Add NVFP4 MTP scale tensors#23563
Merged
Merged
Conversation
CISC
approved these changes
May 23, 2026
am17an
approved these changes
May 23, 2026
kashif
pushed a commit
to kashif/llama.cpp
that referenced
this pull request
May 23, 2026
* Add NVFP4 MTP scale tensors * Link Qwen3.5 MTP tensors * Aligned nullptr
spiritbuun
referenced
this pull request
in spiritbuun/buun-llama-cpp
May 25, 2026
* Add NVFP4 MTP scale tensors * Link Qwen3.5 MTP tensors * Aligned nullptr
fewtarius
pushed a commit
to fewtarius/CachyLLama
that referenced
this pull request
May 30, 2026
* Add NVFP4 MTP scale tensors * Link Qwen3.5 MTP tensors * Aligned nullptr
turbo-tan
pushed a commit
to turbo-tan/llama.cpp-tq3
that referenced
this pull request
Jun 2, 2026
* Add NVFP4 MTP scale tensors * Link Qwen3.5 MTP tensors * Aligned nullptr
TheTom
pushed a commit
to TheTom/llama-cpp-turboquant
that referenced
this pull request
Jun 8, 2026
* Add NVFP4 MTP scale tensors * Link Qwen3.5 MTP tensors * Aligned nullptr (cherry picked from commit b0df4c0)
zbrad
pushed a commit
to zbrad/llama.cpp
that referenced
this pull request
Jul 3, 2026
* Add NVFP4 MTP scale tensors * Link Qwen3.5 MTP tensors * Aligned nullptr
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR adds the additional MTP NVFP4 weight scale and input scale tensors for:
eh_projshared_head_headThe latest update of
remap_nvfp4_tensor_name()in #21095 will mapmtp.layers.x.eh_proj.x, if quantized as NVFP4 to the corresponding:If those scale tensors are present in the GGUF, the newly created model will fail to load without this PR. Conversely, NVFP4/MTP models will not work without this.
I posted a working NVFP4 with MTP model here: Qwen3.5-9B-NVFP4-MTP-GGUF
Yes: AI was used to help test this implementation.