wav2vec2-base-timit-demo-colab

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

• Loss: 0.4028
• Wer: 0.3533

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
2.7434	1.0040	500	1.3280	1.0321
0.8808	2.0080	1000	0.5326	0.5339
0.4891	3.0120	1500	0.4431	0.4541
0.3335	4.0161	2000	0.4053	0.4195
0.2476	5.0201	2500	0.4087	0.3998
0.1918	6.0241	3000	0.3971	0.3884
0.155	7.0281	3500	0.4242	0.3697
0.1249	8.0321	4000	0.4094	0.3599
0.1045	9.0361	4500	0.4028	0.3533

Framework versions

• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 1.18.3
• Tokenizers 0.20.3