[quant] supports act_order inputs in Matmulnbits and new quantization algorithm "hqq" #19106
wejoncy
changed the title Jicwen/matmulnbits gptq [quant] matmulnbits support 2-8 bits, act_order, gptq/hqq 2 years ago
wejoncy
changed the title [quant] matmulnbits support 2-8 bits, act_order, gptq/hqq [quant] matmulnbits supports 2-8 bits with act_order and new quantization "hqq" 2 years ago
wejoncy
force pushed
from
01bfa243
to
555e34a7
2 years ago
wejoncy
changed the title [quant] matmulnbits supports 2-8 bits with act_order and new quantization "hqq" [quant] supports 2-8 bits kernel with act_order inputs in Op Matmulnbits and new quantization "hqq" 2 years ago
wejoncy
force pushed
from
8ec1cabc
to
6229221f
2 years ago
wejoncy
force pushed
from
ed1cd8c5
to
1594d7b6
2 years ago
wejoncy
force pushed
from
61175510
to
2e58ea25
2 years ago
wejoncy
force pushed
from
de14fbc6
to
1a2328ec
2 years ago
wejoncy
changed the title [quant] supports 2-8 bits kernel with act_order inputs in Op Matmulnbits and new quantization "hqq" [quant] supports act_order inputs in Matmulnbits and new quantization algorithm "hqq" 2 years ago
wejoncy
marked this pull request as ready for review 2 years ago
wejoncy
force pushed
from
9096cc18
to
2dbf3299
2 years ago
wejoncy
force pushed
from
8084a75a
to
1167ad7e
2 years ago
wejoncy
force pushed
from
e011c2b8
to
012227ce
2 years ago
support matmulnbits with hqq
9f146e92
cuda kernel ready
1a011abd
python ut, hqq ut
138e53be
fix build warning && cpp lint
430bc4f2
wejoncy
force pushed
from
012227ce
to
fd51543e
2 years ago
wejoncy
force pushed
from
0deef36f
to
9449f215
2 years ago
format
6a3caa6a
wejoncy
force pushed
from
9449f215
to
6a3caa6a
2 years ago
lint
e1f9bbe5
yufenglee
approved these changes
on 2024-03-05
wejoncy
merged
7e613ee8
into main 2 years ago
wejoncy
deleted the jicwen/matmulnbits_gptq branch 2 years ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub