base

This model is a fine-tuned version of facebook/wav2vec2-large-960h on the timit_asr dataset. It achieves the following results on the evaluation set:

• Loss: 0.3835
• Cer: 0.1155

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: 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: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
3.5927	4.0	500	0.4652	0.1633
0.5377	8.0	1000	0.3505	0.1337
0.3763	12.0	1500	0.3568	0.1272
0.2966	16.0	2000	0.3485	0.1219
0.2489	20.0	2500	0.3728	0.1184
0.2133	24.0	3000	0.3713	0.1199
0.1808	28.0	3500	0.3835	0.1155

Framework versions

• Transformers 4.17.0
• Pytorch 2.4.0
• Datasets 1.18.3
• Tokenizers 0.20.3