intel/auto-round

Pull Requests Commits

Merge branch 'main' into copilot/fix-issue-with-auto-rounding

yiliu30 committed 43 days ago

Verified 4775df98

yiliu30 committed 43 days ago

3bee2e5c

auto-round-kernel installation method (#1221)

chensuyue committed 44 days ago

Verified ba48dd77

Address review feedback: remove unused import and add None test

Copilot committed 44 days ago

17fbf3ae

Address code review feedback: improve test mocking and remove debug print

Copilot committed 44 days ago

4f77d6d2

Add comprehensive unit tests for get_fp_layer_names function

Copilot committed 44 days ago

97abed09

Add test for ignore_layers with FP8 models

Copilot committed 44 days ago

9f0eaf89

Fix ignore_layers for FP8 models by checking INNER_SUPPORTED_LAYER_TYPES

Copilot committed 44 days ago

f8b62fc4

Copilot committed 44 days ago

d2758f8c

[vllm-ext] allows setting flashinfer workspace (#1259)

yiliu30 committed 44 days ago

Verified 6e1ea9cb

add permission for workflow (#1190)

chensuyue committed 45 days ago

Verified 94db9057

Update HPU CI to gaudi v1.23.0 (#1271)

XuehaoSun committed 45 days ago

Verified 56750e8a

gguf fix bug of tensors not on same tensor and gguf:q2_k_mixed (#1263)

n1ck-guo committed 45 days ago

Verified 2c67712f

Fix critic bug causing GGUF to run on CPU (#1260)

wenhuach21 committed 46 days ago

Verified 1b7535a0

update supported scheme readme (#1257)

wenhuach21 committed 46 days ago

Verified 6a87a651

Update nightly release workflow (#1256)

chensuyue committed 47 days ago

Verified 583278d9

Add nightly release workflow (#1255)

chensuyue committed 47 days ago

Verified b4ab21d7

GGUF format add support for MoE models with non-linear expert layers. (#1244)

n1ck-guo committed 47 days ago

Verified c7d6aee6

fix cuda ut fail (#1237)

n1ck-guo committed 47 days ago

Verified 7a46aee8

Fix low cpu speed issue (#1251)

wenhuach21 committed 50 days ago

Verified 90db4445

WNA16 does not apply optimized RTN for moe layers by default (#1245)

wenhuach21 committed 50 days ago

Verified 9588bf90

Organize test directory structure with logical categorization (#1243)

Copilot committed 50 days ago

Verified 0f6dc76e

Update setting of disable opt rtn (#1249)

WeiweiZhang1 committed 50 days ago

Verified cc521cab

set disable_opt_rtn to optional bool and change default value to None (#1231)

WeiweiZhang1 committed 50 days ago

Verified bef28c9a

[HPU]Export lazy mode env explicit (#1239)

yiliu30 committed 51 days ago

Verified 001d0e39

[API Change]rename fp_layers to ignore_layers (#1233)

wenhuach21 committed 51 days ago

Verified 0ad224c9

Add FP8 dtype info to quant config (#1228)

yiliu30 committed 51 days ago

Verified afea815e

gguf add support for Mistral-Magistral-Devstral and Granite (#1226)

n1ck-guo committed 52 days ago

Verified 6f2f9b9e

[STEP 2] refactor format and export (#1192)

n1ck-guo committed 53 days ago

Verified c2449059

align input scale for moe experts for fp8 (#1216)

SwekeR-463 committed 53 days ago

Verified 41a83773

Older