muhtasham/whisper-tg

Whisper Tajik

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

• Loss: 0.4538
• Wer: 18.9518

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: Use 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: 4000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0245	6.2893	1000	0.3726	20.9634
0.0043	12.5786	2000	0.4167	20.5318
0.0003	18.8679	3000	0.4431	19.2062
0.0002	25.1572	4000	0.4538	18.9518

Framework versions

• Transformers 4.49.0.dev0
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0