intel/auto-round

Pull Requests Commits

Merge branch 'main' into xinhe/9-17

xin3he committed 52 days ago

Verified 155d1af2

fix:reduce memory usage when loading native fused MoE checkpoints (#2382)

n1ck-guo committed 52 days ago

Verified a5f6ddd8

fix: restore FP32 tensors saved as FP16/BF16 during tensor copying (#2375)

xin3he committed 54 days ago

Verified a6c612fa

fix: share NVFP4 global scales for Wan QKV projections (#2379)

changwangss committed 54 days ago

Verified 36079c49

feat: update model-free compressor to preserve activation config and enhance evaluation flow

xin3he committed 54 days ago

9ae98a69

[ARK] Improve XPU W4A16 WOQ decode with dense S4 DPAS path (#2331)

Zhenzhong1 committed 54 days ago

Verified eaf57a2a

patch triangular_solve for Qwen3.8 model for hpu (#2377)

lkk12014402 committed 54 days ago

Verified 8cd9b878

feat: enhance index file handling in _write_index_file and add unit tests for single and multiple shard scenarios

xin3he committed 54 days ago

f7f345b5

feat: enhance max_scale handling in qdq method and add unit tests for scale preservation

xin3he committed 54 days ago

71f9532b

Remove deprecated files (#2373)

chensuyue committed 55 days ago

Verified fd077878

fix: correct device mapping in run_model_evaluation function

xin3he committed 55 days ago

738ff52c

feat: implement optimized RTN toggle for NVFP4 E5M3 quantization and update CLI handling

xin3he committed 55 days ago

d1a1fbca

bug fix: fix NVFP4 evaluation with input_ global_scale handling and testing (#2374)

xin3he committed 55 days ago

Verified 27daf7f8

remove bagel specific constraints (#2372)

WeiweiZhang1 committed 55 days ago

Verified 4b3680b9

add fineweb-edu dataset as calibration bakeup (#2345)

WeiweiZhang1 committed 55 days ago

Verified 3938e152

feat: add NVFP4 static KV cache quantization support (#2366)

yiliu30 committed 55 days ago

Verified 04fca622

docs: clarify diffusion export format (#2367)

changwangss committed 55 days ago

Verified ea838e56

Add blockwise rotation. (#2087)

lkk12014402 committed 56 days ago

Verified 9b746ff9

fix: recognize Modular Diffusers pipelines as diffusion models (#2120)

Ar4ikov committed 56 days ago

Verified 0d76712e

refactor: decouple datatype quantization from algorithms (#2360)

n1ck-guo committed 56 days ago

Verified 2e82448a

fix: save the config of diffusers models on export (#2119)

Ar4ikov committed 56 days ago

Verified 1983b65a

feat(ark): fused Hadamard + MXFP4 activation quantization kernel

jijiaz committed 56 days ago

Verified ecc5b005

perf: avoid unused SVD reconstruction and select the CUDA driver in SVDQuant (#2356)

changwangss committed 56 days ago

Verified 903ba0ce

unify target_bits/options into bits/schemes for AutoScheme (#2333)

n1ck-guo committed 56 days ago

Verified 2b5a3c55

Refactor test timeouts and update dependencies for performance (#2361)

XuehaoSun committed 56 days ago

Verified 1965bdf0

fix: preserve declared FP32 tensors during model dtype conversion (#2359)

changwangss committed 57 days ago

Verified b33eedc9

fix: support list inputs in diffusion tuning cache (#2346)

changwangss committed 57 days ago

Verified c53611fb

fix: update key cache access for compatibility with newer DynamicCache implementation (#2348)

chensuyue committed 58 days ago

Verified f331bb02

fix: reload Diffusers blocks from component checkpoints (#2353)

changwangss committed 58 days ago

Verified 6a8d95f4

fix qwen35 regression (#2352)

wenhuach21 committed 60 days ago

Verified 141e4c99

Older