| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 70.0, | |
| "global_step": 7980, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 1.3719999999999999e-05, | |
| "loss": 8.2074, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "learning_rate": 2.772e-05, | |
| "loss": 3.7205, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 4.1719999999999994e-05, | |
| "loss": 3.1583, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "learning_rate": 5.5719999999999995e-05, | |
| "loss": 2.9766, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 6.971999999999999e-05, | |
| "loss": 2.9032, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "eval_loss": 2.8768463134765625, | |
| "eval_runtime": 54.8752, | |
| "eval_samples_per_second": 29.139, | |
| "eval_steps_per_second": 0.911, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.26, | |
| "learning_rate": 6.908288770053476e-05, | |
| "loss": 2.8346, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.14, | |
| "learning_rate": 6.81470588235294e-05, | |
| "loss": 2.7766, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.02, | |
| "learning_rate": 6.721122994652407e-05, | |
| "loss": 2.3776, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 7.89, | |
| "learning_rate": 6.627540106951871e-05, | |
| "loss": 1.8024, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "learning_rate": 6.533957219251336e-05, | |
| "loss": 1.5724, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "eval_loss": 0.5638473033905029, | |
| "eval_runtime": 54.3288, | |
| "eval_samples_per_second": 29.432, | |
| "eval_steps_per_second": 0.92, | |
| "eval_wer": 0.6437733562266438, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.65, | |
| "learning_rate": 6.440374331550802e-05, | |
| "loss": 1.4498, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 10.53, | |
| "learning_rate": 6.346791443850267e-05, | |
| "loss": 1.3426, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "learning_rate": 6.253208556149732e-05, | |
| "loss": 1.2806, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 12.28, | |
| "learning_rate": 6.159625668449198e-05, | |
| "loss": 1.2595, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 13.16, | |
| "learning_rate": 6.0660427807486626e-05, | |
| "loss": 1.1818, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 13.16, | |
| "eval_loss": 0.3338354229927063, | |
| "eval_runtime": 54.9878, | |
| "eval_samples_per_second": 29.079, | |
| "eval_steps_per_second": 0.909, | |
| "eval_wer": 0.4759115240884759, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 14.04, | |
| "learning_rate": 5.972459893048127e-05, | |
| "loss": 1.1818, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 14.91, | |
| "learning_rate": 5.8788770053475934e-05, | |
| "loss": 1.133, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 15.79, | |
| "learning_rate": 5.785294117647058e-05, | |
| "loss": 1.1239, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 5.6917112299465236e-05, | |
| "loss": 1.0991, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 17.54, | |
| "learning_rate": 5.5981283422459884e-05, | |
| "loss": 1.0798, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 17.54, | |
| "eval_loss": 0.2876473665237427, | |
| "eval_runtime": 49.892, | |
| "eval_samples_per_second": 32.049, | |
| "eval_steps_per_second": 1.002, | |
| "eval_wer": 0.40864559135440864, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 18.42, | |
| "learning_rate": 5.5045454545454545e-05, | |
| "loss": 1.0756, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 19.3, | |
| "learning_rate": 5.410962566844919e-05, | |
| "loss": 1.0653, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 20.18, | |
| "learning_rate": 5.317379679144385e-05, | |
| "loss": 1.0472, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 21.05, | |
| "learning_rate": 5.2237967914438494e-05, | |
| "loss": 1.03, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 21.93, | |
| "learning_rate": 5.1302139037433155e-05, | |
| "loss": 1.0296, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 21.93, | |
| "eval_loss": 0.2693549394607544, | |
| "eval_runtime": 49.7815, | |
| "eval_samples_per_second": 32.12, | |
| "eval_steps_per_second": 1.004, | |
| "eval_wer": 0.4248055751944248, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 22.81, | |
| "learning_rate": 5.03663101604278e-05, | |
| "loss": 1.0121, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 23.68, | |
| "learning_rate": 4.943048128342246e-05, | |
| "loss": 1.0139, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 24.56, | |
| "learning_rate": 4.8494652406417105e-05, | |
| "loss": 1.0081, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 25.44, | |
