| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 400.0, | |
| "global_step": 8800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 4.55, | |
| "learning_rate": 3.7125e-06, | |
| "loss": 16.0016, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 9.09, | |
| "learning_rate": 7.4625e-06, | |
| "loss": 8.8896, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 13.64, | |
| "learning_rate": 1.1212499999999998e-05, | |
| "loss": 5.8994, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 18.18, | |
| "learning_rate": 1.49625e-05, | |
| "loss": 4.9866, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 22.73, | |
| "learning_rate": 1.8712499999999997e-05, | |
| "loss": 4.2706, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 22.73, | |
| "eval_loss": 4.017421722412109, | |
| "eval_runtime": 13.7125, | |
| "eval_samples_per_second": 22.315, | |
| "eval_steps_per_second": 2.844, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 27.27, | |
| "learning_rate": 2.2462499999999997e-05, | |
| "loss": 3.7391, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 31.82, | |
| "learning_rate": 2.6212499999999997e-05, | |
| "loss": 3.4669, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 36.36, | |
| "learning_rate": 2.99625e-05, | |
| "loss": 3.3769, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 40.91, | |
| "learning_rate": 3.37125e-05, | |
| "loss": 3.304, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 45.45, | |
| "learning_rate": 3.7462499999999996e-05, | |
| "loss": 3.2492, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 45.45, | |
| "eval_loss": 3.2308928966522217, | |
| "eval_runtime": 13.6766, | |
| "eval_samples_per_second": 22.374, | |
| "eval_steps_per_second": 2.852, | |
| "eval_wer": 0.9907741676694746, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 4.12125e-05, | |
| "loss": 3.1542, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 54.55, | |
| "learning_rate": 4.4962499999999995e-05, | |
| "loss": 2.9723, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 59.09, | |
| "learning_rate": 4.871249999999999e-05, | |
| "loss": 2.6187, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 63.64, | |
| "learning_rate": 5.2462499999999994e-05, | |
| "loss": 2.2201, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 68.18, | |
| "learning_rate": 5.62125e-05, | |
| "loss": 1.9709, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 68.18, | |
| "eval_loss": 1.0651201009750366, | |
| "eval_runtime": 13.4664, | |
| "eval_samples_per_second": 22.723, | |
| "eval_steps_per_second": 2.896, | |
| "eval_wer": 0.8439630966706779, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 72.73, | |
| "learning_rate": 5.9962499999999994e-05, | |
| "loss": 1.7822, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 77.27, | |
| "learning_rate": 6.37125e-05, | |
| "loss": 1.6511, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 81.82, | |
| "learning_rate": 6.746249999999999e-05, | |
| "loss": 1.5519, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 86.36, | |
| "learning_rate": 7.121249999999999e-05, | |
| "loss": 1.4612, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 90.91, | |
| "learning_rate": 7.49625e-05, | |
| "loss": 1.4088, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 90.91, | |
| "eval_loss": 0.576468288898468, | |
| "eval_runtime": 13.6104, | |
| "eval_samples_per_second": 22.483, | |
| "eval_steps_per_second": 2.865, | |
| "eval_wer": 0.6550340954673085, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 95.45, | |
| "learning_rate": 7.390808823529411e-05, | |
| "loss": 1.3302, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 7.280514705882352e-05, | |
| "loss": 1.2739, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 104.55, | |
| "learning_rate": 7.171323529411764e-05, | |
| "loss": 1.2181, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 109.09, | |
| "learning_rate": 7.062132352941176e-05, | |
| "loss": 1.1776, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 113.64, | |
| "learning_rate": 6.951838235294117e-05, | |
| "loss": 1.1326, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 113.64, | |
| "eval_loss": 0.48421531915664673, | |
| "eval_runtime": 13.6288, | |
| "eval_samples_per_second": 22.452, | |
| "eval_steps_per_second": 2.862, | |
| "eval_wer": 0.5760128359406338, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 118.18, | |
| "learning_rate": 6.841544117647059e-05, | |
| "loss": 1.0897, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 122.73, | |
| "learning_rate": 6.731249999999999e-05, | |
| "loss": 1.0634, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 127.27, | |
| "learning_rate": 6.62095588235294e-05, | |
| "loss": 1.0321, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 131.82, | |
| "learning_rate": 6.510661764705882e-05, | |
| "loss": 0.9962, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 136.36, | |
| "learning_rate": 6.400367647058824e-05, | |
| "loss": 0.9709, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 136.36, | |
| "eval_loss": 0.4784719944000244, | |
| "eval_runtime": 13.5638, | |
| "eval_samples_per_second": 22.56, | |
| "eval_steps_per_second": 2.875, | |
| "eval_wer": 0.6012835940633775, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 140.91, | |
