wav2vec2-large-xls-r-300m-dv

This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.4688
• Wer: 0.5420

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.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: 500
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
4.7748	2.5993	400	0.9977	0.8406
0.5528	5.1954	800	0.5006	0.6392
0.2781	7.7948	1200	0.4461	0.5797
0.1713	10.3909	1600	0.4591	0.5631
0.118	12.9902	2000	0.4688	0.5420

Framework versions

• Transformers 4.55.1
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.4