GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling
Paper • 2604.18556 • Published • 12
GSQ
GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling, https://huggingface.co/papers/2604.18556
-
-
ISTA-DASLab/Kimi-K2.6-2Bit-GSQ
Image-Text-to-Text • 84B • Updated • 46 • 1 -
ISTA-DASLab/Kimi-K2.5-2Bit-GSQ
Image-Text-to-Text • 84B • Updated • 42 • 1 -
ISTA-DASLab/Llama-3.1-70B-Instruct-2Bit-GSQ
Text Generation • 7B • Updated • 22 -
ISTA-DASLab/Llama-3.1-70B-Instruct-3Bit-GSQ
Text Generation • 9B • Updated • 16 -
ISTA-DASLab/Qwen3.6-35B-A3B-2Bit-GSQ
Image-Text-to-Text • 36B • Updated • 1.89k • 4 -
ISTA-DASLab/Qwen3-4B-GGUF-GSQ
Text Generation • 4B • Updated • 29 -
ISTA-DASLab/Qwen3-8B-GGUF-GSQ
Text Generation • 8B • Updated • 60 • 3 -
ISTA-DASLab/Qwen3.5-4B-GGUF-GSQ
Text Generation • 4B • Updated • 170 • 2 -
ISTA-DASLab/Qwen3.8-27B-3Bit-GSQ
Image-Text-to-Text • 27B • Updated • 2.22k • 9