Aias00 opened a new issue, #813:
URL: https://github.com/apache/rocketmq-dashboard/issues/813

   ## Problem
   
   The Studio LLM settings UI allows users to configure advanced model 
parameters such as `maxTokens` and `temperature`, and sends them to the backend 
LLM config API. However, `LlmConfigService.saveConfig(...)` only persists 
provider, API key, model, and base URL into `GeneralSettingsVO`.
   
   After the process restarts, or after the runtime override is rebuilt from 
general settings, `fromGeneralSettings(...)` reconstructs the LLM config with 
hard-coded defaults:
   
   ```java
   .maxTokens(DEFAULT_MAX_TOKENS)
   .temperature(DEFAULT_TEMPERATURE)
   ```
   
   This makes non-default advanced settings look saved at runtime but not 
actually durable.
   
   ## Expected behavior
   
   Saving LLM config should persist `maxTokens` and `temperature`, and loading 
config from general settings should restore those saved values with the same 
default fallback used today.
   
   ## Scope
   
   Track 3 / AI Native. This only covers OpenAI-compatible advanced model 
parameters that are already part of `LlmConfigVO`; provider-specific fields can 
be handled separately.
   
   ## Evidence
   
   - 
`server/src/main/java/org/apache/rocketmq/studio/ops/ai/LlmConfigService.java`
   - 
`server/src/main/java/org/apache/rocketmq/studio/settings/GeneralSettingsVO.java`
   - 
`server/src/main/java/org/apache/rocketmq/studio/settings/GeneralSettingsUpdateDTO.java`
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to