CXDuncan/whisper-small-malayalam

Whisper Small Malayalam

This model is a fine-tuned version of openai/whisper-small on the Malayalam-IndicVoices dataset. It achieves the following results on the evaluation set:

• Loss: 0.0003
• Wer: 51.5300

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: 1e-05
• train_batch_size: 16
• 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: 500
• training_steps: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0665	5.0	1000	0.0446	67.4679
0.0099	10.0	2000	0.0064	57.3925
0.0007	15.0	3000	0.0007	51.2762
0.0003	20.0	4000	0.0003	51.5300

Framework versions

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