e-06, | |
| "loss": 0.0, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.0014455171767622232, | |
| "learning_rate": 3.2171428571428576e-06, | |
| "loss": 0.0, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 9.39, | |
| "grad_norm": 0.009605071507394314, | |
| "learning_rate": 3.1457142857142858e-06, | |
| "loss": 0.0, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 9.47, | |
| "grad_norm": 0.0006564524956047535, | |
| "learning_rate": 3.074285714285715e-06, | |
| "loss": 0.0, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 9.55, | |
| "grad_norm": 0.0008893849444575608, | |
| "learning_rate": 3.002857142857143e-06, | |
| "loss": 0.0, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 9.63, | |
| "grad_norm": 0.001537854433991015, | |
| "learning_rate": 2.9314285714285716e-06, | |
| "loss": 0.0, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 9.71, | |
| "grad_norm": 0.0020573874935507774, | |
| "learning_rate": 2.86e-06, | |
| "loss": 0.0, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.71, | |
| "eval_loss": 0.017799125984311104, | |
| "eval_runtime": 794.7434, | |
| "eval_samples_per_second": 1.551, | |
| "eval_steps_per_second": 0.195, | |
| "eval_wer": 3.094378545642084, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 9.79, | |
| "grad_norm": 0.001788494409993291, | |
| "learning_rate": 2.7885714285714287e-06, | |
| "loss": 0.0, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 9.87, | |
| "grad_norm": 0.0009725405252538621, | |
| "learning_rate": 2.7171428571428577e-06, | |
| "loss": 0.0, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 9.95, | |
| "grad_norm": 0.001284493599087, | |
| "learning_rate": 2.645714285714286e-06, | |
| "loss": 0.0, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 10.03, | |
| "grad_norm": 0.0017106755403801799, | |
| "learning_rate": 2.574285714285715e-06, | |
| "loss": 0.0, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 10.11, | |
| "grad_norm": 0.0007921723299659789, | |
| "learning_rate": 2.502857142857143e-06, | |
| "loss": 0.0, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 10.19, | |
| "grad_norm": 0.0018981642788276076, | |
| "learning_rate": 2.4314285714285717e-06, | |
| "loss": 0.0, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 10.28, | |
| "grad_norm": 0.0014530618209391832, | |
| "learning_rate": 2.3600000000000003e-06, | |
| "loss": 0.0, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 10.36, | |
| "grad_norm": 0.0011344915255904198, | |
| "learning_rate": 2.288571428571429e-06, | |
| "loss": 0.0, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 10.44, | |
| "grad_norm": 0.0006109612877480686, | |
| "learning_rate": 2.2171428571428575e-06, | |
| "loss": 0.0, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 10.52, | |
| "grad_norm": 0.0011850837618112564, | |
| "learning_rate": 2.145714285714286e-06, | |
| "loss": 0.0, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.0005283011123538017, | |
| "learning_rate": 2.0742857142857146e-06, | |
| "loss": 0.0, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 10.68, | |
| "grad_norm": 0.0010328389471396804, | |
| "learning_rate": 2.0028571428571432e-06, | |
| "loss": 0.0, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 10.76, | |
| "grad_norm": 0.0009599952027201653, | |
| "learning_rate": 1.9314285714285714e-06, | |
| "loss": 0.0, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 10.84, | |
| "grad_norm": 0.0011755761224776506, | |
| "learning_rate": 1.8600000000000002e-06, | |
| "loss": 0.0, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 10.92, | |
| "grad_norm": 0.0018132536206394434, | |
| "learning_rate": 1.7885714285714288e-06, | |
| "loss": 0.0, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.0018807642627507448, | |
| "learning_rate": 1.7171428571428572e-06, | |
