paulinenyaboe/swahili_zen_model
Domaine:
natural language processingType de record:
dataset
Créateur:
pau
Hôte:
model:
huggingface.co
from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor
import torch
import librosa
import numpy as np
model_id = "zenlm/zen3-asr"
processor = AutoProcessor.from_pretrained(model_id)
model = AutoModelForSpeechSeq2Seq.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")
def transcribe(audio_path):
audio, sr = librosa.load(audio_path, sr=16000)