| { | |
| "best_metric": 2.1115864203149175, | |
| "best_model_checkpoint": "./checkpoint-1000", | |
| "epoch": 33.333333333333336, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.2000000000000006e-07, | |
| "loss": 1.715, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_loss": 1.6869975328445435, | |
| "eval_runtime": 192.3622, | |
| "eval_samples_per_second": 1.242, | |
| "eval_steps_per_second": 0.156, | |
| "eval_wer": 97.23309365613908, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 9.200000000000001e-07, | |
| "loss": 1.3716, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "eval_loss": 1.2431073188781738, | |
| "eval_runtime": 160.0584, | |
| "eval_samples_per_second": 1.493, | |
| "eval_steps_per_second": 0.187, | |
| "eval_wer": 93.12824246837171, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 1.42e-06, | |
| "loss": 0.969, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "eval_loss": 0.8620495796203613, | |
| "eval_runtime": 184.7193, | |
| "eval_samples_per_second": 1.294, | |
| "eval_steps_per_second": 0.162, | |
| "eval_wer": 90.79821607354145, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 0.6238, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_loss": 0.4657336473464966, | |
| "eval_runtime": 195.5848, | |
| "eval_samples_per_second": 1.222, | |
| "eval_steps_per_second": 0.153, | |
| "eval_wer": 72.78601984163102, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "learning_rate": 2.42e-06, | |
| "loss": 0.3129, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "eval_loss": 0.283780962228775, | |
| "eval_runtime": 184.0918, | |
| "eval_samples_per_second": 1.298, | |
| "eval_steps_per_second": 0.163, | |
| "eval_wer": 34.486210976608724, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 2.92e-06, | |
| "loss": 0.2271, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.23963618278503418, | |
| "eval_runtime": 168.6349, | |
| "eval_samples_per_second": 1.417, | |
| "eval_steps_per_second": 0.178, | |
| "eval_wer": 21.498134158551014, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "learning_rate": 3.4200000000000007e-06, | |
| "loss": 0.1854, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "eval_loss": 0.21348817646503448, | |
| "eval_runtime": 131.9979, | |
| "eval_samples_per_second": 1.811, | |
| "eval_steps_per_second": 0.227, | |
| "eval_wer": 7.618094111222354, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 0.1527, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "eval_loss": 0.19052761793136597, | |
| "eval_runtime": 122.4712, | |
| "eval_samples_per_second": 1.951, | |
| "eval_steps_per_second": 0.245, | |
| "eval_wer": 6.189132611267862, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "learning_rate": 4.42e-06, | |
| "loss": 0.1214, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "eval_loss": 0.17034465074539185, | |
| "eval_runtime": 130.0187, | |
| "eval_samples_per_second": 1.838, | |
| "eval_steps_per_second": 0.231, | |
| "eval_wer": 5.242559388368071, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 4.92e-06, | |
| "loss": 0.0973, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "eval_loss": 0.1484680473804474, | |
| "eval_runtime": 113.2953, | |
| "eval_samples_per_second": 2.11, | |
| "eval_steps_per_second": 0.265, | |
| "eval_wer": 2.639483025393647, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "learning_rate": 5.420000000000001e-06, | |
| "loss": 0.0669, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 9.17, | |
| "eval_loss": 0.12224661558866501, | |
| "eval_runtime": 111.1904, | |
| "eval_samples_per_second": 2.149, | |
| "eval_steps_per_second": 0.27, | |
| "eval_wer": 1.7202147993082735, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 5.92e-06, | |
| "loss": 0.0291, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.09692100435495377, | |
| "eval_runtime": 110.9954, | |
| "eval_samples_per_second": 2.153, | |
| "eval_steps_per_second": 0.27, | |
| "eval_wer": 2.0478747610812778, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.83, | |
| "learning_rate": 6.42e-06, | |
| "loss": 0.0122, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 10.83, | |
| "eval_loss": 0.10275180637836456, | |
| "eval_runtime": 110.9654, | |
| "eval_samples_per_second": 2.154, | |
| "eval_steps_per_second": 0.27, | |
| "eval_wer": 1.8112314553563305, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 11.67, | |
| "learning_rate": 6.92e-06, | |
| "loss": 0.0087, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 11.67, | |
| "eval_loss": 0.1043357104063034, | |
| "eval_runtime": 110.8965, | |
| "eval_samples_per_second": 2.155, | |
