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Outdated suggestions cannot be applied.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.
We had vectorized the scale and zeropoint calculations for the minmax observer, but the memoryless observer was still using the un-vectorized code. Moved the
get_qparams_along_dim
to the base observer class so all observers use it, and resolved some shape issues.With this change the token,channel and tensor strategies can all use the same logic for calling quant/dequant so this simplified the forward pass code a lot too
Testing
Running a llama1.1b model with w8a8 dynamic per token:
Before change: 10 sec/iteration
After change: 4.5 iterations/sec
A huge speedup!