| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.15533980582525, | |
| "global_step": 5108, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 1.487205882352941e-05, | |
| "loss": 16.4084, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.91, | |
| "learning_rate": 2.9430882352941174e-05, | |
| "loss": 6.2142, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.87, | |
| "learning_rate": 4.398970588235294e-05, | |
| "loss": 4.1261, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "learning_rate": 5.8548529411764694e-05, | |
| "loss": 3.2894, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 7.83, | |
| "eval_cer": 1.0, | |
| "eval_loss": 3.1501412391662598, | |
| "eval_runtime": 111.2121, | |
| "eval_samples_per_second": 25.141, | |
| "eval_steps_per_second": 0.396, | |
| "eval_wer": 1.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 7.310735294117646e-05, | |
| "loss": 3.083, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 11.76, | |
| "learning_rate": 7.5e-05, | |
| "loss": 2.9638, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 13.72, | |
| "learning_rate": 7.5e-05, | |
| "loss": 2.534, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 15.68, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.8586, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 15.68, | |
| "eval_cer": 0.24023761138033453, | |
| "eval_loss": 0.8871385455131531, | |
| "eval_runtime": 114.0709, | |
| "eval_samples_per_second": 24.511, | |
| "eval_steps_per_second": 0.386, | |
| "eval_wer": 0.6721492743607463, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 17.64, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.6008, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.4898, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 21.56, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.4105, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 23.52, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.3431, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 23.52, | |
| "eval_cer": 0.19387212755979366, | |
| "eval_loss": 0.5813264846801758, | |
| "eval_runtime": 110.7438, | |
| "eval_samples_per_second": 25.247, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.5502418797512094, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 25.49, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.2993, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 27.45, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.2563, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 29.41, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.2298, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 31.37, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.2052, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 31.37, | |
| "eval_cer": 0.1664738679589391, | |
| "eval_loss": 0.49559420347213745, | |
| "eval_runtime": 111.374, | |
| "eval_samples_per_second": 25.105, | |
| "eval_steps_per_second": 0.395, | |
| "eval_wer": 0.47878369039391844, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.186, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 35.29, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.1473, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 37.25, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.1235, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 39.21, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.1097, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 39.21, | |
| "eval_cer": 0.1397321661195352, | |
| "eval_loss": 0.44468843936920166, | |
| "eval_runtime": 109.918, | |
| "eval_samples_per_second": 25.437, | |
| "eval_steps_per_second": 0.4, | |
| "eval_wer": 0.41428242340474547, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 41.17, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.0951, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 43.14, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.0785, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 45.1, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.0575, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 47.06, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.0528, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 47.06, | |
| "eval_cer": 0.13331249022979522, | |
| "eval_loss": 0.4439217448234558, | |
| "eval_runtime": 111.2432, | |
| "eval_samples_per_second": 25.134, | |
| "eval_steps_per_second": 0.396, | |
| "eval_wer": 0.3960838516470859, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 49.02, | |
| "learning_rate": 7.5e-05, | |
| "loss": 1.0359, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 50.97, | |
| "learning_rate": 7.377299412915851e-05, | |
| "loss": 1.0118, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 52.93, | |
| "learning_rate": 7.098434442270058e-05, | |
| "loss": 1.0083, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 54.89, | |
| "learning_rate": 6.819569471624266e-05, | |
| "loss": 0.9939, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 54.89, | |
