Pushing the Limits of Large Language Model Quantization via the Linearity Theorem
Paper
•
2411.17525
•
Published
HIGGS
Models prequantized with [HIGGS](https://arxiv.org/abs/2411.17525) zero-shot quantization. Requires the latest `transformers` to run.
-
-
ISTA-DASLab/Llama-3.3-70B-Instruct-HIGGS-GPTQ-4bit
Updated • 46 • 2 -
ISTA-DASLab/Llama-3.1-8B-Instruct-HIGGS-GPTQ-4bit
Text Generation • Updated • 9 -
ISTA-DASLab/Llama-3.1-8B-Instruct-HIGGS-GPTQ-3bit
Text Generation • Updated • 19 -
ISTA-DASLab/Llama-3.1-8B-HIGGS-GPTQ-4bit
Text Generation • Updated • 12 -
ISTA-DASLab/Llama-3.1-8B-HIGGS-GPTQ-3bit
Text Generation • Updated • 22 -
ISTA-DASLab/Llama-3.3-70B-Instruct-HIGGS-4bit
Text Generation • Updated • 29 • 3 -
ISTA-DASLab/Llama-3.3-70B-Instruct-HIGGS-3bit
Text Generation • Updated • 9 -
ISTA-DASLab/Llama-3.1-70B-Instruct-HIGGS-4bit
Text Generation • Updated • 8 -
ISTA-DASLab/Llama-3.1-70B-Instruct-HIGGS-3bit
Text Generation • Updated • 7 -
ISTA-DASLab/Llama-3.1-8B-Instruct-HIGGS-4bit
Text Generation • Updated • 30 -
ISTA-DASLab/Llama-3.1-8B-Instruct-HIGGS-3bit
Text Generation • Updated • 11 -
ISTA-DASLab/Llama-3.1-8B-HIGGS-4bit
Text Generation • Updated • 38 -
ISTA-DASLab/Llama-3.1-8B-HIGGS-3bit
Text Generation • Updated • 29 -
ISTA-DASLab/gemma-2-9b-it-HIGGS-4bit
Text Generation • Updated • 11 -
ISTA-DASLab/gemma-2-9b-it-HIGGS-3bit
Text Generation • Updated • 12 -
ISTA-DASLab/gemma-2-9b-HIGGS-4bit
Text Generation • Updated • 14 -
ISTA-DASLab/gemma-2-9b-HIGGS-3bit
Text Generation • Updated • 12