AI Briefing#gguf
The latest AI and developer news about #gguf, with the original source and a short summary.
The latest AI and developer news about #gguf, with the original source and a short summary.
Unsloth has released Dynamic v3.0 quantized models for Qwen3.8-27B, demonstrating improved accuracy.
Liquid AI has released the LFM2.5 Q4_0 models, which minimize quality degradation through Quantization-Aware Distillation (QAD).
The DFlash developer released DFlash 2, which supports Qwen 3.8 27B and Muse Glimmer, and submitted a llama.cpp pull request.
Unsloth has released the massive GGUF quantized files for Kimi K3.
NVIDIA supports on-device execution by quantizing ASR, TTS, and speech codecs to GGUF.
The llama.cpp mainline now supports voice cloning with Qwen3-TTS 1.7B.
llama.cpp has been changed to load MTP tensors for certain architectures by default even when MTP is disabled, increasing VRAM usage.
audio.cpp 0.4 adds Higgs Audio v3 TTS and Fish Audio S2 Pro among others, now supporting 35 model families with Q8 GGUF improving speed and VRAM.
ik_llama.cpp has added support for the openPangu-2.0-Flash model along with a GGUF format update.
Qwythos-9B-v2, which applies FTPO technology to solve the repetitive generation problem, has been released.
Sberbank has released the GigaChat 3.5 model with 432B parameters along with a GGUF quantized version.
A Pull Request to support the DeepSeek V4 model has been merged into the llama.cpp project.
Unsloth has released **MTP GGUF weights** for the 31B, 26B-A4B, and 12B models of Gemma 4.
A Pull Request has been submitted to llama.cpp for support of Stepfun AI's Step-3.7-Flash model.
NVIDIA's Parakeet STT model has been ported to run in a C++/ggml environment without Python, greatly improving performance and efficiency.
A Qwen3.6 GGUF model has been released that integrates an MTP head, enabling self-speculative decoding without a separate draft model.
ByteShape released Qwen 3.6 35B GGUF NTP and MTP quantizations along with benchmarks.
llama.cpp added MTP support, and Qwen3.6 MTP GGUF models were released.
This analyzes the components of the GGUF file format—chat templates, special tokens, sampler settings—and the areas that still need improvement.
This explains in detail the structure of the GGUF file format and the role of metadata it includes, such as chat templates and special tokens.
Ollama has switched from a GGML-based build to direct llama.cpp support.
Unsloth has released GGUF versions of Qwen models with Multi-Token Prediction (MTP) technology applied.
The number of GGUF-format model uploads on Hugging Face has roughly doubled over the past two months.
BeeLlama.cpp, a new fork that maximizes the performance of llama.cpp, has been released.