Add complete TabPFNv2 usage comments in classification#28
Add complete TabPFNv2 usage comments in classification#28emast4 wants to merge 1 commit intoPriorLabs:mainfrom
Conversation
Existing setup code fails when using TabPFNv2 because ModelVersion is never imported.
|
Codex usage limits have been reached for code reviews. Please check with the admins of this repo to increase the limits by adding credits. |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request updates the classification documentation to include commented-out instructions for using TabPFNv2. The feedback suggests importing ModelVersion directly from the main tabpfn package instead of the internal constants sub-module to adhere to public API best practices.
| # To use TabPFNv2: | ||
| # from tabpfn.constants import ModelVersion |
There was a problem hiding this comment.
It is generally recommended to use the public API for imports rather than internal sub-modules like constants. In TabPFN v2.x, ModelVersion is exported directly at the package level, so you can import it from tabpfn instead of tabpfn.constants.
Additionally, ensure that when users uncomment this, they also update the model initialization to include the desired device (e.g., device="cuda"), as the create_default_for_version method might otherwise default to CPU.
# To use TabPFNv2:
# from tabpfn import ModelVersion
1 participant
Existing setup code fails when using TabPFNv2 because ModelVersion is never imported.