Initial commit: Add Dockerfile, app.py, and requirements.txt

Dockerfile

@@ -0,0 +1,61 @@
+# ----------------------------------------------------------
+# 1. Base image
+# ----------------------------------------------------------
+FROM ubuntu:22.04
+
+# Avoid interactive prompts during build
+ENV DEBIAN_FRONTEND=noninteractive
+
+# ----------------------------------------------------------
+# 2. Install dependencies
+# ----------------------------------------------------------
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    wget \
+    curl \
+    ca-certificates \
+    git \
+    build-essential \
+    python3 \
+    python3-pip \
+    && rm -rf /var/lib/apt/lists/*
+
+# ----------------------------------------------------------
+# 3. Install Ollama
+#    Check https://github.com/jmorganca/ollama/releases for the latest version
+# ----------------------------------------------------------
+ARG OLLAMA_VERSION="v0.0.14"
+RUN wget -q \
+    https://github.com/jmorganca/ollama/releases/download/${OLLAMA_VERSION}/ollama-linux-x64 \
+    -O /usr/local/bin/ollama && \
+    chmod +x /usr/local/bin/ollama
+
+# ----------------------------------------------------------
+# 4. Copy Python requirements and install
+# ----------------------------------------------------------
+WORKDIR /app
+COPY requirements.txt /app/requirements.txt
+RUN pip3 install --no-cache-dir -r requirements.txt
+
+# ----------------------------------------------------------
+# 5. Copy application files
+# ----------------------------------------------------------
+COPY app.py /app/app.py
+
+# ----------------------------------------------------------
+# 6. Environment variables & expose port
+# ----------------------------------------------------------
+# By default, set an example API key (override at runtime/Secrets)
+ENV OLLAMA_API_KEY=change_me
+
+# We must expose 7860 for Hugging Face Spaces
+EXPOSE 7860
+
+# ----------------------------------------------------------
+# 7. Start Ollama in background and then run the FastAPI server
+# ----------------------------------------------------------
+# The model is pulled from HF: abanm/Dubs-Q8_0-GGUF
+#
+# "ollama serve" runs on default port 11411. We'll run that in the background,
+# then run uvicorn on 7860.
+CMD [ "sh", "-c", "ollama serve --model https://huggingface.co/abanm/Dubs-Q8_0-GGUF & uvicorn app:app --host 0.0.0.0 --port 7860" ]
+ "https://huggingface.co/your-account/your-model"]

app.py

@@ -0,0 +1,47 @@
+import os
+import subprocess
+
+from fastapi import FastAPI, Request, HTTPException
+import uvicorn
+
+app = FastAPI()
+
+# Load your API key from environment (set this in HF Secrets for security)
+OLLAMA_API_KEY = os.environ.get("OLLAMA_API_KEY", "change_me")
+
+@app.post("/generate")
+async def generate(request: Request):
+    """Endpoint that generates text based on the prompt."""
+    # 1. Check API key
+    auth_header = request.headers.get("Authorization")
+    if not auth_header or not auth_header.startswith("Bearer "):
+        raise HTTPException(status_code=401, detail="Missing or invalid Authorization header")
+
+    token = auth_header.split(" ")[1]
+    if token != OLLAMA_API_KEY:
+        raise HTTPException(status_code=401, detail="Invalid API key")
+
+    # 2. Parse the request JSON
+    body = await request.json()
+    prompt_text = body.get("prompt", "")
+    if not prompt_text:
+        raise HTTPException(status_code=400, detail="No prompt provided")
+
+    # 3. Option A: Call Ollama via its CLI directly
+    #    (The model is served in the background via `ollama serve`)
+    process = subprocess.Popen(
+        ["ollama", "run", prompt_text],
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        text=True
+    )
+    output, error = process.communicate()
+
+    if process.returncode != 0 or error:
+        raise HTTPException(status_code=500, detail=f"Ollama error: {error.strip()}")
+
+    return {"response": output.strip()}
+
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)

requirements.txt

@@ -0,0 +1,3 @@
+fastapi
+uvicorn
+httpx