intel/auto-round

Pull Requests Commits

yiliu30 committed 23 days ago

aea1c1c0

calib for kv/attn

yiliu30 committed 23 days ago

c0aff354

yiliu30 committed 23 days ago

493c3845

fix bf16 option in AutoScheme (#1079)

wenhuach21 committed 25 days ago

Verified 5329bcda

refine sampler (#1077)

wenhuach21 committed 27 days ago

Verified 727e573e

Support mxfp nvfp lmhead quant (#1051)

WeiweiZhang1 committed 27 days ago

Verified c4a14799

improve accuracy for 2bit with auto-round-best (#1078)

wenhuach21 committed 27 days ago

Verified 7907c329

Add MXFP8 MOE/Linear and MXFP4 Linear (#1034)

yiliu30 committed 28 days ago

Verified 61bcfdb6

simplify what's new and add publication_list (#1070)

xin3he committed 28 days ago

Verified 7345fe57

update alg_ext and add ut (#1064)

n1ck-guo committed 28 days ago

Verified e18f9b66

Support transformers loading quantized moe model (#1067)

mengniwang95 committed 29 days ago

Verified b1b60e47

Add LLMC integration test (#1053)

yiliu30 committed 29 days ago

Verified 9c02a926

export kv scheme (#1068)

yiliu30 committed 29 days ago

Verified e96756c1

Optimize CPU CI pipelines (#1071)

XuehaoSun committed 29 days ago

Verified 2ef7f318

Add compatibility test for ARM (#1073)

XuehaoSun committed 30 days ago

Verified be56548a

support for gguf mixed q2_k_s (#1059)

n1ck-guo committed 30 days ago

Verified e5755588

remove tbb (#1069)

yiliu30 committed 30 days ago

Verified e1ec8552

Add static FP8 attention support (#1061)

yiliu30 committed 31 days ago

Verified 6bde0e11

fix transformers==4.57.1 in CI (#1066)

XuehaoSun committed 31 days ago

Verified ede64fa4

refine readme (#1063)

wenhuach21 committed 31 days ago

Verified 2102257e

Revert "Add static FP8 attention support (#1045)" (#1060)

yiliu30 committed 31 days ago

Verified 6d40e20a

Enhance tokenizer saving (#1057)

mengniwang95 committed 31 days ago

Verified 4ec50d1c

Add static FP8 attention support (#1045)

yiliu30 committed 31 days ago

Verified c5b1c412

[High Risk]reduce vram usage for optimized RTN mode (#1043)

wenhuach21 committed 32 days ago

Verified e1b89d23

Announce llmc integration (#1055)

yiliu30 committed 34 days ago

Verified 90871a71

set add_bos_token=True for llama model (#1046)

n1ck-guo committed 35 days ago

Verified 924a4682

Update get_block_names func (#1047)

mengniwang95 committed 35 days ago

Verified 548acbb0

add memory monitor and import auto-scheme on demand (#1049)

wenhuach21 committed 35 days ago

Verified 247a9a7b

add environment.md and remove AutoRoundMLLM usage in readme (#1042)

xin3he committed 37 days ago

Verified c640c725

fix accuracy regression (#1041)

wenhuach21 committed 38 days ago

Verified 6a8ac7e0

Older