Hacktoberfest 2026: the issues maintainers tagged for October, open and beginner-friendly. Browse Hacktoberfest issues

PTQ-Level Efficiency, QAT-Level Accuracy|训练后量化的效率,量化感知训练的精度

Is intel/auto-round beginner-friendly?

In the last 30 days, intel/auto-round merged 40 of 61 pull requests from outside contributors, and its maintainers usually reply within 1 day. 1 beginner-friendly issue is open now.

Stars
1.7k
Forks
193
Open beginner issues
1
Indexed issues
131
Avg merge
2d 14h
Merged PRs (30d)
55
Dominant language
Python
License
Apache-2.0
Last GitHub push
Oct 21, 2026
Latest indexed
Sep 17, 2026
Contributing guide
Contributing guide
Code of conduct
Code of conduct
Beginner labels
good first issue

How to contribute to intel/auto-round

  1. Read the contributing guide first: it says how the maintainers want changes proposed, tested and reviewed.
  2. Read its code of conduct: it covers issues and pull requests as well as chat.
  3. Your contributions will be published under the project's Apache-2.0 license.
  4. Pick the open beginner-friendly issue below and comment that you want to work on it before you start.

Issues somebody may already be working on are listed last. List everything by date

  • [Feature]: support NVFP4_E5M3_GS
    Open
    enhancement

    Difficulty 4/5 3-5 days Newbie friendliness 44/100

    intel/auto-round#2532 ·

    Maintainers usually reply within 1 day

  • chat_template: vLLM 0.31 renames reasoning_content to reasoning (Qwen templates lose previous reasoning)
    Open

    Difficulty 2/5 1-3 hours Newbie friendliness 76/100

    intel/auto-round#2525 ·

    Maintainers usually reply within 1 day

  • [Feature]: move calib dataset ,data utils to calibration folder
    Open
    enhancement

    Difficulty 3/5 1-2 days Newbie friendliness 20/100

    intel/auto-round#2507 ·

    Maintainers usually reply within 1 day

  • [Feature]: eval NeUQI with typical tasks and provide usage suggestion to users
    Open
    enhancement

    Difficulty 4/5 3-5 days Newbie friendliness 22/100

    intel/auto-round#2502 ·

    Maintainers usually reply within 1 day

  • [Bug]: Inference speed becomes slower after AutoRound quantization
    Open
    bug

    Difficulty 5/5 Over a week Newbie friendliness 15/100

    intel/auto-round#2482 · 9 comments ·

    Maintainers usually reply within 1 day

  • [Feature]: Improve CLI user experience by migrating from argparse to Click or Typer
    Open
    enhancement

    Difficulty 5/5 Over a week Newbie friendliness 25/100

    intel/auto-round#2475 ·

    Maintainers usually reply within 1 day

  • [Feature]: move diffusion code from signround quantizer
    May be free again @changwangss claimed this 35 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2399 · 1 comment · 1 assignee ·

    Maintainers usually reply within 1 day

  • RFC: Data-parallel block tuning (--parallel_quantization)
    May be free again @avtc claimed this 32 days ago, and no pull request is open. Open
    enhancement

    Difficulty 5/5 Over a week Newbie friendliness 25/100

    intel/auto-round#2378 · 2 comments · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Bug]: MiniMax-M2.7_mxfp4_mixed_fp8kv_LLMC vllm eval failed
    May be free again @xin3he claimed this 39 days ago, and no pull request is open. Open
    bug

    intel/auto-round#2369 · 1 assignee ·

    Maintainers usually reply within 1 day

  • Add a CI check to prevent coverage regressions in pull requests
    May be free again @chensuyue claimed this 44 days ago, and no pull request is open. Open

    intel/auto-round#2347 · 3 comments · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: expand AWQ experiments to low-bit INT and Hadamard combinations
    May be free again @WeiweiZhang1 claimed this 35 days ago, and no pull request is open. Open
    enhancement high priority

