carefree0910
carefree0910
> Update: I dig into it a little bit and found that the `GraphicsData` parsed by `SVGScene` will contain several `points` which is `NaN`, which is likely the reason that...
### Expected Behavior No overlapping area at join ### Current Behavior ### Possible Solution Not sure, this might be something deeeeep XD > A workaround is to draw lines without...
**EDIT** I'm able to fix this by assigning `width` and `height` to the `canvas` in the first fiddle, but still not sure why this is happening XD --- I'm trying...
First of all, thanks for your great work! And I found out that the `ln_pre` seems to be a redundant component: https://github.com/openai/CLIP/blob/3b473b0e682c091a9e53623eebc1ca1657385717/clip/model.py#L212 Because you've already applied `pre_norm` mechanism in `ResidualAttentionBlock`....
Minimal buggy svg: ```svg ```
First of all, thanks for your amazing work! And it seems that your `TransformerEncoderLayer` implementation is a bit different from the 'mainstream' implementations, because you create your residual link **after**...
## Willingness to contribute The MLflow Community encourages new feature contributions. Would you or another member of your organization be willing to contribute an implementation of this feature (either as...