| "eval_steps_per_second": 0.271, | |
| "eval_wer": 2.1206880859197232, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 7.420000000000001e-06, | |
| "loss": 0.005, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "eval_loss": 0.11149879544973373, | |
| "eval_runtime": 110.87, | |
| "eval_samples_per_second": 2.156, | |
| "eval_steps_per_second": 0.271, | |
| "eval_wer": 2.1661964139437515, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.0041, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "eval_loss": 0.11916878074407578, | |
| "eval_runtime": 110.7405, | |
| "eval_samples_per_second": 2.158, | |
| "eval_steps_per_second": 0.271, | |
| "eval_wer": 1.929553108218804, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 14.17, | |
| "learning_rate": 8.42e-06, | |
| "loss": 0.0036, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 14.17, | |
| "eval_loss": 0.11888804286718369, | |
| "eval_runtime": 113.1767, | |
| "eval_samples_per_second": 2.112, | |
| "eval_steps_per_second": 0.265, | |
| "eval_wer": 2.694093019022481, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 8.920000000000001e-06, | |
| "loss": 0.003, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 0.1166142076253891, | |
| "eval_runtime": 113.7728, | |
| "eval_samples_per_second": 2.101, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.557568034950396, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 15.83, | |
| "learning_rate": 9.42e-06, | |
| "loss": 0.0025, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 15.83, | |
| "eval_loss": 0.11832274496555328, | |
| "eval_runtime": 113.8233, | |
| "eval_samples_per_second": 2.1, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 3.1855829616819875, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.0022, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "eval_loss": 0.11866947263479233, | |
| "eval_runtime": 113.9373, | |
| "eval_samples_per_second": 2.098, | |
| "eval_steps_per_second": 0.263, | |
| "eval_wer": 3.34941294256849, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "learning_rate": 9.58e-06, | |
| "loss": 0.0019, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 17.5, | |
| "eval_loss": 0.12195830047130585, | |
| "eval_runtime": 110.795, | |
| "eval_samples_per_second": 2.157, | |
| "eval_steps_per_second": 0.271, | |
| "eval_wer": 1.7384181305178845, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 18.33, | |
| "learning_rate": 9.080000000000001e-06, | |
| "loss": 0.0014, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 18.33, | |
| "eval_loss": 0.12314373254776001, | |
| "eval_runtime": 113.6372, | |
| "eval_samples_per_second": 2.103, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.102484754710112, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 19.17, | |
| "learning_rate": 8.580000000000001e-06, | |
| "loss": 0.0014, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 19.17, | |
| "eval_loss": 0.12057469040155411, | |
| "eval_runtime": 113.9208, | |
| "eval_samples_per_second": 2.098, | |
| "eval_steps_per_second": 0.263, | |
| "eval_wer": 2.1935014107581687, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 8.08e-06, | |
| "loss": 0.0011, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 0.12358372658491135, | |
| "eval_runtime": 113.6203, | |
| "eval_samples_per_second": 2.103, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.1115864203149175, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 20.83, | |
| "learning_rate": 7.58e-06, | |
| "loss": 0.0008, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 20.83, | |
| "eval_loss": 0.12533117830753326, | |
| "eval_runtime": 113.505, | |
| "eval_samples_per_second": 2.106, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.1752980795485577, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 21.67, | |
| "learning_rate": 7.08e-06, | |
| "loss": 0.0008, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 21.67, | |
| "eval_loss": 0.12367217987775803, | |
| "eval_runtime": 113.5595, | |
| "eval_samples_per_second": 2.105, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.102484754710112, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "learning_rate": 6.5800000000000005e-06, | |
| "loss": 0.0008, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 22.5, | |
| "eval_loss": 0.12699657678604126, | |
| "eval_runtime": 113.7775, | |
| "eval_samples_per_second": 2.101, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.102484754710112, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 23.33, | |
| "learning_rate": 6.08e-06, | |
