Views
No views yet
openai/whisper-medium pour la
reconnaissance automatique de la parole (ASR) en darija (dialecte marocain),
sur un ensemble combiné de datasets publics.| Métrique | Valeur |
|---|---|
| WER (Word Error Rate) | 0.0346 (3.46%) |
| CER (Character Error Rate) | 0.0135 (1.35%) |
| Stat | Valeur |
|---|---|
| Époques | 10.0 |
| Train loss (moyenne) | 0.07550942687588065 |
| Durée d'entraînement | 20:20:27 |
| Vitesse | 5.479 samples/s, 0.171 steps/s |
adiren7/darija_speech_to_textmohamedmou/moroccan-darija-asr-dataset-splitafyfbadreddine77/darija-asr-datasetntariklk/darija-merged-asratlasia/Moroccan-Darija-Wiki-Audio-DatasetDatasmartly/moroccan_darija_audio (accès refusé — dataset privé, non inclus dans l'entraînement)RHEZLOUNE/darija-restaurant-audioanaszil/Segmented-Moroccan-Darija-Wiki-Audio-Dataset1from transformers import WhisperForConditionalGeneration, WhisperProcessor
2import librosa
3import torch
4
5processor = WhisperProcessor.from_pretrained("amineouaki/whisper-medium-darija-combined")
6model = WhisperForConditionalGeneration.from_pretrained("amineouaki/whisper-medium-darija-combined")
7
8speech, _ = librosa.load("audio.wav", sr=16000)
9inputs = processor(speech, sampling_rate=16000, return_tensors="pt")
10
11with torch.no_grad():
12 generated_ids = model.generate(inputs.input_features, max_length=225)
13
14print(processor.batch_decode(generated_ids, skip_special_tokens=True)[0])