Skip to content

Update cache system guide#910

Merged
michaelbenayoun merged 11 commits intomainfrom
neuron_model_cache_update
Jul 22, 2025
Merged

Update cache system guide#910
michaelbenayoun merged 11 commits intomainfrom
neuron_model_cache_update

Conversation

@michaelbenayoun
Copy link
Member

What does this PR do?

As per title.

@HuggingFaceDocBuilderDev

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

**Note: it is not available for models exported using any other NeuronModelXX classes, that use a different export mechanism.**
After reading this guide, you'll understand:
- How the Neuron model cache system works and its benefits
- The difference between training cache (Trainium) and inference cache (Inferentia)
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I would not mention Trainium and Inferentia here: you could do inference on Trainium and small model's training on Inferentia, I would not confuse hte hardware and the scope.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Good point, addressed that.

Comment on lines +109 to +112
- **Operates at runtime**: Caches compiled graphs as they're generated during training
- **Uses context managers**: Automatically patches the Neuron compiler to intercept and cache compilations
- **Supports private repos**: You can create and use private cache repositories
- **Uploads automatically**: New compilations are automatically uploaded to your cache repo during training
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: you use a capital letter after colons

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Fixed it when relevant

Copy link
Collaborator

@tengomucho tengomucho left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@michaelbenayoun michaelbenayoun merged commit 53f1dd8 into main Jul 22, 2025
6 checks passed
@michaelbenayoun michaelbenayoun deleted the neuron_model_cache_update branch July 22, 2025 12:40
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants