ggml : quantization refactoring #3833
Merged
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
|
Building without GGML_USE_K_QUANTS is currently the only way to quantize a pure Q4_0 model, which has been helpful while developing the Nomic Vulkan backend, as we run the whole model on the GPU but did not have a Q6_K matmul shader at first. Maybe we should add a flag to quantize to disable the k-quants logic? |
Yes. Do you want to add it to this PR? |
Done. |
cebtenzzre
force-pushed
the
ggml-quants
branch
from
2ed9ad3 to
8a86b95
Compare
Nexesenex
pushed a commit
to Nexesenex/croco.cpp
that referenced
this pull request
Oct 30, 2023
* ggml : factor all quantization code in ggml-quants ggml-ci * ggml-quants : fix Zig and Swift builds + quantize tool ggml-ci * quantize : --pure option for disabling k-quant mixtures --------- Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
This was referenced Oct 30, 2023
olexiyb
pushed a commit
to Sanctum-AI/llama.cpp
that referenced
this pull request
Nov 23, 2023
* ggml : factor all quantization code in ggml-quants ggml-ci * ggml-quants : fix Zig and Swift builds + quantize tool ggml-ci * quantize : --pure option for disabling k-quant mixtures --------- Co-authored-by: cebtenzzre <cebtenzzre@gmail.com>
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Outdated suggestions cannot be applied.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.
Moving all CPU quantization code into separate source files.
This is a refactoring change - there should be no functional difference.
k_quants.h/.c->ggml-quants.h/.cggml.cintoggml-quants.cGGML_USE_K_QUANTSifdefs (i.e. always build with K-quants support)In the future,
ggml-quants.hshould probably be moved into aggml-impl.hprivate API header as discussed in ggerganov/ggml#549