| "loss": 0.0, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 11.08, | |
| "grad_norm": 0.003156110644340515, | |
| "learning_rate": 1.6457142857142857e-06, | |
| "loss": 0.0, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 11.17, | |
| "grad_norm": 0.0010045066010206938, | |
| "learning_rate": 1.5742857142857143e-06, | |
| "loss": 0.0, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 11.25, | |
| "grad_norm": 0.0008904847782105207, | |
| "learning_rate": 1.502857142857143e-06, | |
| "loss": 0.0, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 11.33, | |
| "grad_norm": 0.0013107324484735727, | |
| "learning_rate": 1.4314285714285717e-06, | |
| "loss": 0.0, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 11.33, | |
| "eval_loss": 0.017873059958219528, | |
| "eval_runtime": 799.0063, | |
| "eval_samples_per_second": 1.543, | |
| "eval_steps_per_second": 0.194, | |
| "eval_wer": 3.1287605294825513, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 11.41, | |
| "grad_norm": 0.0013505747774615884, | |
| "learning_rate": 1.3600000000000001e-06, | |
| "loss": 0.0, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 11.49, | |
| "grad_norm": 0.0013679265975952148, | |
| "learning_rate": 1.2885714285714287e-06, | |
| "loss": 0.0, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 11.57, | |
| "grad_norm": 0.0021877065300941467, | |
| "learning_rate": 1.2171428571428573e-06, | |
| "loss": 0.0, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 11.65, | |
| "grad_norm": 0.0011414324399083853, | |
| "learning_rate": 1.1457142857142859e-06, | |
| "loss": 0.0, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 11.73, | |
| "grad_norm": 0.0008932927739806473, | |
| "learning_rate": 1.0742857142857145e-06, | |
| "loss": 0.0, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 11.81, | |
| "grad_norm": 0.0016671326011419296, | |
| "learning_rate": 1.0028571428571428e-06, | |
| "loss": 0.0, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 11.89, | |
| "grad_norm": 0.0011023151455447078, | |
| "learning_rate": 9.314285714285714e-07, | |
| "loss": 0.0, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 11.97, | |
| "grad_norm": 0.0008648216025903821, | |
| "learning_rate": 8.6e-07, | |
| "loss": 0.0, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 12.06, | |
| "grad_norm": 0.003709783311933279, | |
| "learning_rate": 7.885714285714287e-07, | |
| "loss": 0.0, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 12.14, | |
| "grad_norm": 0.0005944801960140467, | |
| "learning_rate": 7.171428571428572e-07, | |
| "loss": 0.0, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 12.22, | |
| "grad_norm": 0.0010830622632056475, | |
| "learning_rate": 6.457142857142858e-07, | |
| "loss": 0.0, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.0018471087096258998, | |
| "learning_rate": 5.742857142857143e-07, | |
| "loss": 0.0, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 12.38, | |
| "grad_norm": 0.0018315908964723349, | |
| "learning_rate": 5.028571428571429e-07, | |
| "loss": 0.0, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 12.46, | |
| "grad_norm": 0.0010556795168668032, | |
| "learning_rate": 4.3142857142857146e-07, | |
| "loss": 0.0, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 12.54, | |
| "grad_norm": 0.001286807470023632, | |
| "learning_rate": 3.6e-07, | |
| "loss": 0.0, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 12.62, | |
| "grad_norm": 0.0025280534755438566, | |
| "learning_rate": 2.885714285714286e-07, | |
| "loss": 0.0, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.0008814557804726064, | |
| "learning_rate": 2.1714285714285715e-07, | |
| "loss": 0.0, | |
| "step": 3925 | |
| }, | |
| { | |
| "epoch": 12.78, | |
| "grad_norm": 0.0007184511632658541, | |
