cuda : fix defrag with quantized KV #9319
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
github-actions
bot
added
Nvidia GPU
slaren
force-pushed
the
sl/fix-cuda-defrag
branch
from
290a6e5 to
e462919
Compare
ggerganov
approved these changes
Sep 5, 2024
Nexesenex
added a commit
to Nexesenex/croco.cpp
that referenced
this pull request
Sep 5, 2024
cuda : fix defrag with quantized KV (ggml-org#9319)
MaggotHATE
added a commit
to MaggotHATE/Llama_chat
that referenced
this pull request
Sep 11, 2024
* Important: this guards assert in ggml-backend.c introduced in ggml-org/llama.cpp#9319 , be aware * Merged recent Seed commit * Added a small .txt guide on code that needs to be added to make clblast work on current llama.cpp * minor display styling
dsx1986
pushed a commit
to dsx1986/llama.cpp
that referenced
this pull request
Oct 29, 2024
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 15, 2024
arthw
pushed a commit
to arthw/llama.cpp
that referenced
this pull request
Nov 18, 2024
Nexesenex
pushed a commit
to Nexesenex/croco.cpp
that referenced
this pull request
Feb 25, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
There were several issues with KV defragmentation with quantized KV:
ggml_cpyfrom quant to quant, which was not supported in the CUDA backendggml_backend_schedcannot fallback to the CPU backend either when the destination is pre-allocated, which was not correctly detectedThis fixes the issues in
ggml_backend_schedand adds support to the CUDA backend forggml_cpywhen the types are the same and the tensors are contiguous (usingcudaMemcpyAsync).Other backends may also be affected.
Fixes #9314