intel/auto-round

Pull Requests Commits

Address code review feedback: improve documentation and remove unused code

Copilot committed 102 days ago

a6a69b63

Replace get_module and set_module with PyTorch native APIs

Copilot committed 102 days ago

83a0ed16

Copilot committed 102 days ago

b6678ceb

Refactor FP8 dequantization and detection using registry pattern (#1348)

scopophobic committed 103 days ago

Verified eaabde4a

refactor auto_round api, using local symbol table to pass all args (#1358)

n1ck-guo committed 103 days ago

Verified 528c9c82

Autoround in vLLM Office Hours (#1322)

yiliu30 committed 103 days ago

Verified b3f6f1fb

Fix cpu ut for transformers v5 (#1333)

Kaihui-intel committed 103 days ago

Verified f6e7e303

Fix CPU CI (#1344)

XuehaoSun committed 104 days ago

Verified 8eec5d8a

add copilot-instructions (#1341)

xin3he committed 104 days ago

Verified 0fe5c0a7

refactor eval and add UT (#1324)

xin3he committed 105 days ago

Verified 14eacb1f

Add Chinese README (#1335)

ZaneMark committed 105 days ago

Verified 71b385d2

Fix issue where predefined ignore layers incorrectly apply to GGUF (#1340)

wenhuach21 committed 105 days ago

Verified ca133621

Fix attribute bug for diffusion model (#1338)

mengniwang95 committed 105 days ago

Verified a4ee3a00

update release package and bump version (#1327)

chensuyue committed 106 days ago

Verified 00702194

format add check for scheme args (#1332)

n1ck-guo committed 106 days ago

Verified 7f1dee6e

Delay materializing the replaced model weights until quantization (#1307)

yiliu30 committed 106 days ago

Verified d9d20a12

(feat): add support for g2 fp8 on cpu with LUT (#1254)

SwekeR-463 committed 107 days ago

Verified 6190869b

fix tied weights keys bug (#1330)

wenhuach21 committed 108 days ago

Verified 25364d21

Update publication dates in README and add recipe configuration detai… (#1329)

xin3he committed 108 days ago

Verified 46224d4a

[GGUF] using quant_nontext_module to control whether quant vision model (#1317)

n1ck-guo committed 108 days ago

Verified 24e43976

simplify pull_request_template.md (#1328)

xin3he committed 108 days ago

Verified cf37d8e6

add blockwise way in pure rtn to ease module replacements (#1320)

wenhuach21 committed 109 days ago

Verified 57d1a8ac

set predefined ignore layers (#1318)

wenhuach21 committed 109 days ago

Verified 7c397e14

fix low_cpu and set it to True as default (#1288)

wenhuach21 committed 109 days ago

Verified f654e52d

Add issue templates and PR template with required fields (#1309)

Copilot committed 109 days ago

Verified eaf5885a

fix XPU CI hang issue (#1261)

xin3he committed 110 days ago

Verified 57af5b7f

fix bug of format (#1287)

n1ck-guo committed 110 days ago

Verified 5ba31b6c

Enable q_scale calibration for deepseek v2 on Gaudi (#1299)

mengniwang95 committed 110 days ago

Verified d8ed35a8

open source delta loss (#1300)

wenhuach21 committed 111 days ago

Verified 587eadb8

fix ignore layers regression (#1302)

wenhuach21 committed 111 days ago

Verified 3352a830

Older