graph : add optional scale parameter to build_lora_mm#20427
Merged
CISC merged 1 commit intoggml-org:masterfrom Mar 11, 2026
Merged
graph : add optional scale parameter to build_lora_mm#20427CISC merged 1 commit intoggml-org:masterfrom
CISC merged 1 commit intoggml-org:masterfrom
Conversation
ggerganov
approved these changes
Mar 11, 2026
Comment on lines
768
to
+771
| ggml_tensor * build_lora_mm( | ||
| ggml_tensor * w, | ||
| ggml_tensor * cur) const; | ||
| ggml_tensor * cur, | ||
| ggml_tensor * w_s = nullptr) const; |
Member
There was a problem hiding this comment.
For a follow-up PR, we should move the cur at the front, to increase consistency of the interfaces:
ggml_tensor * build_lora_mm(
ggml_tensor * cur,
ggml_tensor * w,
ggml_tensor * w_s = nullptr) const;It will touch a lot of lines though.
CISC
approved these changes
Mar 11, 2026
Member
CISC
left a comment
There was a problem hiding this comment.
Concur with @ggerganov on ordering.
ProgenyAlpha
pushed a commit
to ProgenyAlpha/llama.cpp
that referenced
this pull request
Mar 12, 2026
tekintian
added a commit
to tekintian/llama.cpp
that referenced
this pull request
Mar 12, 2026
* 'master' of github.com:ggml-org/llama.cpp: (33 commits) convert : better mtp check and fix return [no ci] (ggml-org#20419) vulkan: fix SSM_CONV PP scaling with large ubatch sizes (ggml-org#20379) New conversations now auto-select the first loaded model (ggml-org#20403) ggml-virtgpu: Fix some build commands (ggml-org#20341) metal : avoid divisions in bin kernel (ggml-org#20426) ci: Setup self-hosted CI for Intel Linux Vulkan backend (ggml-org#20154) vulkan: fix l2_norm epsilon handling (ggml-org#20350) vulkan: fix OOB check in flash_attn_mask_opt (ggml-org#20296) vulkan: Fix ErrorOutOfHostMemory on Intel GPU when loading large models with --no-mmap (ggml-org#20059) opencl: use larger workgroup size for get_rows (ggml-org#20316) opencl: add cumsum op (ggml-org#18981) hip: compile debug builds with -O2 on hip to avoid a compiler bug (ggml-org#20392) common/parser: add GigaChatV3/3.1 models support (ggml-org#19931) model : add support for Phi4ForCausalLMV (ggml-org#20168) graph : add optional scale parameter to build_lora_mm [no ci] (ggml-org#20427) common : fix --n-cpu-moe, --cpu-moe for models with fused gate + up (ggml-org#20416) ggml-webgpu: Add supports for `GGML_OP_REPEAT` (ggml-org#20230) llama : enable chunked fused GDN path (ggml-org#20340) llama : whitespace cleanup (ggml-org#20422) ggml : add NVFP4 quantization type support (ggml-org#19769) ...
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.
@ggerganov As discused in #19769. Add optional
w_sparameter tobuild_lora_mm()for applying a multiplicative scale after the matmul. This cleans up the pattern used by bitnet and NVFP4 models where a per-tensor scale is applied after each weight multiplication.