| "learning_rate": 1.4571428571428574e-07, | |
| "loss": 0.0, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 12.86, | |
| "grad_norm": 0.0024200354237109423, | |
| "learning_rate": 7.428571428571429e-08, | |
| "loss": 0.0, | |
| "step": 3975 | |
| }, | |
| { | |
| "epoch": 12.94, | |
| "grad_norm": 0.0013352702371776104, | |
| "learning_rate": 2.8571428571428576e-09, | |
| "loss": 0.0, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 12.94, | |
| "eval_loss": 0.017900340259075165, | |
| "eval_runtime": 798.968, | |
| "eval_samples_per_second": 1.543, | |
| "eval_steps_per_second": 0.194, | |
| "eval_wer": 3.1287605294825513, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 13.03, | |
| "grad_norm": 0.0006061609601601958, | |
| "learning_rate": 5.301333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4025 | |
| }, | |
| { | |
| "epoch": 13.11, | |
| "grad_norm": 0.000903214851859957, | |
| "learning_rate": 5.268000000000001e-06, | |
| "loss": 0.0, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 13.19, | |
| "grad_norm": 0.0009095696732401848, | |
| "learning_rate": 5.234666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4075 | |
| }, | |
| { | |
| "epoch": 13.27, | |
| "grad_norm": 0.0009569237590767443, | |
| "learning_rate": 5.201333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 13.35, | |
| "grad_norm": 0.0013346378691494465, | |
| "learning_rate": 5.168000000000001e-06, | |
| "loss": 0.0, | |
| "step": 4125 | |
| }, | |
| { | |
| "epoch": 13.43, | |
| "grad_norm": 0.0007105122786015272, | |
| "learning_rate": 5.134666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 13.51, | |
| "grad_norm": 0.001324965967796743, | |
| "learning_rate": 5.101333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4175 | |
| }, | |
| { | |
| "epoch": 13.59, | |
| "grad_norm": 0.002936706179752946, | |
| "learning_rate": 5.0680000000000004e-06, | |
| "loss": 0.0, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 13.67, | |
| "grad_norm": 0.0006480725132860243, | |
| "learning_rate": 5.034666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4225 | |
| }, | |
| { | |
| "epoch": 13.75, | |
| "grad_norm": 0.0008248516824096441, | |
| "learning_rate": 5.001333333333333e-06, | |
| "loss": 0.0, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 13.83, | |
| "grad_norm": 0.000976664014160633, | |
| "learning_rate": 4.9680000000000005e-06, | |
| "loss": 0.0, | |
| "step": 4275 | |
| }, | |
| { | |
| "epoch": 13.92, | |
| "grad_norm": 0.0007901455392129719, | |
| "learning_rate": 4.934666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.0008691786788403988, | |
| "learning_rate": 4.901333333333333e-06, | |
| "loss": 0.0, | |
| "step": 4325 | |
| }, | |
| { | |
| "epoch": 14.08, | |
| "grad_norm": 0.00039108938653953373, | |
| "learning_rate": 4.868000000000001e-06, | |
| "loss": 0.0, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 14.16, | |
| "grad_norm": 0.000672497262712568, | |
| "learning_rate": 4.834666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4375 | |
| }, | |
| { | |
| "epoch": 14.24, | |
| "grad_norm": 0.0006290368619374931, | |
| "learning_rate": 4.801333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 14.32, | |
| "grad_norm": 0.0012541558826342225, | |
| "learning_rate": 4.768000000000001e-06, | |
| "loss": 0.0, | |
| "step": 4425 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "grad_norm": 0.0015149698592722416, | |
| "learning_rate": 4.734666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 14.48, | |
| "grad_norm": 0.001397727057337761, | |
| "learning_rate": 4.701333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4475 | |
| }, | |
| { | |
| "epoch": 14.56, | |
| "grad_norm": 0.001000353367999196, | |
| "learning_rate": 4.668e-06, | |
| "loss": 0.0, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 14.56, | |
