whisper: validate get_rows support for cpu extra buffer #3323
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
ggerganov
approved these changes
Jul 14, 2025
|
Could you share some sample numbers of the performance before and after this change? |
|
Here's benchmark Comparison (GET_ROWS vs. Baseline) for Pixel 8:
|
|
This PR has a dependency on llama.cpp PR #14676, which introduces KleidiAI support for the get_rows operator. |
bygreencn
added a commit
to bygreencn/whisper.cpp
that referenced
this pull request
Sep 24, 2025
* ggerganov/master: (89 commits) whisper: validate get_rows support for cpu extra buffer (ggml-org#3323) examples : update links in wasm examples (ggml-org#3318) sync : resolve conflicts (#0) talk-llama : sync llama.cpp sync : ggml sync : resolve conflicts (ggml/0) vulkan: support SET_ROWS (llama/14587) vulkan: optimizations for deepseek prompt processing (llama/14555) model : support LiquidAI LFM2 hybrid family (llama/14620) HIP : Add HIP 7.0+ compatibility for hipBLAS compute types (llama/14634) opencl: add tiled mul_mat_f16_f32 (llama/14535) opencl: add `set_rows` for `f16` and `f32` (llama/14547) SYCL: Initial set_rows kernel implementation (llama/14562) cuda : support Falcon-H1 state size for SSM_SCAN (llama/14602) ggml : add ggml_scale_bias (llama/14417) ggml : prevent integer overflow in gguf tensor size calculation (llama/14595) vulkan: optimize flash attention split_k_reduce (llama/14554) vulkan : fix rope with partial rotation and non-cont src (llama/14582) cuda : fix rope with partial rotation and non-cont src (llama/14580) CUDA: add bilinear interpolation for upscale (llama/14563) ...
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Outdated suggestions cannot be applied.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 patch enables KleidiAI to accelerate the Q4_0 matmul operation when it shares weight data with the get_rows operator.