| { | |
| "best_metric": 0.21267759799957275, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-nyagen-female-model/checkpoint-2100", | |
| "epoch": 7.339449541284404, | |
| "eval_steps": 100, | |
| "global_step": 2400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.3058103975535168, | |
| "grad_norm": 2.4750263690948486, | |
| "learning_rate": 0.00028799999999999995, | |
| "loss": 7.8679, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3058103975535168, | |
| "eval_loss": 0.9127748608589172, | |
| "eval_runtime": 15.6762, | |
| "eval_samples_per_second": 10.207, | |
| "eval_steps_per_second": 2.552, | |
| "eval_wer": 0.6897243107769424, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6116207951070336, | |
| "grad_norm": 1.507306694984436, | |
| "learning_rate": 0.00029703398558187435, | |
| "loss": 0.627, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6116207951070336, | |
| "eval_loss": 0.3264550268650055, | |
| "eval_runtime": 15.9193, | |
| "eval_samples_per_second": 10.051, | |
| "eval_steps_per_second": 2.513, | |
| "eval_wer": 0.43859649122807015, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9174311926605505, | |
| "grad_norm": 1.2747225761413574, | |
| "learning_rate": 0.0002939443872296601, | |
| "loss": 0.5004, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9174311926605505, | |
| "eval_loss": 0.29508110880851746, | |
| "eval_runtime": 15.668, | |
| "eval_samples_per_second": 10.212, | |
| "eval_steps_per_second": 2.553, | |
| "eval_wer": 0.40401002506265665, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.2232415902140672, | |
| "grad_norm": 1.1947221755981445, | |
| "learning_rate": 0.0002908547888774459, | |
| "loss": 0.4611, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2232415902140672, | |
| "eval_loss": 0.27579107880592346, | |
| "eval_runtime": 15.5786, | |
| "eval_samples_per_second": 10.271, | |
| "eval_steps_per_second": 2.568, | |
| "eval_wer": 0.3659147869674185, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.529051987767584, | |
| "grad_norm": 0.9117503762245178, | |
| "learning_rate": 0.0002877651905252317, | |
| "loss": 0.4455, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.529051987767584, | |
| "eval_loss": 0.266660213470459, | |
| "eval_runtime": 15.6734, | |
| "eval_samples_per_second": 10.208, | |
| "eval_steps_per_second": 2.552, | |
| "eval_wer": 0.3598997493734336, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.834862385321101, | |
| "grad_norm": 0.6930361390113831, | |
| "learning_rate": 0.0002846755921730175, | |
| "loss": 0.4142, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.834862385321101, | |
| "eval_loss": 0.26014643907546997, | |
| "eval_runtime": 15.6292, | |
| "eval_samples_per_second": 10.237, | |
| "eval_steps_per_second": 2.559, | |
| "eval_wer": 0.35137844611528823, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.140672782874618, | |
| "grad_norm": 1.517812967300415, | |
| "learning_rate": 0.00028158599382080326, | |
| "loss": 0.3706, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.140672782874618, | |
| "eval_loss": 0.25211504101753235, | |
| "eval_runtime": 15.5542, | |
| "eval_samples_per_second": 10.287, | |
| "eval_steps_per_second": 2.572, | |
| "eval_wer": 0.34786967418546366, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.4464831804281344, | |
| "grad_norm": 0.9740973114967346, | |
| "learning_rate": 0.00027849639546858907, | |
| "loss": 0.3867, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.4464831804281344, | |
| "eval_loss": 0.24611802399158478, | |
| "eval_runtime": 15.6894, | |
| "eval_samples_per_second": 10.198, | |
| "eval_steps_per_second": 2.549, | |
| "eval_wer": 0.31278195488721805, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.7522935779816513, | |
| "grad_norm": 2.278139114379883, | |
| "learning_rate": 0.0002754067971163748, | |
| "loss": 0.3537, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.7522935779816513, | |
| "eval_loss": 0.24492569267749786, | |
| "eval_runtime": 15.6703, | |
| "eval_samples_per_second": 10.21, | |
| "eval_steps_per_second": 2.553, | |
| "eval_wer": 0.3157894736842105, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 3.058103975535168, | |
| "grad_norm": 0.7251365780830383, | |
| "learning_rate": 0.0002723171987641606, | |
| "loss": 0.3821, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.058103975535168, | |
| "eval_loss": 0.24130229651927948, | |
| "eval_runtime": 15.7231, | |
| "eval_samples_per_second": 10.176, | |
| "eval_steps_per_second": 2.544, | |
| "eval_wer": 0.2932330827067669, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.363914373088685, | |
| "grad_norm": 10.20130729675293, | |
| "learning_rate": 0.00026922760041194643, | |
| "loss": 0.3626, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.363914373088685, | |
| "eval_loss": 0.23597605526447296, | |
| "eval_runtime": 15.6773, | |
| "eval_samples_per_second": 10.206, | |
| "eval_steps_per_second": 2.551, | |
| "eval_wer": 0.3082706766917293, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 3.669724770642202, | |
| "grad_norm": 0.775827944278717, | |
| "learning_rate": 0.00026613800205973223, | |
| "loss": 0.3312, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.669724770642202, | |
| "eval_loss": 0.2335551679134369, | |
| "eval_runtime": 15.7243, | |
| "eval_samples_per_second": 10.175, | |
| "eval_steps_per_second": 2.544, | |
| "eval_wer": 0.29974937343358393, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 3.9755351681957185, | |
| "grad_norm": 1.4778733253479004, | |
