| label,log_path,started,checkpoint_dir,completed_training,early_stopped,total_params,trainable_params,train_samples,val_samples,batch_size,epochs_requested,observed_total_epochs_field,observed_epochs_in_log,steps_per_epoch,avg_epoch_seconds,avg_epoch_minutes,median_epoch_minutes,min_epoch_seconds,max_epoch_seconds,approx_batches_per_second,approx_samples_per_second,total_epochs_reported,total_steps_reported,approx_total_hours_if_using_logged_epoch_average,note | |
| good_log,<local_project_root>/logs/good_log.out,,/work/ratul1/supantha/glycan-SD-VS/bert_training_v3/v3.1_cluster_training/checkpoints_v4_bpe,True,True,36997516,36997516,208562,52141,128,100,100,85,1630,620.2,10.34,9.97,594,690,2.6282,336.28,85,138550,14.64,Recovered completed pretraining log for the smaller checkpoint-writing run that saves into checkpoints_v4_bpe. | |
| current_log,<local_project_root>/logs/current_log.out,Tue Dec 23 10:04:07 PM CST 2025,/work/ratul1/supantha/glycan-SD-VS/bert_training_v3/v3.1_cluster_training/checkpoints_v4_bpe_topology,False,False,136470307,136470307,208562,52141,64,100,100,17,3259,3158.82,52.65,50.75,3040,3529,1.0317,66.03,,,,Recovered partial pretraining log for the heavier topology-enabled run that saves into checkpoints_v4_bpe_topology. | |