arun100/whisper-base-fr-derived-2

Whisper Base French

This model is a fine-tuned version of qanastek/whisper-base-french-cased on the google/fleurs fr_fr dataset. It achieves the following results on the evaluation set:

• Loss: 0.5402
• Wer: 23.7955

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-07
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.3835	105.26	1000	0.5892	25.4237
0.2837	210.53	2000	0.5526	23.8955
0.2323	315.79	3000	0.5432	24.0122
0.1961	421.05	4000	0.5402	23.7955
0.1863	526.32	5000	0.5395	23.7955

Framework versions

• Transformers 4.37.0.dev0
• Pytorch 2.1.2+cu121
• Datasets 2.16.2.dev0
• Tokenizers 0.15.0