wav2vec2-xlsr-ne-en-finetuned

This model is a fine-tuned version of facebook/wav2vec2-large-xlsr-53 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7203
• Wer: 0.3468

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: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.4721	4.9958	1184	0.9764	0.6322
0.6	9.9916	2368	0.6466	0.4239
0.428	14.9873	3552	0.6566	0.4119
0.3331	19.9831	4736	0.6387	0.3603
0.2801	24.9789	5920	0.6760	0.3490
0.2493	29.9747	7104	0.7004	0.3415

Framework versions

• Transformers 4.57.3
• Pytorch 2.9.1+cu128
• Datasets 2.21.0
• Tokenizers 0.22.1