openai_models/whisper.py

from transformers import pipeline, AutoModelForCausalLM, AutoModelForSpeechSeq2Seq, AutoProcessor
import torch
import streamlit as st


@st.cache_resource
def whisper_os():
    device = "cuda:0" if torch.cuda.is_available() else "cpu"
    torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32

    assistant_model_id = "distil-whisper/distil-large-v3"

    assistant_model = AutoModelForCausalLM.from_pretrained(
        assistant_model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
    )
    assistant_model.to(device)

    model_id = "openai/whisper-large-v3"

    model = AutoModelForSpeechSeq2Seq.from_pretrained(
        model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True,
        use_safetensors=True, attn_implementation="sdpa")

    processor = AutoProcessor.from_pretrained(model_id)

    pipe = pipeline(
        "automatic-speech-recognition",
        model=model,
        tokenizer=processor.tokenizer,
        feature_extractor=processor.feature_extractor,
        max_new_tokens=128,
        generate_kwargs={"assistant_model": assistant_model},
        torch_dtype=torch_dtype,
        device=device,
    )

    return pipe