Closed kitalia closed 3 months ago
headers.insert(OPENAI_BETA_HEADER, "assistants=v1".parse().unwrap());
in config.rs line 96 should be updated to
headers.insert(OPENAI_BETA_HEADER, "assistants=v2".parse().unwrap());
that would allow temperature configuration on the openai api frontend, it's an easy fix
UPDATE
it isn't an easy fix because it would make the deserialisation of the response fail due to the new response format
Error: failed deserialization of: { "id": "run_ XXX", "object": "thread.run", "created_at": 1714390612, "assistantid": "asst XXX", "threadid": "thread XXX", "status": "queued", "started_at": null, "expires_at": 1714391212, "cancelled_at": null, "failed_at": null, "completed_at": null, "required_action": null, "last_error": null, "model": "gpt-3.5-turbo-0125", "instructions": "XXX" "tools": [ { "type": "file_search" } ], "tool_resources": {}, "metadata": {}, "temperature": 1.0, "top_p": 0.9, "max_completion_tokens": null, "max_prompt_tokens": null, "truncation_strategy": { "type": "auto", "last_messages": null }, "incomplete_details": null, "usage": null, "response_format": "auto", "tool_choice": "auto" }
This is released in v0.21.0
headers.insert(OPENAI_BETA_HEADER, "assistants=v1".parse().unwrap());
in config.rs line 96 should be updated to
headers.insert(OPENAI_BETA_HEADER, "assistants=v2".parse().unwrap());
that would allow temperature configuration on the openai api frontend, it's an easy fix
UPDATE
it isn't an easy fix because it would make the deserialisation of the response fail due to the new response format