Do not cast integer to float unintentionally #547
Merged
+21
−15
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Outdated suggestions cannot be applied.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.
Resolves #544
Also resolves #545 (we just don't wrap those layers with metadata)
Also resolves #464
The issue in 544 was that we were casting embedding layer integer inputs to f16, and then casting back to s64 which causes all sorts of issues because of loss of precision. Now we just never cast integer types at all. We get the same outputs now with f16: