rohitp1/kkkh_w2lm_base_plus_finetune_teacher_noise_libri360_50_epochs_batch_16

dgx1_w2lm_base_plus_finetune_teacher_noise_libri360_50_epochs_batch_16

This model is a fine-tuned version of patrickvonplaten/wavlm-libri-clean-100h-base-plus on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1396
• Wer: 0.0921

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 16
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 256
• total_train_batch_size: 4096
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.2
• num_epochs: 50
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2792	5.91	150	0.2173	0.1413
0.1681	11.81	300	0.1653	0.1146
0.1325	17.72	450	0.1510	0.1040
0.1148	23.63	600	0.1451	0.0992
0.1044	29.53	750	0.1418	0.0964
0.0969	35.44	900	0.1403	0.0947
0.0924	41.35	1050	0.1403	0.0929
0.0899	47.25	1200	0.1396	0.0921

Framework versions

• Transformers 4.29.2
• Pytorch 1.12.1
• Datasets 2.6.1
• Tokenizers 0.13.1