CUDA: fix crash on large batch size for MoE models#13384
Merged
JohannesGaessler merged 1 commit intoMay 9, 2025
Conversation
github-actions
Bot
added
Nvidia GPU
Contributor
|
I'm testing now! Great work as usual! |
Contributor
|
@JohannesGaessler Looks like it works! Great work! |
CISC
approved these changes
May 9, 2025
gabe-l-hart
added a commit
to gabe-l-hart/llama.cpp
that referenced
this pull request
May 9, 2025
* origin/master: (39 commits) server : vision support via libmtmd (ggml-org#12898) sycl : implementation of reordered Q4_0 MMVQ for Intel GPUs (ggml-org#12858) metal : optimize MoE for large batches (ggml-org#13388) CUDA: FA support for Deepseek (Ampere or newer) (ggml-org#13306) llama : do not crash if there is no CPU backend (ggml-org#13395) CUDA: fix crash on large batch size for MoE models (ggml-org#13384) imatrix : Add --parse-special for enabling parsing of special tokens in imatrix calculation (ggml-org#13389) llama-run: add support for downloading models from ModelScope (ggml-org#13370) mtmd : fix batch_view for m-rope (ggml-org#13397) llama : one-off chat template fix for Mistral-Small-2503 (ggml-org#13398) rpc : add rpc_msg_set_tensor_hash_req (ggml-org#13353) vulkan: Allow up to 4096 elements for mul_mat_id row_ids (ggml-org#13326) server : (webui) rename has_multimodal --> modalities (ggml-org#13393) ci : limit write permission to only the release step + fixes (ggml-org#13392) mtmd : Expose helper_decode_image_chunk (ggml-org#13366) server : (webui) fix a very small misalignment (ggml-org#13387) server : (webui) revamp the input area, plus many small UI improvements (ggml-org#13365) convert : support rope_scaling type and rope_type (ggml-org#13349) mtmd : fix the calculation of n_tokens for smolvlm (ggml-org#13381) context : allow cache-less context for embeddings (ggml-org#13108) ...
Seunghhon
pushed a commit
to Seunghhon/llama.cpp
that referenced
this pull request
Apr 26, 2026
ljubomirj
pushed a commit
to ljubomirj/llama.cpp
that referenced
this pull request
May 6, 2026
my-other-github-account
pushed a commit
to my-other-github-account/llama.cpp
that referenced
this pull request
May 15, 2026
my-other-github-account
pushed a commit
to my-other-github-account/llama.cpp
that referenced
this pull request
May 15, 2026
phibya
pushed a commit
to ziee-ai/llama.cpp
that referenced
this pull request
May 29, 2026
fewtarius
pushed a commit
to fewtarius/CachyLLama
that referenced
this pull request
May 30, 2026
AlexiAlp
pushed a commit
to minghaop/llama.cpp
that referenced
this pull request
Jun 2, 2026
AlexiAlp
pushed a commit
to minghaop/llama.cpp
that referenced
this pull request
Jun 2, 2026
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
Fixes #13376 .
I was not able to reproduce the issue but what I think is happening is that the total number of columns being assembled for
MUL_MAT_IDis > 65535, the maximum dimension of CUDA grids in the y and z dimension. Presumably the issue can then be fixed by just swapping the x and y dimension since the limit for the x dimension is 2^31-1. @danielhanchen please confirm whether or not the fix works.