Chinmay ShrivastavaandJonSnow1807 9fbe30c2a0 CUDA: keep the MUL_MAT_ID precision in the per-expert fallback (#31345)
* tests: add CUDA-only MUL_MAT/MUL_MAT_ID cases with results above the fp16 range

IQ1_M weights have no MMQ kernel, so once the batch is bigger than the MMVQ
limit the path is dequantize + cuBLAS. These cases produce results around
1.3e5, past the fp16 max of 65504, so an fp16 result buffer turns them into inf
while the CPU reference stays finite. They are only added for the CUDA backend.
Vulkan deliberately accumulates quantized matmuls in fp16 at default precision
and would fail them by design.

* CUDA: use an fp32 result for quantized src0 in the cuBLAS fallback

On the F16 compute path, quantized weights went through cuBLAS with fp16
accumulation and an fp16 result buffer that was converted to fp32 afterwards.
Anything above 65504 became inf and the next layer turned into NaN (see #30951:
gpt-oss-120b with IQ1_M experts at -ub 512). Set prefer_f32_output for
quantized src0 so the GEMM still uses fp16 operands but accumulates and stores
in fp32, the same mode already used on Volta, RDNA4 and CDNA. Every other
quantized path already accumulates in fp32.

Diagnosis and fix by denizyuret in #30951.

* CUDA: keep the MUL_MAT_ID precision in the per-expert fallback

The per-expert dst_slice kept op = GGML_OP_NONE, so ggml_prec_get_acc returned UNDEFINED and a requested precision was ignored. Set the op to GGML_OP_MUL_MAT. Also revert the fp32 result change and remove the tests.

---------

Co-authored-by: JonSnow1807 <JonSnow1807@users.noreply.github.com>
2026-11-20 09:42:16 +02:00
2026-11-19 02:39:05 +01:00
2026-11-19 02:39:05 +01:00
2026-11-18 00:04:15 +01:00
2026-06-12 15:53:26 +02:00
2026-10-17 02:11:56 +02:00
2026-11-16 13:28:31 +02:00
2026-11-17 11:04:12 +02:00
2026-02-02 08:38:55 +02:00
2026-11-19 13:20:47 +02:00

llama.cpp

llama.cpp

Description

The main goal of llama.cpp is to enable LLM (and MLLM) inference with minimal setup and state-of-the-art performance on a wide range of hardware - locally and in the cloud.

The llama.cpp project is built on top of the ggml library.

Quick start

A few options to get llama.cpp installed on your machine:

# curl
curl -LsSf https://llama.app/install.sh | sh

# powershell
irm https://llama.app/install.ps1 | iex

Once installed:

# Download and run a model directly from Hugging Face
llama cli -hf ggml-org/Qwen3.5-0.8B-GGUF

# Launch OpenAI-compatible API server
llama serve -hf ggml-org/Qwen3.5-0.8B-GGUF
VLM session with `llama cli` VLM session with llama cli Llama Chat, the built-in chat app, against `llama serve` running Qwen 3.6 Llama Chat, the built-in chat app, against llama serve

Documentation

Tools

Development

Contributing

  • Contributors can open PRs
  • Collaborators will be invited based on contributions
  • Maintainers can push to branches in the llama.cpp repo and merge PRs into the master branch
  • Any help with managing issues, PRs and projects is very appreciated!
  • Read the CONTRIBUTING.md for more information

Acknowledgements

  • yhirose/cpp-httplib - Single-header HTTP server, used by llama-server - MIT license
  • nothings/stb - Single-header image format decoder, used by multimodal subsystem - Public domain
  • nlohmann/json - Single-header JSON library, used by various tools/examples - MIT License
  • mackron/miniaudio - Single-header audio format decoder, used by multimodal subsystem - Public domain
  • sheredom/subprocess.h - Single-header process launching solution for C and C++ - Public domain
S
Description
No description provided
Readme MIT
580 MiB
0 Stars 1 Watchers 0 Forks
Languages
C++ 54.9%
C 16.7%
Python 7%
Cuda 5.6%
TypeScript 3.9%
Other 11.7%