feat(rag): add OpenVINO NPU embedding services
This commit is contained in:
@@ -0,0 +1,16 @@
|
||||
[Unit]
|
||||
Description=Obsidian Vault Reindex Endpoint
|
||||
After=network.target
|
||||
|
||||
[Service]
|
||||
Type=simple
|
||||
ExecStart=/usr/bin/python3 /home/will/lab/swarm/scripts/obsidian-reindex-server.py
|
||||
Restart=on-failure
|
||||
RestartSec=5
|
||||
Environment=PORT=18810
|
||||
Environment=RAG_COLLECTION=obsidian_bge_npu
|
||||
Environment=RAG_EMBED_MODEL=bge-base-en-v1.5-int8-ov
|
||||
Environment=OLLAMA_BASE_URL=http://127.0.0.1:18817
|
||||
|
||||
[Install]
|
||||
WantedBy=default.target
|
||||
@@ -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
|
||||
@@ -0,0 +1,16 @@
|
||||
[Unit]
|
||||
Description=RAG/Embedding Health HTTP Service (port 18814)
|
||||
After=network.target
|
||||
|
||||
[Service]
|
||||
Type=simple
|
||||
ExecStart=/usr/bin/python3 /home/will/lab/swarm/scripts/rag-embedding-health-server.py
|
||||
Restart=on-failure
|
||||
RestartSec=5
|
||||
Environment=PORT=18814
|
||||
Environment=RAG_HEALTH_TIMEOUT=180
|
||||
Environment=OLLAMA_BASE_URL=http://127.0.0.1:18817
|
||||
Environment=RAG_EMBED_MODEL=bge-base-en-v1.5-int8-ov
|
||||
|
||||
[Install]
|
||||
WantedBy=default.target
|
||||
Reference in New Issue
Block a user