| "eval_loss": 0.018149884417653084, | |
| "eval_runtime": 806.023, | |
| "eval_samples_per_second": 1.53, | |
| "eval_steps_per_second": 0.192, | |
| "eval_wer": 2.888086642599278, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 14.64, | |
| "grad_norm": 0.0004112539754714817, | |
| "learning_rate": 4.634666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4525 | |
| }, | |
| { | |
| "epoch": 14.72, | |
| "grad_norm": 0.00042669521644711494, | |
| "learning_rate": 4.601333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 14.81, | |
| "grad_norm": 0.0005972330691292882, | |
| "learning_rate": 4.568e-06, | |
| "loss": 0.0, | |
| "step": 4575 | |
| }, | |
| { | |
| "epoch": 14.89, | |
| "grad_norm": 0.000703622295986861, | |
| "learning_rate": 4.534666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 14.97, | |
| "grad_norm": 0.0008054905338212848, | |
| "learning_rate": 4.501333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4625 | |
| }, | |
| { | |
| "epoch": 15.05, | |
| "grad_norm": 0.0006107372464612126, | |
| "learning_rate": 4.468e-06, | |
| "loss": 0.0, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 15.13, | |
| "grad_norm": 0.0011443241965025663, | |
| "learning_rate": 4.4346666666666675e-06, | |
| "loss": 0.0, | |
| "step": 4675 | |
| }, | |
| { | |
| "epoch": 15.21, | |
| "grad_norm": 0.0007337784627452493, | |
| "learning_rate": 4.401333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 15.29, | |
| "grad_norm": 0.0005867818254046142, | |
| "learning_rate": 4.368e-06, | |
| "loss": 0.0, | |
| "step": 4725 | |
| }, | |
| { | |
| "epoch": 15.37, | |
| "grad_norm": 0.00047605938743799925, | |
| "learning_rate": 4.334666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 15.45, | |
| "grad_norm": 0.0007535215700045228, | |
| "learning_rate": 4.301333333333333e-06, | |
| "loss": 0.0, | |
| "step": 4775 | |
| }, | |
| { | |
| "epoch": 15.53, | |
| "grad_norm": 0.0004895786987617612, | |
| "learning_rate": 4.2680000000000005e-06, | |
| "loss": 0.0, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 15.61, | |
| "grad_norm": 0.0005502207204699516, | |
| "learning_rate": 4.234666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4825 | |
| }, | |
| { | |
| "epoch": 15.7, | |
| "grad_norm": 0.0008273226558230817, | |
| "learning_rate": 4.201333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 15.78, | |
| "grad_norm": 0.0008853155304677784, | |
| "learning_rate": 4.168000000000001e-06, | |
| "loss": 0.0, | |
| "step": 4875 | |
| }, | |
| { | |
| "epoch": 15.86, | |
| "grad_norm": 0.0008551743812859058, | |
| "learning_rate": 4.134666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 15.94, | |
| "grad_norm": 0.0012048580683767796, | |
| "learning_rate": 4.1013333333333334e-06, | |
| "loss": 0.0, | |
| "step": 4925 | |
| }, | |
| { | |
| "epoch": 16.02, | |
| "grad_norm": 0.00043612357694655657, | |
| "learning_rate": 4.068000000000001e-06, | |
| "loss": 0.0, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 16.1, | |
| "grad_norm": 0.0005169506184756756, | |
| "learning_rate": 4.034666666666667e-06, | |
| "loss": 0.0, | |
| "step": 4975 | |
| }, | |
| { | |
| "epoch": 16.18, | |
| "grad_norm": 0.000585487752687186, | |
| "learning_rate": 4.0013333333333335e-06, | |
| "loss": 0.0, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 16.18, | |
| "eval_loss": 0.018417511135339737, | |
| "eval_runtime": 808.2166, | |
| "eval_samples_per_second": 1.526, | |
| "eval_steps_per_second": 0.192, | |
| "eval_wer": 2.9224686264397457, | |
| "step": 5000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 8000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 26, | |
| "save_steps": 1000, | |
| "total_flos": 2.303142362873856e+19, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |