Whisper Large v3 French PraxySante - Fine-tuned
This model is a fine-tuned version of bofenghuang/whisper-large-v3-french on the PxCorpus PxSLU, the MediaSpeech, the Multilingual TedX Fr and the Common Voice 17 datasets. It achieves the following results on the evaluation set:
- Loss: 0.6630
- Wer: 27.7159
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: 8
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 8
- gradient_accumulation_steps: 2
- total_train_batch_size: 128
- total_eval_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 3.0
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
0.5569 | 1.6129 | 25 | 0.6630 | 27.7159 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.3.0
- Datasets 2.21.0
- Tokenizers 0.19.1
- Downloads last month
- 3
Inference API (serverless) is not available, repository is disabled.
Model tree for FILALIHicham/praxy-whisper-large-v3-fr-2
Base model
bofenghuang/whisper-large-v3-french
Finetuned
this model
Datasets used to train FILALIHicham/praxy-whisper-large-v3-fr-2
Evaluation results
- Wer on PxCorpus PxSLUself-reported27.716
- Wer on MediaSpeechself-reported27.716
- Wer on Multilingual TedX Frself-reported27.716
- Wer on Common Voice 17self-reported27.716