Executor config,Make the following changes to torch-flow.yml, butBut it still executes according to the default configuration。
kkiskkk opened this issue · comments
Calleysperx commented
Make the following changes to torch-flow.yml:
`jtype: Flow
version: '1'
with:
port: 51001
protocol: http
executors:
- name: clip_t
uses:
jtype: CLIPEncoder
metas:
py_modules:
- clip_server.executors.clip_torch
timeout_ready: 3000000
replicas: 1`
But it still executes according to the default configuration
Ziniu Yu commented
Hi, how do you start the flow? You can get the instructions from our docs https://clip-as-service.jina.ai/user-guides/server/#yaml-config
Please let me know if you have any further questions.