Improve ONNX import book section #2059
Conversation
|
CC @SimonBrandner and @skewballfox |
antimora
changed the title
burn-book/src/import/onnx-model.md
Outdated
| 4. **Trainability**: Imported models can be further trained or fine-tuned using Burn's native | ||
| training loop. |
There was a problem hiding this comment.
using Burn's native training loop.
Though it was already there before, I find the formulation a bit weird, especially due to that last part. You can write you own training loop if you want 🤷♂️
Maybe we can just say: "Imported models can be further trained or fine-tuned with Burn."
|
@laggui I made the fixes (and more). You can view rendered difference since your last review here: |
|
@antimora sorry I was out of town, will try to take a look at this tomorrow afternoon! |
No worries! You will make it up with another upcoming big PR |
There was a problem hiding this comment.
@antimora I think your new fix addresses laggui's concerns. Rest looks good to me!
|
Hi great committers, I wonder when and how to deploy the fix of burn-book on the https://burn.dev/book . I believe this content will improve docs so I look forward to being reflected. Thank you! |
Pull Request Template
Checklist
Changes
Improved the ONNX import section with more specifics on record types and build parameters.
Testing
Rendered the book (see the rendered page here):