| "learning_rate": 6.290073529411764e-05, | |
| "loss": 0.935, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 145.45, | |
| "learning_rate": 6.179779411764705e-05, | |
| "loss": 0.9179, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 6.069485294117646e-05, | |
| "loss": 0.8854, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 154.55, | |
| "learning_rate": 5.9591911764705876e-05, | |
| "loss": 0.87, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 159.09, | |
| "learning_rate": 5.848897058823529e-05, | |
| "loss": 0.8433, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 159.09, | |
| "eval_loss": 0.5048245191574097, | |
| "eval_runtime": 14.3488, | |
| "eval_samples_per_second": 21.326, | |
| "eval_steps_per_second": 2.718, | |
| "eval_wer": 0.5419173686321701, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 163.64, | |
| "learning_rate": 5.73860294117647e-05, | |
| "loss": 0.8145, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 168.18, | |
| "learning_rate": 5.6283088235294115e-05, | |
| "loss": 0.7986, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 172.73, | |
| "learning_rate": 5.5180147058823523e-05, | |
| "loss": 0.7804, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 177.27, | |
| "learning_rate": 5.407720588235294e-05, | |
| "loss": 0.7639, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 181.82, | |
| "learning_rate": 5.297426470588235e-05, | |
| "loss": 0.7404, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 181.82, | |
| "eval_loss": 0.5051876902580261, | |
| "eval_runtime": 14.251, | |
| "eval_samples_per_second": 21.472, | |
| "eval_steps_per_second": 2.737, | |
| "eval_wer": 0.533894905736061, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 186.36, | |
| "learning_rate": 5.187132352941176e-05, | |
| "loss": 0.7352, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 190.91, | |
| "learning_rate": 5.076838235294117e-05, | |
| "loss": 0.7073, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 195.45, | |
| "learning_rate": 4.9665441176470586e-05, | |
| "loss": 0.6933, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "learning_rate": 4.8562499999999995e-05, | |
| "loss": 0.6729, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 204.55, | |
| "learning_rate": 4.747058823529411e-05, | |
| "loss": 0.6589, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 204.55, | |
| "eval_loss": 0.5237030982971191, | |
| "eval_runtime": 13.4774, | |
| "eval_samples_per_second": 22.705, | |
| "eval_steps_per_second": 2.894, | |
| "eval_wer": 0.5896510228640193, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 209.09, | |
| "learning_rate": 4.6367647058823526e-05, | |
| "loss": 0.644, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 213.64, | |
| "learning_rate": 4.5264705882352935e-05, | |
| "loss": 0.6221, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 218.18, | |
| "learning_rate": 4.416176470588235e-05, | |
| "loss": 0.6119, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 222.73, | |
| "learning_rate": 4.305882352941176e-05, | |
| "loss": 0.601, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 227.27, | |
| "learning_rate": 4.196691176470588e-05, | |
| "loss": 0.5831, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 227.27, | |
| "eval_loss": 0.5166385769844055, | |
| "eval_runtime": 13.5084, | |
| "eval_samples_per_second": 22.653, | |
| "eval_steps_per_second": 2.887, | |
| "eval_wer": 0.5447252306458082, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 231.82, | |
| "learning_rate": 4.086397058823529e-05, | |
| "loss": 0.579, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 236.36, | |
| "learning_rate": 3.9761029411764705e-05, | |
| "loss": 0.5577, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 240.91, | |
| "learning_rate": 3.8658088235294113e-05, | |
| "loss": 0.547, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 245.45, | |
| "learning_rate": 3.755514705882353e-05, | |
| "loss": 0.5452, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 250.0, | |
| "learning_rate": 3.645220588235294e-05, | |
| "loss": 0.5375, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 250.0, | |
| "eval_loss": 0.5291843414306641, | |
| "eval_runtime": 13.5842, | |
| "eval_samples_per_second": 22.526, | |
| "eval_steps_per_second": 2.871, | |
| "eval_wer": 0.5487364620938628, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 254.55, | |
| "learning_rate": 3.534926470588235e-05, | |
| "loss": 0.5233, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 259.09, | |
| "learning_rate": 3.424632352941176e-05, | |
| "loss": 0.5081, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 263.64, | |
| "learning_rate": 3.314338235294117e-05, | |
| "loss": 0.5033, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 268.18, | |
| "learning_rate": 3.2040441176470585e-05, | |
| "loss": 0.4857, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 272.73, | |
| "learning_rate": 3.093749999999999e-05, | |
| "loss": 0.4784, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 272.73, | |
| "eval_loss": 0.548018217086792, | |
| "eval_runtime": 13.4861, | |
| "eval_samples_per_second": 22.69, | |
| "eval_steps_per_second": 2.892, | |