| "learning_rate": 0.00026307929969104015, | |
| "loss": 0.3322, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 3.9755351681957185, | |
| "eval_loss": 0.22845473885536194, | |
| "eval_runtime": 15.7244, | |
| "eval_samples_per_second": 10.175, | |
| "eval_steps_per_second": 2.544, | |
| "eval_wer": 0.2967418546365915, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 4.281345565749236, | |
| "grad_norm": 63.480228424072266, | |
| "learning_rate": 0.0002599897013388259, | |
| "loss": 0.3654, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.281345565749236, | |
| "eval_loss": 0.22345833480358124, | |
| "eval_runtime": 15.723, | |
| "eval_samples_per_second": 10.176, | |
| "eval_steps_per_second": 2.544, | |
| "eval_wer": 0.28521303258145364, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 4.587155963302752, | |
| "grad_norm": 0.6586478352546692, | |
| "learning_rate": 0.0002569001029866117, | |
| "loss": 0.3241, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.587155963302752, | |
| "eval_loss": 0.21982760727405548, | |
| "eval_runtime": 15.6138, | |
| "eval_samples_per_second": 10.247, | |
| "eval_steps_per_second": 2.562, | |
| "eval_wer": 0.2807017543859649, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.892966360856269, | |
| "grad_norm": 1.3852250576019287, | |
| "learning_rate": 0.0002538105046343975, | |
| "loss": 0.2908, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 4.892966360856269, | |
| "eval_loss": 0.21671359241008759, | |
| "eval_runtime": 15.6441, | |
| "eval_samples_per_second": 10.228, | |
| "eval_steps_per_second": 2.557, | |
| "eval_wer": 0.27669172932330827, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 5.198776758409786, | |
| "grad_norm": 1.687046766281128, | |
| "learning_rate": 0.0002507209062821833, | |
| "loss": 0.3299, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.198776758409786, | |
| "eval_loss": 0.21697847545146942, | |
| "eval_runtime": 15.6568, | |
| "eval_samples_per_second": 10.219, | |
| "eval_steps_per_second": 2.555, | |
| "eval_wer": 0.2746867167919799, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 5.504587155963303, | |
| "grad_norm": 1.117961049079895, | |
| "learning_rate": 0.00024763130792996907, | |
| "loss": 0.3128, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.504587155963303, | |
| "eval_loss": 0.21468114852905273, | |
| "eval_runtime": 15.7336, | |
| "eval_samples_per_second": 10.169, | |
| "eval_steps_per_second": 2.542, | |
| "eval_wer": 0.268671679197995, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 5.81039755351682, | |
| "grad_norm": 0.5044408440589905, | |
| "learning_rate": 0.0002445417095777549, | |
| "loss": 0.3094, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 5.81039755351682, | |
| "eval_loss": 0.21403858065605164, | |
| "eval_runtime": 15.6158, | |
| "eval_samples_per_second": 10.246, | |
| "eval_steps_per_second": 2.562, | |
| "eval_wer": 0.27418546365914787, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 6.116207951070336, | |
| "grad_norm": 0.6591036319732666, | |
| "learning_rate": 0.00024145211122554065, | |
| "loss": 0.309, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.116207951070336, | |
| "eval_loss": 0.21581442654132843, | |
| "eval_runtime": 15.7498, | |
| "eval_samples_per_second": 10.159, | |
| "eval_steps_per_second": 2.54, | |
| "eval_wer": 0.27017543859649124, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 6.422018348623853, | |
| "grad_norm": 1.306911826133728, | |
| "learning_rate": 0.00023836251287332646, | |
| "loss": 0.3075, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.422018348623853, | |
| "eval_loss": 0.21267759799957275, | |
| "eval_runtime": 15.7059, | |
| "eval_samples_per_second": 10.187, | |
| "eval_steps_per_second": 2.547, | |
| "eval_wer": 0.26516290726817043, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 6.72782874617737, | |
| "grad_norm": 4.904253005981445, | |
| "learning_rate": 0.00023527291452111223, | |
| "loss": 0.2823, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 6.72782874617737, | |
| "eval_loss": 0.2155403196811676, | |
| "eval_runtime": 15.7469, | |
| "eval_samples_per_second": 10.161, | |
| "eval_steps_per_second": 2.54, | |
| "eval_wer": 0.2671679197994987, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 7.033639143730887, | |
| "grad_norm": 1.5606273412704468, | |
| "learning_rate": 0.00023221421215242018, | |
| "loss": 0.3062, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.033639143730887, | |
| "eval_loss": 0.21326474845409393, | |
| "eval_runtime": 15.6903, | |
| "eval_samples_per_second": 10.197, | |
| "eval_steps_per_second": 2.549, | |
| "eval_wer": 0.2556390977443609, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 7.339449541284404, | |
| "grad_norm": 1.3775653839111328, | |
| "learning_rate": 0.00022912461380020596, | |
| "loss": 0.3012, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.339449541284404, | |
| "eval_loss": 0.2165716141462326, | |
| "eval_runtime": 15.6703, | |
| "eval_samples_per_second": 10.21, | |
| "eval_steps_per_second": 2.553, | |
| "eval_wer": 0.25413533834586466, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 7.339449541284404, | |
| "step": 2400, | |
| "total_flos": 9.832669015298243e+18, | |
| "train_loss": 0.6780771311124166, | |
| "train_runtime": 2242.1481, | |
| "train_samples_per_second": 17.461, | |
| "train_steps_per_second": 4.375 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 9810, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.832669015298243e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |