Hugging Face's logo

jusjinuk
/
Llama-3.3-70B-Instruct-3bit-GuidedQuant-LNQ

PyTorch
llama
Model card Files Community
jusjinuk commited on
Commit
f609895
·
verified ·
1 Parent(s): 19a9b30

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -0
README.md CHANGED
@@ -12,6 +12,7 @@ license: mit
12
  - Backend kernel: Any-Precision-LLM kernel (`ap-gemv`)
13
  - Calibration data: RedPajama (1024 sentences / 4096 tokens)
14
  - Calibration objective: Next-token prediction
 
15
 
16
  # How to run
17
  - Follow the instruction in https://github.com/snu-mllab/GuidedQuant.
 
12
  - Backend kernel: Any-Precision-LLM kernel (`ap-gemv`)
13
  - Calibration data: RedPajama (1024 sentences / 4096 tokens)
14
  - Calibration objective: Next-token prediction
15
+ - num_groups (for GuidedQuant Hessian): 1
16
 
17
  # How to run
18
  - Follow the instruction in https://github.com/snu-mllab/GuidedQuant.