| "learning_rate": 4.755882352941176e-05, | |
| "loss": 0.994, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 26.32, | |
| "learning_rate": 4.6622994652406414e-05, | |
| "loss": 1.0014, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 26.32, | |
| "eval_loss": 0.26258525252342224, | |
| "eval_runtime": 49.6202, | |
| "eval_samples_per_second": 32.225, | |
| "eval_steps_per_second": 1.008, | |
| "eval_wer": 0.3732956267043733, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 27.19, | |
| "learning_rate": 4.569652406417112e-05, | |
| "loss": 0.991, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 28.07, | |
| "learning_rate": 4.4760695187165766e-05, | |
| "loss": 0.9786, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 28.95, | |
| "learning_rate": 4.382486631016043e-05, | |
| "loss": 0.9689, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 29.82, | |
| "learning_rate": 4.2889037433155075e-05, | |
| "loss": 0.9631, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 30.7, | |
| "learning_rate": 4.195320855614973e-05, | |
| "loss": 0.9616, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 30.7, | |
| "eval_loss": 0.23906062543392181, | |
| "eval_runtime": 49.6357, | |
| "eval_samples_per_second": 32.215, | |
| "eval_steps_per_second": 1.007, | |
| "eval_wer": 0.32936067063932933, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 31.58, | |
| "learning_rate": 4.101737967914438e-05, | |
| "loss": 0.9375, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 32.46, | |
| "learning_rate": 4.008155080213904e-05, | |
| "loss": 0.9629, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 3.9145721925133686e-05, | |
| "loss": 0.946, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 34.21, | |
| "learning_rate": 3.820989304812834e-05, | |
| "loss": 0.9511, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 35.09, | |
| "learning_rate": 3.727406417112299e-05, | |
| "loss": 0.9303, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 35.09, | |
| "eval_loss": 0.23517830669879913, | |
| "eval_runtime": 50.0462, | |
| "eval_samples_per_second": 31.95, | |
| "eval_steps_per_second": 0.999, | |
| "eval_wer": 0.3217856782143218, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 35.96, | |
| "learning_rate": 3.633823529411765e-05, | |
| "loss": 0.9236, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 36.84, | |
| "learning_rate": 3.5402406417112296e-05, | |
| "loss": 0.9278, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 37.72, | |
| "learning_rate": 3.446657754010695e-05, | |
| "loss": 0.9218, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 38.6, | |
| "learning_rate": 3.35307486631016e-05, | |
| "loss": 0.922, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 39.47, | |
| "learning_rate": 3.259491978609625e-05, | |
| "loss": 0.9248, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 39.47, | |
| "eval_loss": 0.23506982624530792, | |
| "eval_runtime": 49.0987, | |
| "eval_samples_per_second": 32.567, | |
| "eval_steps_per_second": 1.018, | |
| "eval_wer": 0.3206746793253207, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 40.35, | |
| "learning_rate": 3.165909090909091e-05, | |
| "loss": 0.8909, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 41.23, | |
| "learning_rate": 3.072326203208556e-05, | |
| "loss": 0.912, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 42.11, | |
| "learning_rate": 2.9787433155080212e-05, | |
| "loss": 0.8901, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 42.98, | |
| "learning_rate": 2.8851604278074863e-05, | |
| "loss": 0.9007, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 43.86, | |
| "learning_rate": 2.7915775401069517e-05, | |
| "loss": 0.8837, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 43.86, | |
| "eval_loss": 0.23407745361328125, | |
| "eval_runtime": 49.1749, | |
| "eval_samples_per_second": 32.517, | |
| "eval_steps_per_second": 1.017, | |
| "eval_wer": 0.31027168972831026, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 44.74, | |
| "learning_rate": 2.6979946524064168e-05, | |
| "loss": 0.8891, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 45.61, | |
| "learning_rate": 2.6044117647058823e-05, | |
| "loss": 0.882, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 46.49, | |
| "learning_rate": 2.5108288770053474e-05, | |
| "loss": 0.8921, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 47.37, | |
| "learning_rate": 2.4172459893048128e-05, | |
| "loss": 0.8794, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 48.25, | |
| "learning_rate": 2.323663101604278e-05, | |
