bytedance/ABQ-LLM

View on GitHub

An acceleration library that supports arbitrary bit-width combinatorial quantization operations

Stars: 249
Forks: 22
Open beginner issues: 0
Indexed issues: 15

Dominant language: C++
License: Apache-2.0
Last GitHub push: Sep 30, 2024
Latest indexed: Sep 14, 2026
Contributing guide: No contributing guide
Code of conduct: No code of conduct
Beginner labels: No beginner labels indexed
PR merge metrics: No merged PRs in 30d

Beginner friendly Everything

Open All

15 open issues indexed

Why is this repo no longer maintained Open

Difficulty 5/5 Over a week Newbie friendliness 10/100

bytedance/ABQ-LLM#20 · Opened Jun 17, 2025
Poor Performance of W4A4 on MMLU task Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

bytedance/ABQ-LLM#19 · 1 comment · Opened Nov 25, 2024
[warning] loading model fails! Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

bytedance/ABQ-LLM#18 · 1 comment · Opened Nov 8, 2024
Question about W4A16 Benchmark Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

bytedance/ABQ-LLM#17 · 2 comments · Opened Oct 22, 2024
[Question] The end-to-end generation speed and W4A4 Open

Difficulty 5/5 Over a week Newbie friendliness 15/100

bytedance/ABQ-LLM#16 · Opened Oct 13, 2024
No reduction in model size Open

Difficulty 3/5 1-2 days Newbie friendliness 35/100

bytedance/ABQ-LLM#15 · 6 comments · Opened Oct 1, 2024
Seqlen of Kernel Benchmark Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

bytedance/ABQ-LLM#14 · 2 comments · Opened Sep 30, 2024
CUDA kernel of weight only quantization Open

Difficulty 1/5 Under an hour Newbie friendliness 25/100

bytedance/ABQ-LLM#13 · 1 comment · Opened Sep 30, 2024
About W2A16 weight only matmul Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

bytedance/ABQ-LLM#10 · 2 comments · Opened Sep 26, 2024
Is there a plan to support model Qwen2? Open

Difficulty 5/5 Over a week Newbie friendliness 20/100

bytedance/ABQ-LLM#9 · 2 comments · 1 reaction · Opened Sep 24, 2024
Custom datasets support for chat models Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

bytedance/ABQ-LLM#8 · 1 comment · Opened Sep 23, 2024
The link to model zoo Open

Difficulty 1/5 Under an hour Newbie friendliness 25/100

bytedance/ABQ-LLM#7 · 1 comment · Opened Sep 23, 2024
How to reproduce the results of end-to-end throughput experiment? Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

bytedance/ABQ-LLM#3 · 4 comments · Opened Sep 9, 2024
What is the difference between pack4 and pack, and why is pack4 used in the file test_w4a4_int_wmma.cu? Open

Difficulty 2/5 1-3 hours Newbie friendliness 35/100

bytedance/ABQ-LLM#2 · 2 comments · Opened Sep 4, 2024
The methods in this article are really cool !!! Open

Difficulty 5/5 Over a week Newbie friendliness 10/100

bytedance/ABQ-LLM#1 · 1 comment · 1 reaction · Opened Sep 2, 2024

bytedance/ABQ-LLM

Get new issues in your inbox