Initial scaffold: LLM Trainer Dashboard

Full-stack app with FastAPI backend (SSH/paramiko, pipeline streaming,
GPU stats, xterm.js terminal, Ollama model manager) and React + Tailwind
frontend (8 panels: Connection, Documents, Pipeline, QA Pairs, Training,
Terminal, Models, Config). Docker Compose included.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

docker-compose.yml

@@ -0,0 +1,29 @@
+version: "3.9"
+
+services:
+
+  backend:
+    build: ./backend
+    container_name: llm-trainer-backend
+    restart: unless-stopped
+    ports:
+      - "8080:8080"
+    environment:
+      - OLLAMA_URL=http://192.168.2.47:11434
+    networks:
+      - llm-net
+
+  frontend:
+    build: ./frontend
+    container_name: llm-trainer-frontend
+    restart: unless-stopped
+    ports:
+      - "3000:80"
+    depends_on:
+      - backend
+    networks:
+      - llm-net
+
+networks:
+  llm-net:
+    driver: bridge