b10274

ggml-org/llama.cppb10274Aug 4, 2026by github-actions[bot]same day as b10273

What changed since b10273

low risk

b10274 is a single-fix follow-up that removes duplicate empty audio chunks for short inputs in the mtmd multimodal path.

Bug fixes

  • Short audio/text inputs in the mtmd multimodal pipeline no longer produce duplicated empty audio chunks.
  • Test script changes from the fix were reverted, keeping the test harness unchanged.

Upgrade notes

This is a one-commit delta over b10273, so no config or API changes are needed; the sampler context-size removal landed in b10273, not here. Upgrade only if you use multimodal/audio input with short prompts. Pin b10274 if you want the audio-chunk fix without any other behavioral drift.

Analyzed Oct 31, 2026

Upgrading?

Release notes summary

Corrects a bug involving duplicate empty audio chunks when processing short inputs in the multi-modal text-to-speech module.

Key Highlights

  • Fixes duplicate empty audio chunks for short inputs
  • Ensures clean audio processing for short queries

Release Notes (excerpt)

<details open> mtmd: correcting duplicate empty audio chunks for short inputs (#26536) * correcting duplicate empty audio chunks for short inputs * tests.sh code restored </details> **Website:** - <https://llama.app> **macOS/iOS:** - macOS Apple Silicon (arm64) - macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED - macOS Intel (x64) - iOS XCFramework **Linux:** - Ubuntu x64 (CPU) - Ubuntu arm64 (CPU) - Ubuntu s390x (CPU) - Ubuntu x64 (Vulkan) - Ubuntu arm64 (Vulkan)…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories