Ellight/whisper-tiny-hi

Whisper tiny Hindi

This model is a fine-tuned version of openai/whisper-tiny on the Common Voice 17.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.7189

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.001
• train_batch_size: 32
• 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: 50
• num_epochs: 1
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.575	0.4484	100	0.8603
0.6631	0.8969	200	0.7189

Framework versions

• PEFT 0.11.2.dev0
• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.3.dev0
• Tokenizers 0.19.1