Search results

SmoothQuant on t5?

Status: Open.

#105 In mit-han-lab/smoothquant;

· Manojbhat09 opened on Feb 8, 2025

SmoothQuant kv-cache quantization

Status: Open.

#104 In mit-han-lab/smoothquant;

· wuguowuge opened on Feb 7, 2025

Sensitivity analysis of model parameters

Status: Open.

#102 In mit-han-lab/smoothquant;

· shengqihailuo1 opened on Jan 16, 2025

RuntimeError: “triu_tril_cuda_template“ not implemented for ‘BFloat16‘

Status: Open.

#101 In mit-han-lab/smoothquant;

· cuicui2023 opened on Jan 3, 2025

error when getting the scales for my model

Status: Open.

#100 In mit-han-lab/smoothquant;

· rib12316 opened on Dec 19, 2024

How to use this for custom model?

Status: Open.

#98 In mit-han-lab/smoothquant;

· siddagra opened on Dec 7, 2024

how to visual outier of activation

Status: Open.

#97 In mit-han-lab/smoothquant;

· harleyszhang opened on Nov 3, 2024

The upper and lower bounds seems to not 8 bits in some cases

Status: Open.

#96 In mit-han-lab/smoothquant;

· zhangyu68 opened on Oct 30, 2024

Why only 4 layers?

Status: Open.

#95 In mit-han-lab/smoothquant;

· VincentXWD opened on Sep 7, 2024

Support for Qwen2

Status: Open.

#94 In mit-han-lab/smoothquant;

· JiaXinLI98 opened on Jul 31, 2024

How to quantize the out_proj and fc2 module in OPT model family

Status: Open.

#93 In mit-han-lab/smoothquant;

· yanchenmochen opened on Jul 30, 2024

How to quantize llama3?

Status: Open.

#92 In mit-han-lab/smoothquant;

· jpyo0803 opened on Jul 22, 2024