microsoft/TransformerCompression

View on GitHub

For releasing code related to compression methods for transformers, accompanying our publications

Stars: 460
Forks: 58
Open beginner issues: 0
Indexed issues: 20
Avg merge: 8h 40m
Merged PRs (30d): 1

Dominant language: Python
License: MIT
Last GitHub push: Sep 10, 2026
Latest indexed: Sep 19, 2026
Contributing guide: No contributing guide
Code of conduct: Code of conduct
Beginner labels: good first issue help wanted

Beginner friendly Everything

Open All

20 open issues indexed

Loading models is not easy with `from_pretrained` Open

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

microsoft/TransformerCompression#188 · Opened Jun 10, 2025
Error when loading sliced llama3.1-70b-Instruct Open

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

microsoft/TransformerCompression#185 · Opened Feb 10, 2025
transformers dependency version upgrade Open

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

microsoft/TransformerCompression#183 · Opened Jan 4, 2025
adapt the lm_eval tool Open

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

microsoft/TransformerCompression#182 · Opened Dec 2, 2024
Some confusion about the equation.6 in the sliceGPT paper. Open

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

microsoft/TransformerCompression#181 · 1 comment · Opened Nov 25, 2024
Speed up test Open

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

microsoft/TransformerCompression#179 · Opened Oct 30, 2024
Issues with LLAMA 3-8B-Instruct model Open

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

microsoft/TransformerCompression#177 · Opened Sep 2, 2024
NaN weights after rotating and slicing Open

Difficulty 4/5 3-5 days Newbie friendliness 20/100

microsoft/TransformerCompression#176 · 1 comment · Opened Aug 17, 2024
Quarot: DeepSeek-V2 Support Open

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

microsoft/TransformerCompression#174 · Opened Jul 12, 2024
How to evaluate the sliced model Open

Difficulty 3/5 1-2 days Newbie friendliness 25/100

microsoft/TransformerCompression#173 · Opened Jul 10, 2024
How to finetune with multi-gpus under data parallel setting? Open

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

microsoft/TransformerCompression#167 · 1 comment · Opened Jun 8, 2024
model inference Open

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

microsoft/TransformerCompression#160 · Opened May 22, 2024
Command R and R+ support Open

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

microsoft/TransformerCompression#133 · 1 comment · Opened Apr 12, 2024
`run_benchmark.py` runs error when using `--distribute-model` Open

bug

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

microsoft/TransformerCompression#122 · 1 comment · Opened Mar 29, 2024
How to export entire model? Open

good first issue help wanted

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

microsoft/TransformerCompression#114 · 10 comments · Opened Mar 6, 2024
Convert Layernorm to RMSnorm Open

Difficulty 4/5 3-5 days Newbie friendliness 20/100

microsoft/TransformerCompression#110 · 6 comments · Opened Mar 1, 2024
Mistral Support Open

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

microsoft/TransformerCompression#81 · 6 comments · 2 reactions · Opened Jan 29, 2024
loading models is painful and not HF compatible Open

microsoft/TransformerCompression#73 · 1 comment · 1 assignee · Opened Jan 23, 2024
c4 dataset download fails Open

bug

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

microsoft/TransformerCompression#56 · 1 comment · Opened Jan 11, 2024
Common command-line args Open

enhancement

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

microsoft/TransformerCompression#52 · Opened Jan 10, 2024

microsoft/TransformerCompression

Get new issues in your inbox