Overview

7 Releases published by1 person

• v0.3.13-metal

  publishedJul 15, 2025

• v0.3.13-cu124

  publishedJul 15, 2025

• v0.3.13-cu121

  publishedJul 15, 2025

• v0.3.13-cu122

  publishedJul 15, 2025

• v0.3.13-cu123

  publishedJul 15, 2025

• v0.3.14-metal

  publishedJul 18, 2025

• v0.3.14-cu124

  publishedJul 18, 2025

1 Pull request merged by1 person

• Better chat format for Qwen2.5-VL

  #2040 mergedJul 15, 2025

1 Pull request opened by1 person

• Actually create a random seed when using seed = -1 on load

  #2042 openedJul 16, 2025

9 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Segmentation fault (core dumped) appearing randomly

  #2005 commented onJul 12, 2025 • 0 new comments

• Setting seed to -1 (random) or using default LLAMA_DEFAULT_SEED generates a deterministic reply chain

  #1809 commented onJul 12, 2025 • 0 new comments

• Can't install llama-cpp-python with HIPBLAS/ROCm on Windows

  #1489 commented onJul 15, 2025 • 0 new comments

• Generate answer from embedding vectors

  #1897 commented onJul 16, 2025 • 0 new comments

• Windows11：ERROR: ERROR: Failed to build installable wheels for some pyproject.toml based projects (llama-cpp-python)

  #2035 commented onJul 17, 2025 • 0 new comments

• FileNotFoundError: Shared library with base name 'llama' not found

  #568 commented onJul 18, 2025 • 0 new comments

• Inferencing Flan-T5 - GGML_ASSERT error

  #2038 commented onJul 18, 2025 • 0 new comments

• Add support for XTC and DRY samplers

  #1843 commented onJul 17, 2025 • 0 new comments

• ARM Runners support CUDA SBSA

  #2039 commented onJul 18, 2025 • 0 new comments