    Difficulty 5/5 Over a week Newbie friendliness 35/100

    intel/auto-round#2298 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Bug]: unexpected enable_quant_input: false in config.json for FP8_BLOCK
    May be free again @n1ck-guo claimed this 56 days ago, and no pull request is open. Open
    bug

    intel/auto-round#2224 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: svdquant supports llm and other schemes(W2A16)
    May be free again @changwangss claimed this 56 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2218 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Bug]: qwen3.8-27B q2ks is larger than unsloth's Q2KL
    Open
    enhancement low priority

    Difficulty 4/5 3-5 days Newbie friendliness 30/100

    intel/auto-round#2217 · 1 comment ·

    Maintainers usually reply within 1 day

  • [Bug]: ARK SageV1 BF16 D128 causes DEVICE_LOST on Windows Arc B580
    May be free again @Zhenzhong1 claimed this 53 days ago, and no pull request is open. Open
    bug

    intel/auto-round#2213 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: mxfp4 dual-scale investigate
    May be free again @changwangss claimed this 67 days ago, and no pull request is open. Open

    intel/auto-round#2199 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: auto-scheme shared_layers supports vllm pattern as default
    May be free again @lvliang-intel claimed this 56 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2183 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Bug]: dump auto-round key hype-parameters to conifg.json
    May be free again @n1ck-guo claimed this 56 days ago, and no pull request is open. Open
    bug

    intel/auto-round#2154 · 1 assignee ·

    Maintainers usually reply within 1 day

  • Qwen3-30B-A3B-Instruct-2507 bf16 vllm eval failed on H200
    May be free again @yiliu30 claimed this 75 days ago, and no pull request is open. Open

    intel/auto-round#2151 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: refine architecture code
    May be free again @n1ck-guo claimed this 56 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2126 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: hadamard vllm inference support
    May be free again @lkk12014402 claimed this 83 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2122 · 2 comments · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: each calibration dataset set its own default apply_chat_template value
    May be free again @lkk12014402 claimed this 59 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2099 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: refine auto-round -h
    May be free again @n1ck-guo claimed this 59 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2078 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: Support FP/INT W4A4 recipe
    May be free again @WeiweiZhang1 claimed this 111 days ago, and no pull request is open. Open

    intel/auto-round#2009 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: support and eval VSRAQ
    May be free again @WeiweiZhang1 claimed this 56 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#2008 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Bug]: test/test_mlx/test_mlx_format.py failed on Mac OS
    Open
    bug

    Difficulty 3/5 1-2 days Newbie friendliness 35/100

    intel/auto-round#1955 ·

    Maintainers usually reply within 1 day

  • [Feature]: Introduce built-in ar, ar-light, and ar-best recipes for each quantization scheme
    May be free again @n1ck-guo claimed this 111 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#1940 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: AutoScheme is currently underperforming compared to Unsloth's model for gguf
    Open
    enhancement

    intel/auto-round#1896 · 17 comments · 2 assignees ·

    Maintainers usually reply within 1 day

  • [Feature]: support teq algorithm
    May be free again @WeiweiZhang1 claimed this 144 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#1888 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: refine export code to provide unified interface for pack_layer and save_quantized_as_format
    May be free again @n1ck-guo claimed this 111 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#1789 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: We need a memory constrained converting.
    May be free again @thuang6 claimed this 192 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#1702 · 1 comment · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: the ram usage of AutoScheme needs further optimizaiton
    May be free again @lvliang-intel claimed this 173 days ago, and no pull request is open. Open
    enhancement

    intel/auto-round#1697 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Upstream]: SGlang inference with AR kernels
    May be free again @luoyu-intel claimed this 219 days ago, and no pull request is open. Open

    intel/auto-round#1587 · 2 assignees ·

    Maintainers usually reply within 1 day

  • Ensuring correct device dispatch after module replacement
    May be free again @yiliu30 claimed this 274 days ago, and no pull request is open. Open
    good first issue

    intel/auto-round#1308 · 3 comments · 1 assignee ·

    Maintainers usually reply within 1 day

  • Enhance AutoScheme
    May be free again @xin3he claimed this 240 days ago, and no pull request is open. Open

    intel/auto-round#910 · 1 comment · 1 assignee ·

    Maintainers usually reply within 1 day

  • Use triton-windows on Windows for torch compile and triton backend
    Open
    enhancement

