whisper-large-v3-turbo-Marathi-Version1

This model is a fine-tuned version of openai/whisper-large-v3-turbo on the common_voice_17_0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.3116
• Wer: 54.0389

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-05
• 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
0.3595	4.0	2000	0.3886	65.0826
0.3104	8.0	4000	0.3528	59.4175
0.279	12.0	6000	0.3360	57.2381
0.262	16.0	8000	0.3277	56.5516
0.2512	20.0	10000	0.3194	55.5252
0.2698	24.0	12000	0.3160	55.0920
0.2233	28.0	14000	0.3155	54.6388
0.2247	32.0	16000	0.3130	54.0256
0.2164	36.0	18000	0.3120	54.2255
0.2198	40.0	20000	0.3116	54.0389

Framework versions

• PEFT 0.14.0
• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0