Dockerfile

# Use lightweight Python base image
FROM python:3.10-slim

# Set working directory
WORKDIR /app

# Install system dependencies (for PyTorch and git)
RUN apt-get update && apt-get install -y git && rm -rf /var/lib/apt/lists/*

# Copy requirements and install
COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

# Copy application code
COPY . .

# ✅ Hugging Face cache directory (use /tmp, writable in Spaces)
ENV HF_HOME=/tmp

# Pre-download model into /tmp to avoid cold start
RUN python -c "from transformers import AutoTokenizer, AutoModelForCausalLM; \
    from peft import PeftModel; \
    base_model_id='microsoft/DialoGPT-small'; \
    adapter_id='rabiyulfahim/Python_coding_model'; \
    AutoTokenizer.from_pretrained(base_model_id, cache_dir='/tmp'); \
    base_model = AutoModelForCausalLM.from_pretrained(base_model_id, cache_dir='/tmp'); \
    PeftModel.from_pretrained(base_model, adapter_id, cache_dir='/tmp')"



# Expose the correct port (Hugging Face default is 7860)
EXPOSE 7860

# Command to run the app
CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]