mtmd: fix whisper audio tail truncation by exposing padded buffer to FFT#22770
Merged
ngxson merged 1 commit intoMay 7, 2026
Merged
Conversation
4 tasks
Contributor
Author
|
cc @ngxson |
CISC
approved these changes
May 7, 2026
cetarthoriphros
pushed a commit
to cetarthoriphros/llama.cpp
that referenced
this pull request
May 9, 2026
meh
pushed a commit
to meh/llama.cpp
that referenced
this pull request
May 10, 2026
rsenthilkumar6
pushed a commit
to rsenthilkumar6/llama.cpp
that referenced
this pull request
May 19, 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Overview
Fixes the audio input >30s problem
Additional information
In log_mel_spectrogram, the whisper-style padding branch allocated samples_padded with a 30s silence tail but never reassigned samples and n_samples to point at it (unlike the no_padding and center_padding branches), so the pad never made it into the mel and the chunking loop dropped the final partial slice along with real audio.
Fixes #22591. Reproduced & tested with :
ggml-org/Qwen3-Omni-30B-A3B-Instruct-GGUF/
Qwen3-Omni-30B-A3B-Instruct-Q8_0.gguf
mmproj-Qwen3-Omni-30B-A3B-Instruct-Q8_0.gguf
mmproj-Qwen3-Omni-30B-A3B-Instruct-bf16.gguf
Testing
Speak for more than 30 seconds and give a random magic word at the end. The LLM must know it.
No patch
With patch
Requirements