xiaowei
2a350cf781
Agnes 2.0-flash是推理模型，推理token占大量预算。 800仍不够→JSON截断(parse error: unexpected end of JSON input)。 1200确保推理+正文都够用。
14 lines
266 B
Desktop File
14 lines
266 B
Desktop File
[Unit]
|
|
Description=BGE-M3 Embedding Server (ONNX Runtime)
|
|
After=network.target
|
|
|
|
[Service]
|
|
Type=simple
|
|
User=muc
|
|
ExecStart=/usr/bin/python3 /home/muc/projects/memoryweave/deploy/bge_embed_server.py
|
|
Restart=on-failure
|
|
RestartSec=5
|
|
|
|
[Install]
|
|
WantedBy=multi-user.target
|