Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix record nested value de/serialization #1751
Fix record nested value de/serialization #1751
Changes from all commits
d268991
7037593
40b3c71
File filter
Filter by extension
Conversations
Jump to
There are no files selected for viewing
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Shouldn't we add more element types such as
bf16
,f16
,f64
,u64
,u32
?There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yeah we're gonna have to (at some point anyway). I started with the two types for my use cases (BF16 for Llama and F32 for all other models like the ResNet family for testing purposes). They're probably the most common too.
Wanted to get a review of the implementation before going further. As mentioned by @antimora currently this doesn't scale very well to add a concrete implementation for each type, but it's the easiest solution I came up with for now.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't see another way: we need to compiler to know the size of the vector of elements at compile time! Maybe we could read the vector as bytes instead and cast them later on?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hmmm maybe. I think for now I'll stick to the current method to wrap up this PR and in the future we could refactor this if needed. At the same time, it's not like the number of types will not be manageable.. so "not scaling" isn't necessarily an issue right now.