Rziane/whisper-large-v3-turbo-CAENNAIS

whisper-medium-CAENNAIS

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

• Loss: 0.5740
• Wer: 26.7396

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: 32
• eval_batch_size: 16
• seed: 42
• optimizer: Use OptimizerNames.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
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	1.0	56	0.7664	33.4990
No log	2.0	112	0.4936	28.0649
No log	3.0	168	0.4702	23.7906
No log	4.0	224	0.4987	28.4957
No log	5.0	280	0.4999	23.7575
No log	6.0	336	0.5567	25.3810
No log	7.0	392	0.5685	23.4924
No log	8.0	448	0.5738	25.0497
0.3662	9.0	504	0.6081	24.6852
0.3662	10.0	560	0.5740	26.7396

Framework versions

• Transformers 4.46.2
• Pytorch 2.4.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.0