Issues

• DoRA and BoRA support

  ggml-org/ggml#1648 · Filter by author uis246 opened on Oct 3, 2026

• Why doesn't ggml distrbute libggml-cpu-{arch}.so got all CPU architectures?

  ggml-org/ggml#1630 · Filter by author yurivict opened on Sep 16, 2026
  1

• [Feature request] Use a stable GGML CPU thread pool instead of re-creating it for each request

  ggml-org/ggml#1628 · Filter by author Digipom opened on Sep 14, 2026

• ggml-cuda: missing <cuda/iterator>/<cuda/execution>/<cuda/stream_ref> includes break build with CCCL 3.1+ (argsort.cu, top-k.cu)

  ggml-org/ggml#1624 · Filter by author masoft2016 opened on Sep 11, 2026

• MUSA: flash-attn TILE GQA path (ncols2>1) emits all-NaN logits when attention window length is a multiple of 256

  ggml-org/ggml#1621 · Filter by author shengjuntu opened on Sep 7, 2026

• vulkan: vkGetDeviceQueue2 returns VK_NULL_HANDLE on older AMD drivers, first vkQueueSubmit aborts

  ggml-org/ggml#1615 · Filter by author thewh1teagle opened on Sep 4, 2026
  1

• Convert ggml to Rust

  ggml-org/ggml#1609 · Filter by author thewh1teagle opened on Aug 27, 2026

• cpu: support GGML_CPU_ALL_VARIANTS with static linking

  ggml-org/ggml#1608 · Filter by author thewh1teagle opened on Aug 27, 2026

• libggml isn't throwing an error if intel gpu fence timeout happens but DOES if GGML_VK_PERF_LOGGER=1

  ggml-org/ggml#1565 · Filter by author erikarn opened on Jul 15, 2026

• CUDA backend crashes with "an illegal memory access" on SM 8.9 (RTX 4060 Laptop) when processing ViT-style small matmul shapes

  ggml-org/ggml#1557 · Filter by author Raphael620 opened on Jul 9, 2026

• Question about requirements.txt dependencies

  ggml-org/ggml#1555 · Filter by author eshwanthkartitr opened on Jul 6, 2026

• libggml 0.13.1 fails in MUL_MAT_ID() on i915 (Intel comet lake) on freebsd-16

  ggml-org/ggml#1554 · Filter by author erikarn opened on Jul 6, 2026
  1

• ggml-metal: a transient command-buffer failure (e.g. iOS background-GPU rejection) permanently latches the whole backend into an error state

  ggml-org/ggml#1545 · Filter by author sindresorhus opened on Jun 17, 2026
  1

• Meta: gguf-org impersonation

  ggml-org/ggml#1542 · Filter by author Green-Sky opened on Jun 14, 2026

• support prebuild

  ggml-org/ggml#1537 · Filter by author Cyberhan123 opened on Jun 11, 2026

• ggml-0.14.0 build fails on OpenBSD/riscv64

  ggml-org/ggml#1535 · Filter by author catap opened on Jun 10, 2026

• ggml-metal: im2col _ext kernel dispatched with base-kernel grid geometry corrupts large 1-D convs

  ggml-org/ggml#1532 · Filter by author shakfu opened on Jun 8, 2026
  1

• error: no member named 'nr' in 'ggml_backend_meta_split_state'

  ggml-org/ggml#1530 · Filter by author yurivict opened on Jun 7, 2026
  1

• GGML_OP_CONV_TRANSPOSE_1D is Way Too Slow at TTS Scale (CUDA/HIP/Metal)

  ggml-org/ggml#1529 · Filter by author Rafa00127 opened on Jun 7, 2026

• Recommended way to add graph-internal distributed collectives for downstream ggml projects?

  ggml-org/ggml#1525 · Filter by author yiming-l21 opened on May 31, 2026

• Three backward-pass gaps blocking standard ViT training: LayerNorm, GELU, conv_2d/CONT

  ggml-org/ggml#1514 · Filter by author OriPekelman opened on May 27, 2026
  1

• SYCL: Provided range and/or offset does not fit in int

  ggml-org/ggml#1511 · Filter by author MrDrMcCoy opened on May 26, 2026

• Segfault when loading arbitrary file as mmproj or model

  ggml-org/ggml#1505 · Filter by author joeyhub opened on May 23, 2026

• ggml-cpu sched aliases intermediate grad-chain tensors in long backward graphs (CUDA correct)

  ggml-org/ggml#1501 · Filter by author OriPekelman opened on May 21, 2026
  1

• ggml-cuda: missing __syncthreads() in ssm-scan.cu

  ggml-org/ggml#1498 · Filter by author FrankZhuQS opened on May 19, 2026
  1