smrc/new-whisper-small-fr-qc

Whisper Small FR QC

This model is a fine-tuned version of openai/whisper-small on the Linguistique 1.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0002
• Wer: 0.1776

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: 16
• 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: 500
• training_steps: 5000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.0171	6.3694	1000	0.0153	1.2211
0.0007	12.7389	2000	0.0006	0.1332
0.0003	19.1083	3000	0.0003	0.1332
0.0002	25.4777	4000	0.0002	0.1776
0.0002	31.8471	5000	0.0002	0.1776

Framework versions

• Transformers 4.45.2
• Pytorch 2.0.0+cu118
• Datasets 3.0.1
• Tokenizers 0.20.1