asr_en_model

This model is a fine-tuned version of facebook/wav2vec2-large-960h-lv60-self on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0593
• Wer: 0.1522

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: 36
• eval_batch_size: 8
• 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: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
6.7979	4.5	500	2.4200	0.9999
0.5517	9.01	1000	0.0731	0.1567
0.1188	13.51	1500	0.0645	0.1535
0.0826	18.02	2000	0.0626	0.1528
0.0627	22.52	2500	0.0593	0.1522

Framework versions

• Transformers 4.11.3
• Pytorch 1.13.1+cu117
• Datasets 1.14.0
• Tokenizers 0.10.3