| "loss": 0.8887, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 48.25, | |
| "eval_loss": 0.23113244771957397, | |
| "eval_runtime": 48.9674, | |
| "eval_samples_per_second": 32.654, | |
| "eval_steps_per_second": 1.021, | |
| "eval_wer": 0.3114836885163115, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 49.12, | |
| "learning_rate": 2.2300802139037433e-05, | |
| "loss": 0.8553, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 2.1364973262032084e-05, | |
| "loss": 0.8751, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 50.88, | |
| "learning_rate": 2.042914438502674e-05, | |
| "loss": 0.8557, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 51.75, | |
| "learning_rate": 1.949331550802139e-05, | |
| "loss": 0.8626, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 52.63, | |
| "learning_rate": 1.8557486631016044e-05, | |
| "loss": 0.8529, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 52.63, | |
| "eval_loss": 0.22297005355358124, | |
| "eval_runtime": 49.2944, | |
| "eval_samples_per_second": 32.438, | |
| "eval_steps_per_second": 1.014, | |
| "eval_wer": 0.3000706999293001, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 53.51, | |
| "learning_rate": 1.762165775401069e-05, | |
| "loss": 0.8487, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 54.39, | |
| "learning_rate": 1.6685828877005346e-05, | |
| "loss": 0.8585, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 55.26, | |
| "learning_rate": 1.575e-05, | |
| "loss": 0.8412, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 56.14, | |
| "learning_rate": 1.4814171122994651e-05, | |
| "loss": 0.8457, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 57.02, | |
| "learning_rate": 1.3878342245989304e-05, | |
| "loss": 0.8404, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 57.02, | |
| "eval_loss": 0.22788779437541962, | |
| "eval_runtime": 48.6265, | |
| "eval_samples_per_second": 32.883, | |
| "eval_steps_per_second": 1.028, | |
| "eval_wer": 0.3054236945763054, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 57.89, | |
| "learning_rate": 1.2942513368983956e-05, | |
| "loss": 0.8342, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 58.77, | |
| "learning_rate": 1.2006684491978609e-05, | |
| "loss": 0.844, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 59.65, | |
| "learning_rate": 1.107085561497326e-05, | |
| "loss": 0.8276, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 60.53, | |
| "learning_rate": 1.0144385026737967e-05, | |
| "loss": 0.8264, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 61.4, | |
| "learning_rate": 9.20855614973262e-06, | |
| "loss": 0.8242, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 61.4, | |
| "eval_loss": 0.2298082411289215, | |
| "eval_runtime": 48.4266, | |
| "eval_samples_per_second": 33.019, | |
| "eval_steps_per_second": 1.032, | |
| "eval_wer": 0.30057569942430057, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 62.28, | |
| "learning_rate": 8.272727272727272e-06, | |
| "loss": 0.8291, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 63.16, | |
| "learning_rate": 7.336898395721925e-06, | |
| "loss": 0.8213, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 64.04, | |
| "learning_rate": 6.401069518716577e-06, | |
| "loss": 0.8139, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 64.91, | |
| "learning_rate": 5.46524064171123e-06, | |
| "loss": 0.8177, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 65.79, | |
| "learning_rate": 4.529411764705883e-06, | |
| "loss": 0.8288, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 65.79, | |
| "eval_loss": 0.2333229035139084, | |
| "eval_runtime": 48.8974, | |
| "eval_samples_per_second": 32.701, | |
| "eval_steps_per_second": 1.023, | |
| "eval_wer": 0.2996667003332997, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "learning_rate": 3.5935828877005344e-06, | |
| "loss": 0.8184, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 67.54, | |
| "learning_rate": 2.657754010695187e-06, | |
| "loss": 0.8178, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 68.42, | |
| "learning_rate": 1.7219251336898393e-06, | |
| "loss": 0.8214, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 69.3, | |
| "learning_rate": 7.860962566844919e-07, | |
| "loss": 0.8133, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "step": 7980, | |
| "total_flos": 2.7520612954833367e+19, | |
| "train_loss": 1.2354308326740313, | |
| "train_runtime": 11524.5807, | |
| "train_samples_per_second": 22.03, | |
| "train_steps_per_second": 0.692 | |
| } | |
| ], | |
| "max_steps": 7980, | |
| "num_train_epochs": 70, | |
| "total_flos": 2.7520612954833367e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |