munirrani/fast-whisper-medium-finetune

How to run

from transformers import (
    AutomaticSpeechRecognitionPipeline,
    WhisperForConditionalGeneration,
    WhisperTokenizer,
    WhisperProcessor,
    BitsAndBytesConfig,
)
from peft import PeftModel, PeftConfig

# Use 4-bit quantization instead of 8-bit for now
bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_compute_dtype=torch.float16,
    bnb_4bit_quant_storage=torch.float16,
    bnb_4bit_use_double_quant=True,
    # load_in_8bit=True,
)


peft_model_id = "munirrani/whisper-medium-finetune"
language = "ms"
task = "transcribe"
peft_config = PeftConfig.from_pretrained(peft_model_id)
model = WhisperForConditionalGeneration.from_pretrained(
    peft_config.base_model_name_or_path, quantization_config=bnb_config, device_map="auto"
)

model = PeftModel.from_pretrained(model, peft_model_id)
tokenizer = WhisperTokenizer.from_pretrained(peft_config.base_model_name_or_path, language=language, task=task)
processor = WhisperProcessor.from_pretrained(peft_config.base_model_name_or_path, language=language, task=task)
feature_extractor = processor.feature_extractor
forced_decoder_ids = processor.get_decoder_prompt_ids(language=language, task=task)
pipe = AutomaticSpeechRecognitionPipeline(model=model, tokenizer=tokenizer, feature_extractor=feature_extractor)

def transcribe(audio):
    with torch.cuda.amp.autocast():
        text = pipe(audio, generate_kwargs={"forced_decoder_ids": forced_decoder_ids}, max_new_tokens=255)["text"]
    return text

file_name = "your_audio_file.mp3"
transcribe(file_name)