Create Dockerfile

Dockerfile

@@ -0,0 +1,74 @@
+# Use a lightweight Python base
+FROM python:3.10-slim
+
+# Install system dependencies for OpenVINO and Image processing
+RUN apt-get update && apt-get install -y \
+    libgl1-mesa-glx libglib2.0-0 git \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install optimized CPU libraries
+# optimum[openvino] is the key for fast CPU inference
+RUN pip install --no-cache-dir \
+    flask flask-cors requests \
+    "optimum[openvino,diffusers]" \
+    transformers accelerate torch --extra-index-url https://download.pytorch.org/whl/cpu
+
+# Set up environment
+ENV HOME=/home/user
+WORKDIR $HOME
+RUN mkdir -p $HOME/.cache && chmod -R 777 $HOME
+
+# --- 1. The Python Guard + Inference Script ---
+RUN cat <<EOF > $HOME/app.py
+from flask import Flask, request, jsonify, send_file
+from optimum.intel import OVStableDiffusionPipeline
+from flask_cors import CORS
+import json, os, datetime, io, torch
+
+app = Flask(__name__)
+CORS(app)
+
+DB_PATH = "/home/user/usage.json"
+WL_PATH = "/home/user/whitelist.txt"
+LIMIT = 500
+UNLIMITED_KEY = "sk-ess4l0ri37"
+
+# Load Model Optimized for CPU
+# Using SD 1.5 (Small/Fast) converted to OpenVINO format
+print("Loading Optimized CPU Model...")
+model_id = "heifai/stable-diffusion-v1-5-openvino" 
+pipe = OVStableDiffusionPipeline.from_pretrained(model_id, compile=True)
+print("Model Ready.")
+
+def get_whitelist():
+    if not os.path.exists(WL_PATH):
+        with open(WL_PATH, "w") as f:
+            f.write(f"{UNLIMITED_KEY}\n")
+        return {UNLIMITED_KEY}
+    with open(WL_PATH, "r") as f:
+        return set(line.strip() for line in f.readlines())
+
+@app.route("/api/generate", methods=["POST"])
+def generate():
+    user_key = request.headers.get("x-api-key", "")
+    if user_key not in get_whitelist():
+        return jsonify({"error": "Unauthorized"}), 401
+
+    # Add your usage tracking logic here (from your Ollama script)
+
+    data = request.json
+    prompt = data.get("prompt", "a simple cat")
+    
+    # CPU Optimization: Fewer steps for speed
+    image = pipe(prompt, num_inference_steps=20).images[0]
+    
+    img_io = io.BytesIO()
+    image.save(img_io, 'PNG')
+    img_io.seek(0)
+    return send_file(img_io, mimetype='image/png')
+
+if __name__ == "__main__":
+    app.run(host="0.0.0.0", port=7860)
+EOF
+
+ENTRYPOINT ["python3", "app.py"]