feat(rag): add OpenVINO NPU embeddings service

This commit is contained in:
William Valentin
2026-06-03 18:28:16 -07:00
parent 7745648a13
commit fe4dea0f07
7 changed files with 268 additions and 14 deletions
+19
View File
@@ -0,0 +1,19 @@
[Unit]
Description=OpenVINO NPU Embeddings HTTP Service (port 18817)
After=network.target
[Service]
Type=simple
WorkingDirectory=/home/will/lab/swarm
ExecStart=/home/will/.venvs/npu/bin/python /home/will/lab/swarm/scripts/openvino-embeddings-server.py
Restart=on-failure
RestartSec=5
Environment=OPENVINO_EMBED_PORT=18817
Environment=OPENVINO_EMBED_HOST=0.0.0.0
Environment=OPENVINO_EMBED_DEVICE=NPU
Environment=OPENVINO_EMBED_MODEL=bge-base-en-v1.5-int8-ov
Environment=OPENVINO_EMBED_MODEL_DIR=/home/will/.cache/openvino-models/bge-base-en-v1.5-int8-ov
Environment=OPENVINO_EMBED_MAX_LENGTH=512
[Install]
WantedBy=default.target