wav2vec2-960h-fine-tuning

This model is a fine-tuned version of facebook/wav2vec2-base-960h on the None dataset. It achieves the following results on the evaluation set:

• Loss: 5.8952
• Wer: 99.9699

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: 0.0001
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
15.6622	0.6536	100	56.5642	99.9699
9.4272	1.3072	200	8.8586	99.9699
4.5296	1.9608	300	5.3600	99.9699
3.4224	2.6144	400	5.8952	99.9699

Framework versions

• Transformers 4.42.3
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1