Feat (llm/awq): activation-aware weight scaling #1213
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
pablomlago
changed the title
Giuseppe5
reviewed
Mar 10, 2025
| @@ -251,6 +250,48 @@ def apply(self, model, is_training, quantization_enabled): | |||
| self.enable_param_quantization(model, is_training) | |||
|
|
|||
|
|
|||
| class disable_enable_quantization: | |||
There was a problem hiding this comment.
Potentially interesting for all use cases where we need to do this. I'd expose flags to disable weight/act/bias quantization
Giuseppe5
reviewed
Mar 10, 2025
src/brevitas/graph/equalize.py
Outdated
| @@ -780,9 +781,11 @@ def _no_equalize(): | |||
| for module in chain(src_axes.values(), sink_axes.values()): | |||
| rewriters.extend(module.instantiate_rewriters(rewriter_class, scaling_factors)) | |||
|
|
|||
| # Apply rewriters before offloading | |||
| # Apply rewriters before offloading, if parametrize_inplace is True. Note that parametrizations | |||
| # are not immediately to prevent potential errors if the model is offloaded. | |||
There was a problem hiding this comment.
Can you elaborate a bit more the issue here?
Giuseppe5
reviewed
Mar 10, 2025
| raise ValueError # early exit to break later inference | ||
|
|
||
| # patch layer 0 to catch input and kwargs | ||
| layers[0] = Catcher(layers[0]) | ||
| blocks[0] = Catcher(blocks[0]) |
There was a problem hiding this comment.
I don't think we need this part of the codebase, why can't we do what we do in GPTQ to catch the input to the first block?
We can also move that piece of code to some utils in exmples/common/generative
Giuseppe5
reviewed
Mar 10, 2025
src/brevitas/utils/python_utils.py
Outdated
| @@ -64,3 +65,30 @@ def run(*args, **kwargs): | |||
| return function(*args, **kwargs) | |||
|
|
|||
| return run | |||
|
|
|||
|
|
|||
| def longest_common_prefix(strings: List[str]): | |||
There was a problem hiding this comment.
This seems overly specific to AWQ, not sure if this should live here
Giuseppe5
reviewed
Mar 10, 2025
| "ffn.act": block.ffn.act, | ||
| "ffn.down_proj": block.ffn.down_proj,}, | ||
| )) | ||
| elif "falcon" in str(block.__class__).lower(): |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
Reason for this PR
Implementation of AWQ: Activation-aware Weight Quantization for LLM Compression and Acceleration.
Using weight-only quantization and the configuration:
Changes Made in this PR
RegionAWQ, inheriting fromRegionto aggregate the information of the modules s on which AWQ optimizes the scale.auto_scaleandauto_clipto rely on Brevitas quantizers.Testing Summary
Testing
apply_awqagainst the author's repository.Risk Highlight
Checklist
devbranch.