| "loss": 0.0008, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 23.33, | |
| "eval_loss": 0.12633280456066132, | |
| "eval_runtime": 113.4116, | |
| "eval_samples_per_second": 2.107, | |
| "eval_steps_per_second": 0.265, | |
| "eval_wer": 2.093383089105306, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 24.17, | |
| "learning_rate": 5.580000000000001e-06, | |
| "loss": 0.0007, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 24.17, | |
| "eval_loss": 0.12679414451122284, | |
| "eval_runtime": 113.4198, | |
| "eval_samples_per_second": 2.107, | |
| "eval_steps_per_second": 0.265, | |
| "eval_wer": 2.093383089105306, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 5.0800000000000005e-06, | |
| "loss": 0.0006, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_loss": 0.1280515044927597, | |
| "eval_runtime": 113.5855, | |
| "eval_samples_per_second": 2.104, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.220806407572586, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 25.83, | |
| "learning_rate": 4.58e-06, | |
| "loss": 0.0005, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 25.83, | |
| "eval_loss": 0.12817054986953735, | |
| "eval_runtime": 113.5145, | |
| "eval_samples_per_second": 2.105, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.093383089105306, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "learning_rate": 4.08e-06, | |
| "loss": 0.0007, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 26.67, | |
| "eval_loss": 0.12809617817401886, | |
| "eval_runtime": 113.5669, | |
| "eval_samples_per_second": 2.104, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.1661964139437515, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "learning_rate": 3.58e-06, | |
| "loss": 0.0006, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 27.5, | |
| "eval_loss": 0.12927967309951782, | |
| "eval_runtime": 113.4285, | |
| "eval_samples_per_second": 2.107, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.1115864203149175, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 28.33, | |
| "learning_rate": 3.08e-06, | |
| "loss": 0.0006, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 28.33, | |
| "eval_loss": 0.1292882263660431, | |
| "eval_runtime": 113.3979, | |
| "eval_samples_per_second": 2.108, | |
| "eval_steps_per_second": 0.265, | |
| "eval_wer": 2.2026030763629745, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 29.17, | |
| "learning_rate": 2.5800000000000003e-06, | |
| "loss": 0.0006, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 29.17, | |
| "eval_loss": 0.12911662459373474, | |
| "eval_runtime": 113.514, | |
| "eval_samples_per_second": 2.105, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.1206880859197232, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "learning_rate": 2.08e-06, | |
| "loss": 0.0006, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_loss": 0.12980858981609344, | |
| "eval_runtime": 113.4833, | |
| "eval_samples_per_second": 2.106, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.184399745153363, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 30.83, | |
| "learning_rate": 1.5800000000000001e-06, | |
| "loss": 0.0005, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 30.83, | |
| "eval_loss": 0.13006018102169037, | |
| "eval_runtime": 113.4467, | |
| "eval_samples_per_second": 2.107, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.1115864203149175, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 31.67, | |
| "learning_rate": 1.08e-06, | |
| "loss": 0.0005, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 31.67, | |
| "eval_loss": 0.13012319803237915, | |
| "eval_runtime": 113.2847, | |
| "eval_samples_per_second": 2.11, | |
| "eval_steps_per_second": 0.265, | |
| "eval_wer": 2.1115864203149175, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "learning_rate": 5.800000000000001e-07, | |
| "loss": 0.0005, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 32.5, | |
| "eval_loss": 0.13053424656391144, | |
| "eval_runtime": 113.4539, | |
| "eval_samples_per_second": 2.107, | |
| "eval_steps_per_second": 0.264, | |
| "eval_wer": 2.1115864203149175, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 8e-08, | |
| "loss": 0.0005, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "eval_loss": 0.13054169714450836, | |
| "eval_runtime": 113.3817, | |
| "eval_samples_per_second": 2.108, | |
| "eval_steps_per_second": 0.265, | |
| "eval_wer": 2.1115864203149175, | |
| "step": 1000 | |
| } | |
| ], | |
| "max_steps": 1000, | |
| "num_train_epochs": 34, | |
| "total_flos": 9.19663953739776e+18, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |