onnxruntime
GatherBlockQuantized supports zero points and 8 bits for uint8 dtype
#25214
Merged

GatherBlockQuantized supports zero points and 8 bits for uint8 dtype #25214

tianleiwu merged 3 commits into main from tlwu/gather_block_quant_8bits
tianleiwu
tianleiwu tianleiwu marked this pull request as draft 174 days ago
tianleiwu tianleiwu changed the title GatherBlockQuantized adds bits attributes and support zero points for uint8 dtype GatherBlockQuantized supports zero points and 8 bits for uint8 dtype 174 days ago
tianleiwu tianleiwu marked this pull request as ready for review 172 days ago
tianleiwu tianleiwu requested a review from sushraja-msft sushraja-msft 171 days ago
tianleiwu tianleiwu requested a review from jiafatom jiafatom 171 days ago
tianleiwu tianleiwu requested a review from kunal-vaishnavi kunal-vaishnavi 171 days ago
tianleiwu tianleiwu marked this pull request as draft 171 days ago
tianleiwu tianleiwu force pushed from f7257432 to 63c8f9af 166 days ago
tianleiwu tianleiwu marked this pull request as ready for review 166 days ago
tianleiwu tianleiwu marked this pull request as draft 166 days ago
tianleiwu tianleiwu marked this pull request as ready for review 166 days ago
tianleiwu Support bits=4 and zero point for uint8 input
a4784ef3
tianleiwu tianleiwu force pushed from 39b55db3 to a4784ef3 165 days ago
tianleiwu refine
ff9ca7e1
tianleiwu fix test
8f1d98d9
jiafatom
jiafatom commented on 2025-07-09
jiafatom
jiafatom approved these changes on 2025-07-09
kunal-vaishnavi
kunal-vaishnavi approved these changes on 2025-07-09
tianleiwu tianleiwu merged ce8796dd into main 165 days ago
tianleiwu tianleiwu deleted the tlwu/gather_block_quant_8bits branch 165 days ago

Login to write a write a comment.

Login via GitHub

Assignees
No one assigned
Labels
Milestone