-
-
Notifications
You must be signed in to change notification settings - Fork 3.9k
Description
Is there an existing issue for this?
- Yes, I have searched the existing issues and none of them match my problem.
Product Variant
Self-Hosted
Describe the bug you're experiencing
Hello there,
I'm rather new to the field of resume generators and especially new to hosting LLMs, so please consider this question as newbish. :)
I was trying to use the import feature and go on from there, but even testing the LLM connection drives me crazy currently.
For now I'm using latest Lemonade-server with a Radeon 9070XT, tried the OpenAI and Ollama API, different ports, on Win11 and WSL ubuntu 24.04 and tried a couple different available models (i.e. gemma-3-4b-it-GGUF, gpt-oss-20b, etc. ) but I always get the same error, even though from the lemonade logs the request has been answered fine with 200.
Error:
ERROR [oRPC]: Error: No object generated: response did not match schema.
I never tried a public LLM provider because I want to stay local with my sensitive data.
Are there any strict requirements to the model being available that the available OSS models can't fulfill?
Any help or hint is appreciated and thanks.
Best regards
Benjamin
PS:
chatgpt info:
The error message "ERROR [oRPC]: Error: No object generated: response did not match schema." typically indicates that a response from an operation did not conform to the expected format or structure (schema). This often occurs in programming or API interactions when:
Data Structure Mismatch: The returned data does not align with the defined schema, meaning it might be missing required fields or have fields that are of the wrong type.
Validation Error: The system is unable to validate the response due to rules defined in the schema, leading to rejection of the data.
Unexpected Input: The operation could have received unexpected or malformed input data, causing an invalid response.
What template are you using?
None
Anything else?
No response