Upload 4 files

Dockerfile

@@ -0,0 +1,33 @@
+FROM python:3.10-slim
+
+# Install system dependencies
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    build-essential \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+WORKDIR /code
+
+# Copy requirements file
+COPY ./requirements.txt /code/requirements.txt
+
+# Upgrade pip and install requirements
+RUN pip install --no-cache-dir --upgrade pip && \
+    pip install --no-cache-dir -r /code/requirements.txt
+
+# Create and use non-root user
+RUN useradd -m user
+USER user
+
+# Set environment variables
+ENV HOME=/home/user \
+    PATH=/home/user/.local/bin:$PATH \
+    PYTHONUNBUFFERED=1
+
+WORKDIR $HOME/app
+
+# Copy application code
+COPY --chown=user . $HOME/app
+
+# Run the application
+CMD ["python", "-m", "uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

api.py

@@ -0,0 +1 @@
+from app import app

app.py

@@ -0,0 +1,73 @@
+import os
+import logging
+from fastapi import FastAPI, HTTPException
+from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
+from peft import PeftModel, PeftConfig
+
+# Set up logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+# Initialize FastAPI app
+app = FastAPI()
+
+# Global variables for model, tokenizer, and pipeline
+model = None
+tokenizer = None
+pipe = None
+
+@app.on_event("startup")
+async def load_model():
+    global model, tokenizer, pipe
+    
+    try:
+        # Get Hugging Face token from environment variable
+        hf_token = os.environ.get("HUGGINGFACE_TOKEN")
+        
+        logger.info("Loading PEFT configuration...")
+        config = PeftConfig.from_pretrained("frankmorales2020/Mistral-7B-text-to-sql-flash-attention-2-dataeval")
+        
+        logger.info("Loading base model...")
+        base_model = AutoModelForCausalLM.from_pretrained(
+            "mistralai/Mistral-7B-Instruct-v0.3", 
+            token=hf_token if hf_token else None,
+            use_auth_token=True if not hf_token else None
+        )
+        
+        logger.info("Loading PEFT model...")
+        model = PeftModel.from_pretrained(base_model, "frankmorales2020/Mistral-7B-text-to-sql-flash-attention-2-dataeval")
+
+        logger.info("Loading tokenizer...")
+        tokenizer = AutoTokenizer.from_pretrained(
+            "mistralai/Mistral-7B-Instruct-v0.3", 
+            token=hf_token if hf_token else None,
+            use_auth_token=True if not hf_token else None
+        )
+
+        logger.info("Creating pipeline...")
+        pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer)
+        
+        logger.info("Model, tokenizer, and pipeline loaded successfully.")
+    except Exception as e:
+        logger.error(f"Error loading model or creating pipeline: {e}")
+        raise
+
+@app.get("/")
+def home():
+    return {"message": "Hello World"}
+
+@app.get("/generate")
+async def generate(text: str):
+    if not pipe:
+        raise HTTPException(status_code=503, detail="Model not loaded")
+    
+    try:
+        output = pipe(text, max_length=100, num_return_sequences=1)
+        return {"output": output[0]['generated_text']}
+    except Exception as e:
+        logger.error(f"Error during text generation: {e}")
+        raise HTTPException(status_code=500, detail=f"Error during text generation: {str(e)}")
+
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=7860)

requirements.txt

@@ -0,0 +1,9 @@
+fastapi==0.103.0
+uvicorn[standard]==0.17.*
+torch>=1.13.0
+transformers>=4.34.0,<4.35.0
+numpy<2
+peft>=0.6.0,<0.7.0
+accelerate>=0.24.1,<0.25.0
+huggingface_hub>=0.16.4,<0.18.0
+tokenizers>=0.14.0,<0.15.0