| "eval_wer": 0.5595667870036101, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 277.27, | |
| "learning_rate": 2.9834558823529408e-05, | |
| "loss": 0.4693, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 281.82, | |
| "learning_rate": 2.8731617647058824e-05, | |
| "loss": 0.4673, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 286.36, | |
| "learning_rate": 2.7628676470588235e-05, | |
| "loss": 0.4582, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 290.91, | |
| "learning_rate": 2.6525735294117647e-05, | |
| "loss": 0.4397, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 295.45, | |
| "learning_rate": 2.542279411764706e-05, | |
| "loss": 0.4421, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 295.45, | |
| "eval_loss": 0.5682183504104614, | |
| "eval_runtime": 13.4936, | |
| "eval_samples_per_second": 22.677, | |
| "eval_steps_per_second": 2.89, | |
| "eval_wer": 0.5467308463698355, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 300.0, | |
| "learning_rate": 2.431985294117647e-05, | |
| "loss": 0.4361, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 304.55, | |
| "learning_rate": 2.321691176470588e-05, | |
| "loss": 0.4287, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 309.09, | |
| "learning_rate": 2.211397058823529e-05, | |
| "loss": 0.4181, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 313.64, | |
| "learning_rate": 2.1011029411764703e-05, | |
| "loss": 0.4156, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 318.18, | |
| "learning_rate": 1.9908088235294115e-05, | |
| "loss": 0.4047, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 318.18, | |
| "eval_loss": 0.5681235194206238, | |
| "eval_runtime": 13.3347, | |
| "eval_samples_per_second": 22.948, | |
| "eval_steps_per_second": 2.925, | |
| "eval_wer": 0.5447252306458082, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 322.73, | |
| "learning_rate": 1.8805147058823527e-05, | |
| "loss": 0.3978, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 327.27, | |
| "learning_rate": 1.770220588235294e-05, | |
| "loss": 0.3874, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 331.82, | |
| "learning_rate": 1.659926470588235e-05, | |
| "loss": 0.385, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 336.36, | |
| "learning_rate": 1.5496323529411763e-05, | |
| "loss": 0.3817, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 340.91, | |
| "learning_rate": 1.4393382352941176e-05, | |
| "loss": 0.3779, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 340.91, | |
| "eval_loss": 0.578274667263031, | |
| "eval_runtime": 13.5305, | |
| "eval_samples_per_second": 22.616, | |
| "eval_steps_per_second": 2.882, | |
| "eval_wer": 0.5346971520256719, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 345.45, | |
| "learning_rate": 1.3290441176470588e-05, | |
| "loss": 0.3702, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 350.0, | |
| "learning_rate": 1.21875e-05, | |
| "loss": 0.3723, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 354.55, | |
| "learning_rate": 1.1084558823529412e-05, | |
| "loss": 0.3646, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 359.09, | |
| "learning_rate": 9.981617647058822e-06, | |
| "loss": 0.3478, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 363.64, | |
| "learning_rate": 8.878676470588234e-06, | |
| "loss": 0.3525, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 363.64, | |
| "eval_loss": 0.5856263041496277, | |
| "eval_runtime": 13.4874, | |
| "eval_samples_per_second": 22.688, | |
| "eval_steps_per_second": 2.892, | |
| "eval_wer": 0.5367027677496992, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 368.18, | |
| "learning_rate": 7.775735294117647e-06, | |
| "loss": 0.3574, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 372.73, | |
| "learning_rate": 6.672794117647058e-06, | |
| "loss": 0.3495, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 377.27, | |
| "learning_rate": 5.56985294117647e-06, | |
| "loss": 0.3576, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 381.82, | |
| "learning_rate": 4.466911764705882e-06, | |
| "loss": 0.3441, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 386.36, | |
| "learning_rate": 3.363970588235294e-06, | |
| "loss": 0.3393, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 386.36, | |
| "eval_loss": 0.5959608554840088, | |
| "eval_runtime": 13.5943, | |
| "eval_samples_per_second": 22.509, | |
| "eval_steps_per_second": 2.869, | |
| "eval_wer": 0.5359005214600883, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 390.91, | |
| "learning_rate": 2.2610294117647057e-06, | |
| "loss": 0.3366, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 395.45, | |
| "learning_rate": 1.1580882352941175e-06, | |
| "loss": 0.3404, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 400.0, | |
| "learning_rate": 5.514705882352941e-08, | |
| "loss": 0.3312, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 400.0, | |
| "step": 8800, | |
| "total_flos": 4.6124470831742304e+19, | |
| "train_loss": 1.3667902833765204, | |
| "train_runtime": 19151.8999, | |
| "train_samples_per_second": 14.578, | |
| "train_steps_per_second": 0.459 | |
| } | |
| ], | |
| "max_steps": 8800, | |
| "num_train_epochs": 400, | |
| "total_flos": 4.6124470831742304e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |