app.py

import os
os.environ["TRANSFORMERS_CACHE"] = "/tmp/hf-cache"  # Important for Docker

from fastapi import FastAPI, UploadFile, File
from fastapi.middleware.cors import CORSMiddleware
import torchaudio
import torch
from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
import io

app = FastAPI()

# Allow all origins (for Flutter)
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_methods=["*"],
    allow_headers=["*"],
)

# Load model
processor = Wav2Vec2Processor.from_pretrained("Mustafaa4a/ASR-Somali")
model = Wav2Vec2ForCTC.from_pretrained("Mustafaa4a/ASR-Somali")

@app.get("/")
async def root():
    return {"message": "Somali Speech-to-Text API is running."}

@app.post("/transcribe")
async def transcribe(file: UploadFile = File(...)):
    audio_bytes = await file.read()
    audio_stream = io.BytesIO(audio_bytes)
    
    waveform, sample_rate = torchaudio.load(audio_stream)

    if sample_rate != 16000:
        resampler = torchaudio.transforms.Resample(orig_freq=sample_rate, new_freq=16000)
        waveform = resampler(waveform)

    inputs = processor(waveform.squeeze(), sampling_rate=16000, return_tensors="pt")
    
    with torch.no_grad():
        logits = model(**inputs).logits

    predicted_ids = torch.argmax(logits, dim=-1)
    transcription = processor.decode(predicted_ids[0])
    return {"transcription": transcription}