intel/auto-round

Pull Requests Commits

yiliu30 committed 44 days ago

70f54fe0

yiliu30 committed 44 days ago

da6fe165

yiliu30 committed 44 days ago

329f67de

yiliu30 committed 44 days ago

e566783d

add auto offload back

yiliu30 committed 44 days ago

b9dd50f1

yiliu30 committed 45 days ago

cb10a466

yiliu30 committed 45 days ago

4e72f6f0

yiliu30 committed 45 days ago

fe53fb3a

yiliu30 committed 45 days ago

6800cfe0

fix multiple devices issue in Compressor and AutoScheme (#1007)

wenhuach21 committed 46 days ago

Verified 344b40cf

yiliu30 committed 46 days ago

b01c91ff

Fix non auto device map (#1005)

WeiweiZhang1 committed 47 days ago

Verified cd54e705

fix multiple devices map issue in calibration (#1003)

wenhuach21 committed 47 days ago

Verified ead6f296

Fix diffusion multi-device ut issue (#1002)

mengniwang95 committed 47 days ago

Verified f6745fd9

Support for immediate saving to reduce ram usage (#965)

Kaihui-intel committed 47 days ago

Verified daeb3bb7

Refine exllamav2 ut (#1001)

WeiweiZhang1 committed 48 days ago

Verified 758b2390

refine md tables (#994)

WeiweiZhang1 committed 48 days ago

Verified c0aa30ab

fix mllm device_map ut (#1000)

Kaihui-intel committed 48 days ago

Verified 4aeca3ae

fix cuda ut bug (#999)

n1ck-guo committed 48 days ago

Verified d1bf7e8d

yiliu30 committed 48 days ago

e2e2d42b

Reduce peak gpu memory usage and support moe estimation (#981)

xin3he committed 48 days ago

Verified 84e9a776

fix lm head bug and rm clear_mem_reach_threhold (#997)

wenhuach21 committed 48 days ago

Verified 284eecdd

[CI] Update python to 3.12 and torch to 2.8.0 (#741)

XuehaoSun committed 49 days ago

Verified 268f7dda

yiliu30 committed 49 days ago

0354c2ba

yiliu30 committed 49 days ago

b992c319

root committed 49 days ago

2bd3c4b1

fix bug of cannot create adam compressor (#992)

n1ck-guo committed 49 days ago

Verified 05cab090

root committed 50 days ago

553ee5c8

support model_dtype and fix bug of scheme contains quotes, mllm eval (#985)

n1ck-guo committed 50 days ago

Verified cd308ddb

add support for Magistral-Small (#980)

n1ck-guo committed 50 days ago

Verified dc6a6d3f

Older