cuda : fix rope with partial rotation and non-cont src #14580
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
ggml-ci
|
@qnixsynapse Could you push a SYCL fix directly in this branch, as I don't have an environment to test with? |
|
|
ggml-ci
JohannesGaessler
approved these changes
Jul 8, 2025
ggml/src/ggml-cuda/rope.cu
Outdated
Comment on lines
62
to
63
| dst[i + 0] = x[ix + 0]; | ||
| dst[i + 1] = x[ix + 1]; |
There was a problem hiding this comment.
Suggested change
| dst[i + 0] = x[ix + 0]; | |
| dst[i + 1] = x[ix + 1]; | |
| dst[idst + 0] = x[ix + 0]; | |
| dst[idst + 1] = x[ix + 1]; |
From what I can tell i and idst are the same.
Oh you already seems to fix it... Thank you!! |
|
@jeffbolznv @0cc4m Looks like Vulkan needs a fix too: |
gabe-l-hart
added a commit
to gabe-l-hart/llama.cpp
that referenced
this pull request
Jul 8, 2025
* origin/master: model : fix hunyuan moe chat template (ggml-org#14584) model : add SmolLM3 (ggml-org#14581) memory : fix broken batch splits for recurrent cache (ggml-org#14575) vulkan : fix rope with partial rotation and non-cont src (ggml-org#14582) server: Add ability to mount server at prefix (ggml-org#14544) model : add hunyuan moe (ggml-org#14425) vulkan: increase timeout for CI (ggml-org#14574) cuda : fix rope with partial rotation and non-cont src (ggml-org#14580) CUDA: add bilinear interpolation for upscale (ggml-org#14563) musa: fix build warnings (unused variable) (ggml-org#14561) llama : fix incorrect minicpm3 v_states shape (ggml-org#14571) llama : remove ggml_cont where possible (ggml-org#14568)
qnixsynapse
pushed a commit
to menloresearch/llama.cpp
that referenced
this pull request
Jul 10, 2025
* cuda : fix rope non-cont ggml-ci * cont : fix multi-rope + add test ggml-ci * sycl : try fix ggml-ci * cont : fix sycl + clean-up cuda ggml-ci
qnixsynapse
pushed a commit
to menloresearch/llama.cpp
that referenced
this pull request
Jul 10, 2025
* cuda : fix rope non-cont ggml-ci * cont : fix multi-rope + add test ggml-ci * sycl : try fix ggml-ci * cont : fix sycl + clean-up cuda ggml-ci
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.
The problem was revealed by #14573 and #14568
The problem occurs for ropes with
n_dim < ne00+ non-contsrc.We never really had proper tests for this case because of the
all = false;logic in thetest-backend-ops.