| "eval_cer": 0.13785628680110468, | |
| "eval_loss": 0.43479105830192566, | |
| "eval_runtime": 112.3814, | |
| "eval_samples_per_second": 24.88, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 0.40138217000691084, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 56.85, | |
| "learning_rate": 6.540704500978474e-05, | |
| "loss": 0.9884, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 58.82, | |
| "learning_rate": 6.26183953033268e-05, | |
| "loss": 0.9771, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 60.78, | |
| "learning_rate": 5.982974559686888e-05, | |
| "loss": 0.9577, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 62.74, | |
| "learning_rate": 5.704109589041095e-05, | |
| "loss": 0.9441, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 62.74, | |
| "eval_cer": 0.12230733156166954, | |
| "eval_loss": 0.42357733845710754, | |
| "eval_runtime": 114.6173, | |
| "eval_samples_per_second": 24.394, | |
| "eval_steps_per_second": 0.384, | |
| "eval_wer": 0.365261460492974, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 64.7, | |
| "learning_rate": 5.425244618395303e-05, | |
| "loss": 0.9296, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 66.66, | |
| "learning_rate": 5.14637964774951e-05, | |
| "loss": 0.9203, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 68.62, | |
| "learning_rate": 4.867514677103718e-05, | |
| "loss": 0.9215, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 70.58, | |
| "learning_rate": 4.5886497064579256e-05, | |
| "loss": 0.913, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 70.58, | |
| "eval_cer": 0.11566880308477932, | |
| "eval_loss": 0.43085047602653503, | |
| "eval_runtime": 112.648, | |
| "eval_samples_per_second": 24.821, | |
| "eval_steps_per_second": 0.391, | |
| "eval_wer": 0.3475236120709514, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 72.54, | |
| "learning_rate": 4.309784735812133e-05, | |
| "loss": 0.9033, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 74.5, | |
| "learning_rate": 4.03091976516634e-05, | |
| "loss": 0.8945, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 76.47, | |
| "learning_rate": 3.752054794520548e-05, | |
| "loss": 0.8714, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 78.43, | |
| "learning_rate": 3.473189823874755e-05, | |
| "loss": 0.8678, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 78.43, | |
| "eval_cer": 0.1109895263404721, | |
| "eval_loss": 0.42695942521095276, | |
| "eval_runtime": 109.8691, | |
| "eval_samples_per_second": 25.448, | |
| "eval_steps_per_second": 0.4, | |
| "eval_wer": 0.3337479843354066, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 80.39, | |
| "learning_rate": 3.1943248532289626e-05, | |
| "loss": 0.8612, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 82.35, | |
| "learning_rate": 2.9154598825831697e-05, | |
| "loss": 0.8603, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 84.31, | |
| "learning_rate": 2.6365949119373778e-05, | |
| "loss": 0.8451, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 86.27, | |
| "learning_rate": 2.357729941291585e-05, | |
| "loss": 0.8414, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 86.27, | |
| "eval_cer": 0.10695638580584649, | |
| "eval_loss": 0.41582536697387695, | |
| "eval_runtime": 110.6892, | |
| "eval_samples_per_second": 25.26, | |
| "eval_steps_per_second": 0.398, | |
| "eval_wer": 0.32204561161022804, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 88.23, | |
| "learning_rate": 2.0788649706457926e-05, | |
| "loss": 0.8394, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 90.19, | |
| "learning_rate": 1.7999999999999997e-05, | |
| "loss": 0.8304, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 92.16, | |
| "learning_rate": 1.5211350293542076e-05, | |
| "loss": 0.8188, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 94.12, | |
| "learning_rate": 1.2422700587084147e-05, | |
| "loss": 0.817, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 94.12, | |
| "eval_cer": 0.10721692460007295, | |
| "eval_loss": 0.4184626638889313, | |
| "eval_runtime": 112.2012, | |
| "eval_samples_per_second": 24.92, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 0.323105275282193, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 96.08, | |
| "learning_rate": 9.634050880626226e-06, | |
| "loss": 0.8126, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 98.04, | |
| "learning_rate": 6.845401174168297e-06, | |
| "loss": 0.8105, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 99.99, | |
| "learning_rate": 4.056751467710376e-06, | |
| "loss": 0.7959, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 100.16, | |
| "step": 5108, | |
| "total_flos": 8.553278225483388e+19, | |
| "train_loss": 1.6543282796931436, | |
| "train_runtime": 29157.3737, | |
| "train_samples_per_second": 22.424, | |
| "train_steps_per_second": 0.175 | |
| } | |
| ], | |
| "max_steps": 5108, | |
| "num_train_epochs": 101, | |
| "total_flos": 8.553278225483388e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |