Issues

• CI check: detect orphaned quantization kernels vs. Python registry drift (VPTQ, AQLM, GGUF affected)

  dphnAI/sonar#1783 · Filter by author Nakanokensetsu opened on Sep 10, 2026

• [Installation]: install.sh fails to fetch aphrodite_engine-0.23.0+rocm723-cp38-abi3-linux_x86_64.whl

  dphnAI/sonar#1762 · Filter by author RobbieNeko opened on Aug 7, 2026
  8

• [Performance]: [V100 / sm_70] Volta build enablement + fused-MoE BLOCK_K performance finding

  dphnAI/sonar#1677 · Filter by author KumphanartDansiri opened on Jun 29, 2026

• [Bug]: YAML run option no longer available in 0.9.0

  bug
  dphnAI/sonar#1498 · Filter by author prolix-oc opened on Sep 6, 2025
  1

• [Installation]: Install v0.6.6/v0.6.7 on amd gpu gfx906 failed, v0.6.5 success but cannot run gptq

  dphnAI/sonar#1291 · Filter by author gengchaogit opened on Apr 5, 2025
  6

• [Bug]: Failed to install for CPU

  bug
  dphnAI/sonar#1280 · Filter by author Nero10578 opened on Mar 18, 2025
  1

• [Feature]: Consider integrating SVDquant (W4A4 quantization) from Nunchaku project

  dphnAI/sonar#1092 · Filter by author dengyingxu opened on Jan 24, 2025

• [Bug]: clamp broken with HIP

  bug
  dphnAI/sonar#1084 · Filter by author cb88 opened on Jan 8, 2025
  1

• [Bug]: Docker instance doesn't download model (affects VLLM as well)

  bug
  dphnAI/sonar#911 · Filter by author selalipop opened on Dec 17, 2024
  1

• [Feature]: add back exl2 support?

  dphnAI/sonar#884 · Filter by author joshuakoh1 opened on Dec 12, 2024
  2

• [Bug]: Error at Custom KoboldAI Endpoint! The custom endpoint failed to respond correctly. You may wish to try a different URL or API type.

  bug
  dphnAI/sonar#883 · Filter by author baditaflorin opened on Dec 12, 2024
  1

• [Feature]: pass-through parameter from request to model.forward (already implemented)

  dphnAI/sonar#836 · Filter by author qpwo opened on Nov 25, 2024
  1

• [Installation]: Cannot find CUDA_TOOLKIT_ROOT_DIR while trying to build for ROCm

  dphnAI/sonar#815 · Filter by author RuntimeRacer opened on Nov 14, 2024
  1

• [Bug]: .\gguf_to_torch.py broken along with direct load GGUF

  bug
  dphnAI/sonar#804 · Filter by author sorasoras opened on Nov 3, 2024
  2

• [Installation]: Unable to make openvino / CPU install from source work: "Failed to import from aphrodite._C with No module named 'aphrodite._C'"

  dphnAI/sonar#791 · Filter by author bolaft opened on Oct 28, 2024

• [Bug]: Impossible dependency requirement with GGUF

  bug
  dphnAI/sonar#783 · Filter by author musoles opened on Oct 19, 2024
  5

• [Installation]: I tried compile GFX1100 on WSL2 but it does not seems work

  dphnAI/sonar#780 · Filter by author sorasoras opened on Oct 16, 2024
  13

• [Feature]: Support for tpu v3-8

  dphnAI/sonar#759 · Filter by author Abdulhanan535 opened on Sep 22, 2024

• [Installation]: Failed to initialize NumPy: No module named 'numpy'

  dphnAI/sonar#734 · Filter by author Nero10578 opened on Sep 15, 2024
  2

• [Usage]: GGUFed models on AMD GPUs

  dphnAI/sonar#632 · Filter by author TuzelKO opened on Sep 5, 2024
  1

• Support 3bit quip# model.

  dphnAI/sonar#525 · Filter by author Minami-su opened on Jun 27, 2024

• [Usage]: native nvlink support or not agnostic to mobo

  dphnAI/sonar#518 · Filter by author puppetm4st3r opened on Jun 20, 2024
  1

• [Feature]: Support [RecurrentGemmaForCausalLM]

  dphnAI/sonar#506 · Filter by author hmmsmoon opened on Jun 12, 2024
  3

• [Feature]: An alternative to max_tokens which defaults to minimum(max_tokens, remaining_tokens)

  dphnAI/sonar#492 · Filter by author josephrocca opened on May 30, 2024
  1

• [Feature]: Exllamav2 Q4, Q6, and Q8 cache

  dphnAI/sonar#463 · Filter by author Anthonyg5005 opened on May 9, 2024
  3