fine-tune-wav2vec2-large-xls-r-300m-xty_224s

This model was trained from scratch on the ml-superb-subset dataset. It achieves the following results on the evaluation set:

• Loss: 0.7658
• Wer: 1.0585

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: 4
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 40
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.5419	5.5172	400	0.4368	0.9994
0.423	11.0345	800	0.4315	1.0
0.3795	16.5517	1200	0.3892	1.0151
0.3306	22.0690	1600	0.4055	1.0013
0.2464	27.5862	2000	0.4672	1.0421
0.1454	33.1034	2400	0.6656	1.0333
0.0883	38.6207	2800	0.7658	1.0585

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1