wav2vec2-xls-r-300m-malayalam-colab-CV17.0

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: 1.6154
• Wer: 1.0029
• Cer: 0.4254

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.15
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
15.7515	3.1496	200	7.2856	1.0	1.0
5.2078	6.2992	400	3.9581	1.0	1.0
3.6268	9.4488	600	3.4876	1.0	0.9923
3.4082	12.5984	800	3.3891	1.0	0.9906
3.3259	15.7480	1000	3.3171	0.9984	0.9415
3.0224	18.8976	1200	2.6551	1.0	0.7845
2.1063	22.0472	1400	1.9206	0.9942	0.4722
1.564	25.1969	1600	1.6999	0.9916	0.4298
1.3323	28.3465	1800	1.6358	0.9990	0.4264
1.2413	31.4961	2000	1.6154	1.0029	0.4254

Framework versions

• Transformers 4.42.4
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1