intel/auto-round

Pull Requests Commits

Fix some CI issues (#1842)

lvliang-intel committed 86 days ago

Verified fd495491

Support Exporting Block-Wise FP8 AR Format (#1798)

Zhenzhong1 committed 89 days ago

Verified bf1c81ce

support quarot/spinquant rotation before quantization (#1797)

lkk12014402 committed 90 days ago

Verified 064cf312

Reduce VRAM usage of quantizing VLM models (#1777)

lvliang-intel committed 90 days ago

Verified 64a5817d

add new accuracy test for VLM loading and fix bug (#1829)

xin3he committed 90 days ago

Verified a577f101

Awq algorithm (#1749)

WeiweiZhang1 committed 90 days ago

Verified e8ccff29

add mimo-audio, Qwen-TTS model backbone quantization (#1810)

WeiweiZhang1 committed 90 days ago

Verified 04e14693

add inc integration unit test (#1820)

xin3he committed 90 days ago

Verified 948b482c

Reduce RAM usage of quantizing VLM models and fix some issues of quantizing gemma4 (#1791)

lvliang-intel committed 90 days ago

Verified 9696884c

Update dependencies in CI and installation scripts for cuda compatibility (#1825)

XuehaoSun committed 91 days ago

Verified bd0a1f1e

remove new arch warning (#1832)

n1ck-guo committed 91 days ago

Verified 5ce6308a

Fix few merge errors (#1817)

n1ck-guo committed 91 days ago

Verified 706ac7e3

[Fix Accuracy] fix layer name mismatch of VLM(qwen3.5-2B) in hf loading (#1823)

xin3he committed 91 days ago

Verified 425a6fff

[step2]adjust gguf tuning algorithm (#1824)

wenhuach21 committed 91 days ago

Verified 10b9f523

add auto_round_rtn cli and remove fast (#1808)

n1ck-guo committed 92 days ago

Verified debba756

Feats: Quantize/save/evaluate the Wan-AI/WAN2.2 models in w4a16 format (#1678)

lvliang-intel committed 92 days ago

Verified f1db02a3

[bug fix] model free: fix copy folder of diffusion model, fix unresponsive disable_model_free, and cuda UT (#1809)

xin3he committed 92 days ago

Verified 2c915bbd

[step 1]adjust gguf tuning algorithm (#1649)

wenhuach21 committed 92 days ago

Verified 770b7a3a

chore: clean wheel download directory and update path exclusions in test configurations (#1822)

chensuyue committed 92 days ago

Verified e46231bb

Support ByteDance-Seed/BAGEL-7B-MoT quantization in w4a16 format (#1633)

lvliang-intel committed 92 days ago

Verified 91679a5c

Scope ARK oneDNN XPU build, bump ARK version to 0.13.1, and remove the dpcpp runtime dependency (#1819)

Copilot committed 95 days ago

Verified a9f8fc1a

add smooth K for sagev1 (#1806)

luoyu-intel committed 95 days ago

Verified 017ed92c

chore: add shared agent config layout (#1700)

yiliu30 committed 96 days ago

Verified 8ccf34b0

[ARK CI] Enhance auto-round-lib CI test (#1801)

chensuyue committed 96 days ago

Verified 5bbe39ba

[refactor] decouple calibaration code (#1765)

n1ck-guo committed 96 days ago

Verified 71fb3506

reduce XPU memory usage in CI (#1812)

xin3he committed 96 days ago

Verified bc545633

set default DYNAMO_CACHE_SIZE_LIMIT to 16 (#1786)

wenhuach21 committed 97 days ago

Verified cb9b0b76

Add Dockerfile with torch installed, enhance image build logic to check for changes (#1811)

chensuyue committed 97 days ago

Verified 694ad39a

Fix mixed-precision accuracy regression when AutoScheme runs with CPU offloading and Hadamard rotation enabled (#1753)

lvliang-intel committed 97 days ago

Verified 65ae39e7

Fix GGUF-K RTN routing and fake eval dtype normalization (#1807)

n1ck-guo committed 97 days ago

Verified 351ba32a

Newer Older