Open alejandroballestae opened 1 month ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
View / edit / reply to this conversation on ReviewNB
RajeshThallam commented on 2024-10-14T14:20:48Z ----------------------------------------------------------------
Add links to open the notebook in Colab, Workbench, Colab Enterprise and Github. Follow this notebook for example.
View / edit / reply to this conversation on ReviewNB
RajeshThallam commented on 2024-10-14T14:20:49Z ----------------------------------------------------------------
Can you test with 3.10 or above? Most env are now upgraded to 3.10.
View / edit / reply to this conversation on ReviewNB
RajeshThallam commented on 2024-10-14T14:20:50Z ----------------------------------------------------------------
We are moving away from "Rapid Evaluation" branding and using "Gen AI Evaluation Service". Please update accordingly.
Please add references to product documentation where appropriate.
View / edit / reply to this conversation on ReviewNB
RajeshThallam commented on 2024-10-14T14:20:50Z ----------------------------------------------------------------
Please update module name with new GA version: google-cloud-aiplatform[evaluation]
Test with 1.70.0 version of Vertex AI SDK
View / edit / reply to this conversation on ReviewNB
RajeshThallam commented on 2024-10-14T14:20:51Z ----------------------------------------------------------------
remove references to project ids. Please use [your-project-id]
ot something similar.
View / edit / reply to this conversation on ReviewNB
RajeshThallam commented on 2024-10-14T14:20:52Z ----------------------------------------------------------------
typo contex --> context
The notebook reference does not exist anymore after migration to GA version. Please update accordingly.
This notebook explores the application of statistical/tendency methods to gain insights into the behavior of the Vertex AI Python SDK for Rapid Evaluation for generated answers during Question Answering (QA) tasks. Due to the inherent variability, non-deterministic or probabilistic outcomes in these tasks, relying on a single execution may not provide a comprehensive understanding of the real performance.