Closed hsuyuming closed 5 months ago
@jsondai Do you have any idea regarding for this one?
I can increate output token, but i don't think that is a long term solution for user
Hi, this is a bug that we plan to fix soon. The Rapid Eval SDK can continue the evaluation process with partially generated response with a warning message, instead of throwing an error.
File Name
score_and_select_models_rapid_evaluation_sdk.ipynb
What happened?
When I execute "Running evaluation", I get a RuntimeError error. After further investigation, this error occurs because when we set max_output_tokens to 128 for Gemini, it returns MAX_TOKENS as the finish reason based on reaching the output token limit. Based on line [1], if the finish reason is not STOP and FINISH_REASON_UNSPECIFIED, then the SDK will throw an exception.
[1] https://github.com/googleapis/python-aiplatform/blob/main/vertexai/preview/evaluation/_evaluation.py#L220
Relevant log output
Code of Conduct