whisper-large-v3-turbo-Hindi-Version1

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the fleurs dataset. It achieves the following results on the evaluation set:

• Loss: 0.2303
• Wer: 26.1421

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-06
• train_batch_size: 8
• 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: 1000
• training_steps: 20000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.2701	6.7797	2000	0.2583	28.8462
0.2499	13.5593	4000	0.2446	27.6552
0.2309	20.3390	6000	0.2397	27.1769
0.2379	27.1186	8000	0.2364	26.8059
0.2215	33.8983	10000	0.2343	26.4252
0.226	40.6780	12000	0.2322	26.5228
0.2201	47.4576	14000	0.2314	25.9274
0.2263	54.2373	16000	0.2308	26.4740
0.2154	61.0169	18000	0.2307	26.3862
0.2357	67.7966	20000	0.2303	26.1421

Framework versions

• PEFT 0.14.0
• Transformers 4.46.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.20.1