intel/auto-round

Pull Requests Commits

reduce the memory usage

Zhenzhong1 committed 2 days ago

7d1e25bc

optimize decode

Zhenzhong1 committed 2 days ago

229c4040

[high risk]Speedup 2.5-6X moe tuning and pin ngram to cpu (#2303)

wenhuach21 committed 3 days ago

Verified 678f8f5f

fix: update artifact name format for ark wheel download (#2321)

chensuyue committed 3 days ago

Verified b1738032

Implement AI-Powered CI Failure Analysis and Fix Recommendations (#2277)

chensuyue committed 3 days ago

Verified 18486c39

reduce cuda ut runing time (#2266)

n1ck-guo committed 3 days ago

Verified bdb0d31f

support model_type == "inkling_mm_model" (#2214)

xin3he committed 4 days ago

Verified de801f9f

fix: defer packing for multi-transformer diffusion (#2314)

changwangss committed 4 days ago

Verified b4dc0cc7

fix: update quantization configuration handling in FakeFormat and related tests (#2311)

xin3he committed 4 days ago

Verified 777c4ea7

Optimize diffusion calibration step sampling (#2265)

changwangss committed 5 days ago

Verified ce90add6

add sdxl svdquant (#2251)

changwangss committed 5 days ago

Verified 1c3a5518

fix: handle mantissa round overflow (#2302)

xin3he committed 5 days ago

Verified 9478fcf8

support 5,6,7 bits on cuda and add humming backend (#2281)

wenhuach21 committed 5 days ago

Verified 1f23576a

Fix xpu CI retry issue (#2309)

chensuyue committed 5 days ago

Verified fee969dd

Update CI/CD dockerfile (#2294)

XuehaoSun committed 5 days ago

Verified 726becc3

[risky]reduce ram usage of loading moe without monkey patch and fix qwen3-flash-next group_size issue (#2291)

wenhuach21 committed 7 days ago

Verified 7d8905ef

Optimize INT4 MoE prefill/decode (sym) XPU kernels with dedicated w4a16 tiling (#2111)

Copilot committed 7 days ago

Verified 21bab556

update awq detail data of fake eval (#2285)

WeiweiZhang1 committed 8 days ago

Verified 3942b57e

Allow non-MLLM calibration datasets when not quantizing vision (#2065)

aquilarubra committed 8 days ago

Verified 63dbc8ba

fix gguf format mtp bug (#2297)

n1ck-guo committed 8 days ago

Verified 947ce102

fix reproducible seeds

jijiaz committed 8 days ago

55e4f787

modified repack fp16-bf16 route tests

jijiaz committed 8 days ago

be91de74

modified repack fp16-bf16 route tests

jijiaz committed 8 days ago

0474fab9

fix CUDA UT (#2284)

xin3he committed 8 days ago

Verified 039f027f

feat(orchestrator): add reference_output to quantize_block to skip collect_reference (#2289)

xesdiny committed 8 days ago

Verified c9017ff5

Enhance test logging and retry mechanism for unit tests (#2274)

chensuyue committed 9 days ago

Verified b48c9db6

fix: skip activation fake-quant in FakeActQuantLinear when act quantization is not configured (#2286)

n1ck-guo committed 9 days ago

Verified 888b0f1e

fix: replace bare except clauses with except Exception (#1981)

ramkrishs committed 9 days ago

Verified 81b4db35

fix: asymmetric zero-point packing, format-scoped 8-bit asym, W5/W6/W7 llm_compressor presets (#2273)

avtc committed 9 days ago

Verified 5d2f8847

fix(diffusion): use real images for I2V calibration (#2272)

changwangss committed 10 days ago

Verified 1df1cef5

Older