[Quantization] AutoGPTQ refactor and matmul combination support #694
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Outdated suggestions cannot be applied.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.
This PR refactors the AutoGPTQ integration to better align with the framework design. The PR, meanwhile, supports the AutoGPTQ quantization in MLC LLM with matmul combination.
With this PR, you will be able to compile Llama2 using the following command:
to use the AutoGPTQ quantization. Note that the first run may take around 10 min for AutoGPTQ quantization computation, and the following runs will be much quicker. The AutoGPTQ quantization requires the Python
auto_gptq
package to have version at least 0.2.0.Co-authored-by: Ruihang Lai ruihangl@cs.cmu.edu