abetlen / abetlen/llama-cpp-python

Have you thought about adding quantum cache or 8 bit cache?

未关闭
#1,161 2 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看
enhancement
主要语言
Python
星标
10.6k
派生
1.4k
PR 合并指标
PR 指标待抓取

描述

I've figured out how to change the type. It's a little slower but it does save memory.

The numbers for type_K or V simply correspond to the cache types in ggml.h

```
GGML_TYPE_F32 = 0,
GGML_TYPE_F16 = 1,
GGML_TYPE_Q4_0 = 2,
GGML_TYPE_Q4_1 = 3,
// GGML_TYPE_Q4_2 = 4, support has been removed
// GGML_TYPE_Q4_3 (5) support has been removed
GGML_TYPE_Q5_0 = 6,
GGML_TYPE_Q5_1 = 7,
GGML_TYPE_Q8_0 = 8,
GGML_TYPE_Q8_1 = 9,
```

贡献指南

打开贡献指南

评估

这个 Issue 还没有评估数据。

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。