whisper-large-v3-Tamil-Version1

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

• Loss: 0.2299
• Wer: 40.1989

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• 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
0.2648	5.8309	2000	0.2695	46.8731
0.2368	11.6618	4000	0.2503	45.3660
0.2151	17.4927	6000	0.2414	43.2643
0.2121	23.3236	8000	0.2367	41.9315
0.2069	29.1545	10000	0.2339	40.9165
0.2038	34.9854	12000	0.2322	40.7115
0.1936	40.8163	14000	0.2309	40.6807
0.1871	46.6472	16000	0.2304	40.4142
0.1901	52.4781	18000	0.2298	40.3014
0.1885	58.3090	20000	0.2299	40.1989

Framework versions

• PEFT 0.12.1.dev0
• Transformers 4.45.0.dev0
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1