app.py

# app.py
from fastapi import FastAPI, HTTPException, UploadFile, File
from pydantic import BaseModel
from llama_cpp import Llama
from typing import Optional
import uvicorn
import huggingface_hub
import os
from PIL import Image
import io
import base64

app = FastAPI(
    title="OmniVLM API",
    description="API for text and image processing using OmniVLM model",
    version="1.0.0"
)

# Download the model from Hugging Face Hub
model_path = huggingface_hub.hf_hub_download(
    repo_id="NexaAIDev/OmniVLM-968M",
    filename="omnivision-text-optimized-llm-Q8_0.gguf"
)

# Initialize the model with the downloaded file
llm = Llama(
    model_path=model_path,
    n_ctx=2048,
    n_threads=4,
    n_batch=512,
    verbose=True
)

class GenerationRequest(BaseModel):
    prompt: str
    max_tokens: Optional[int] = 100
    temperature: Optional[float] = 0.7
    top_p: Optional[float] = 0.9

class ImageRequest(BaseModel):
    prompt: Optional[str] = "Describe this image in detail"
    max_tokens: Optional[int] = 200
    temperature: Optional[float] = 0.7

class GenerationResponse(BaseModel):
    generated_text: str

@app.post("/generate", response_model=GenerationResponse)
async def generate_text(request: GenerationRequest):
    try:
        output = llm(
            request.prompt,
            max_tokens=request.max_tokens,
            temperature=request.temperature,
            top_p=request.top_p
        )
        
        return GenerationResponse(generated_text=output["choices"][0]["text"])
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/process-image", response_model=GenerationResponse)
async def process_image(
    file: UploadFile = File(...),
    request: ImageRequest = None
):
    try:
        # Read and validate the image
        image_data = await file.read()
        image = Image.open(io.BytesIO(image_data))
        
        # Convert image to base64
        buffered = io.BytesIO()
        image.save(buffered, format=image.format or "JPEG")
        img_str = base64.b64encode(buffered.getvalue()).decode()
        
        # Create prompt with image
        prompt = f"""
        <image>data:image/jpeg;base64,{img_str}</image>
        {request.prompt if request else "Describe this image in detail"}
        """
        
        # Generate description
        output = llm(
            prompt,
            max_tokens=request.max_tokens if request else 200,
            temperature=request.temperature if request else 0.7
        )
        
        return GenerationResponse(generated_text=output["choices"][0]["text"])
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))

@app.get("/health")
async def health_check():
    return {"status": "healthy"}

if __name__ == "__main__":
    port = int(os.environ.get("PORT", 7860))
    uvicorn.run(app, host="0.0.0.0", port=port)