Multi-threaded quantization #1075
Merged
Conversation
This file contains hidden or 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
Not much gain for simple quantizations, bit it will be important for quantizations that require more CPU cycles.
It now does the job in ~14 seconds on my Mac for Q4_0, Q4_1 and Q4_2. Single-threaded it was taking more than 2 minutes after adding the more elaborate version of Q4_2.
|
This could make more accurate but slow quantization methods more practical. (See #835.) |
ggerganov
approved these changes
Apr 20, 2023
After changing chunk_size to const int as suggested by @ggerganov, clang and GCC starting to warn me that I don't need to capture it in the lambda. So, I removed it from the capture list. But that makes the MSVC build fail. So, making it a constexpr to make every compiler happy.
|
Please resolve conflicts with the master branch |
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Outdated suggestions cannot be applied.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.
This PR adds multi-threading for quantization.
The gain is very minor for small models (e.g., LLaMA 7B) and simple quantization (
Q4_0andQ4_1), but very significant for large models and the now more elaborateQ4_2quantization.quantize-statsnow finishes in just 14.5 seconds (7B) or 44 seconds (13B) on my computer for all 3 quantization types. The single-threaded version took 144 seconds (7B) or 242 seconds (13B).