Create entrypoint.sh

entrypoint.sh

@@ -0,0 +1,61 @@
+#!/bin/bash
+set -e
+
+echo "============================================"
+echo "  Starting Ollama on port 7860..."
+echo "============================================"
+
+ollama serve &
+OLLAMA_PID=$!
+
+echo "Waiting for Ollama to be ready..."
+max_attempts=120
+attempt=0
+until curl -sf http://localhost:7860/api/tags > /dev/null 2>&1; do
+    sleep 1
+    attempt=$((attempt + 1))
+    if [ $attempt -ge $max_attempts ]; then
+        echo "ERROR: Ollama failed to start after ${max_attempts}s"
+        exit 1
+    fi
+    echo "  attempt ${attempt}/${max_attempts}..."
+done
+
+echo "Ollama is running!"
+echo ""
+
+echo "============================================"
+echo "  Pulling models:"
+echo "    - qwen3:0.6b"
+echo "    - qwen3-embedding:0.6b"
+echo "============================================"
+
+ollama pull qwen3:0.6b
+ollama pull qwen3-embedding:0.6b
+
+echo ""
+echo "============================================"
+echo "  Warming up qwen3:0.6b..."
+echo "============================================"
+
+curl -sf http://localhost:7860/api/generate \
+  -d '{"model":"qwen3:0.6b","prompt":"hi","stream":false}' > /dev/null 2>&1 || true
+
+echo ""
+echo "============================================"
+echo "  Warming up qwen3-embedding:0.6b..."
+echo "============================================"
+
+curl -sf http://localhost:7860/api/embeddings \
+  -d '{"model":"qwen3-embedding:0.6b","prompt":"hello"}' > /dev/null 2>&1 || true
+
+echo ""
+echo "============================================"
+echo "  READY!"
+echo "  Base URL: http://localhost:7860"
+echo "  Models:"
+echo "    - qwen3:0.6b"
+echo "    - qwen3-embedding:0.6b"
+echo "============================================"
+
+wait $OLLAMA_PID