Upload 3 files

Dockerfile

@@ -0,0 +1,28 @@
+FROM python:3.11-slim
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+        curl git ca-certificates zstd && rm -rf /var/lib/apt/lists/*
+
+RUN curl -fsSL https://ollama.com/install.sh | sh
+
+RUN useradd -m -u 1000 user
+USER user
+ENV HOME=/home/user \
+    PATH=/home/user/.local/bin:$PATH \
+    HOST=0.0.0.0 \
+    PORT=7860 \
+    OPEN_BROWSER=false \
+    OLLAMA_URL=http://localhost:11434 \
+    OLLAMA_MODEL=qwen2.5:0.5b \
+    OUTPUT_BASENAME=/tmp/papers
+
+WORKDIR /home/user/app
+
+RUN git clone --depth 1 https://github.com/RebornX10/rebornx10.github.io.git . \
+    && pip install --no-cache-dir --user -r requirements.txt
+
+COPY --chown=user start.sh /home/user/start.sh
+
+EXPOSE 7860
+
+CMD ["bash", "/home/user/start.sh"]

README.md

@@ -1,10 +1,22 @@
 ---
-title: Sci Paper Rag
-emoji: 🚀
-colorFrom: green
-colorTo: gray
+title: sci_paper_llm
+emoji: 🔬
+colorFrom: blue
+colorTo: indigo
 sdk: docker
+app_port: 7860
 pinned: false
+license: mit
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# sci_paper_llm — live demo
+
+Build open-access scientific-paper datasets from [OpenAlex](https://openalex.org) and
+query them with a local LLM, running entirely inside this Space.
+
+Source code: <https://github.com/RebornX10/rebornx10.github.io>
+
+> **Free CPU Space caveats:** the LLM runs on CPU with a small model
+> (`qwen2.5:0.5b` by default), so answers are slow and the model is pulled on
+> cold start. For full speed, use a GPU Space or run locally with Docker Compose.
+> Change the model with the `OLLAMA_MODEL` Space variable.

start.sh

@@ -0,0 +1,10 @@
+#!/usr/bin/env bash
+# sci_paper_llm — Samuel Adone (GitHub: RebornX10) — MIT
+set -e
+
+ollama serve &
+until curl -sf http://localhost:11434/api/tags >/dev/null 2>&1; do sleep 1; done
+ollama pull "${OLLAMA_MODEL}"
+
+cd /home/user/app
+exec python main.py