    Difficulty 2/5 1-3 hours Newbie friendliness 48/100

    intel/auto-round#870 · 47 comments ·

    Maintainers usually reply within 1 day

  • [Bug]: FP8 static KV cache accuracy issue
    Possibly taken @xin3he claimed this 6 days ago. Open
    bug

    intel/auto-round#2518 · 1 reaction · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: split special_model_handler to a folder
    Possibly taken @Cross2pro claimed this 3 days ago. Open
    enhancement

    Difficulty 3/5 Half a day Newbie friendliness 42/100

    intel/auto-round#2514 · 2 comments ·

    Maintainers usually reply within 1 day

  • [Bug]: kv cache potential bugs
    Possibly taken @xin3he claimed this 6 days ago. Open
    bug high priority

    Difficulty 3/5 1-2 days Newbie friendliness 45/100

    intel/auto-round#2504 · 2 comments · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Bug]: opt-rtn causes excessive VRAM usage for Nemotron-3.5-Lightning-30B-A3B-BF16
    Possibly taken @n1ck-guo claimed this 6 days ago. Open
    bug

    Difficulty 4/5 3-5 days Newbie friendliness 22/100

    intel/auto-round#2501 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [SGLang-Omni] AR MXFP4/8 model inference​
    Possibly taken @WeiweiZhang1 claimed this 13 days ago. Open

    intel/auto-round#2480 · 1 comment · 1 assignee ·

    Maintainers usually reply within 1 day

  • [SGL] MXFP4/8 scheme support​
    Possibly taken @WeiweiZhang1 claimed this 13 days ago. Open

    intel/auto-round#2479 · 1 comment · 1 assignee ·

    Maintainers usually reply within 1 day

  • [DeepKlox] SA v2/v3, BSA support​
    Possibly taken @yiliu30 claimed this 13 days ago. Open

    intel/auto-round#2478 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [LLMC] Triton kernel in LLM-C XPU support
    Possibly taken @yiliu30 claimed this 13 days ago. Open
    llmc-only

    intel/auto-round#2477 · 1 comment · 1 assignee ·

    Maintainers usually reply within 1 day

  • [LLMC] NVFP4+ scheme support
    Possibly taken @yiliu30 claimed this 13 days ago. Open
    llmc-only

    intel/auto-round#2476 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [Feature]: Save important AutoRound hyperparameters in the llmc format, just like in the AutoRound format.
    Possibly taken @Cross2pro claimed this 4 days ago. Open
    enhancement

    Difficulty 3/5 1-2 days Newbie friendliness 45/100

    intel/auto-round#2469 · 2 comments ·

    Maintainers usually reply within 1 day

  • [Bug]: offloder cleaning failed
    Possibly taken @lvliang-intel claimed this 16 days ago. Open
    bug

    intel/auto-round#2467 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [ARK] Per-head hybrid attention
    Possibly taken @luoyu-intel claimed this 16 days ago. Open

    intel/auto-round#2465 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [ARK] Fused HMT Quant
    Possibly taken @luoyu-intel claimed this 16 days ago. Open

    intel/auto-round#2464 · 1 assignee ·

    Maintainers usually reply within 1 day

  • [ARK] SageAttention V3 (Mixed 4/8)
    Possibly taken @luoyu-intel claimed this 16 days ago. Open

    intel/auto-round#2463 · 1 assignee ·

    Maintainers usually reply within 1 day

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.