README.md

metadata

base_model:
  - Undi95/Meta-Llama-3.1-8B-Claude
  - safe049/ParuMaid-Llama3-Chinese-8B
library_name: transformers
tags:
  - mergekit
  - merge

merge

This is a merge of pre-trained language models created using mergekit.

Merge Details

Merge Method

This model was merged using the TIES merge method using safe049/ParuMaid-Llama3-Chinese-8B as a base.

Models Merged

The following models were included in the merge:

• Undi95/Meta-Llama-3.1-8B-Claude

Configuration

The following YAML configuration was used to produce this model:

# 使用 ties 方法进行模型合并的配置文件

base_model: "safe049/ParuMaid-Llama3-Chinese-8B"  # 设置 ParuMaid-Llama3-Chinese-8B 作为基础模型

models:
  - model: "safe049/ParuMaid-Llama3-Chinese-8B"
  - model: "Undi95/Meta-Llama-3.1-8B-Claude"
    parameters:
      density: 0.5  # 控制合并中的稀疏度
      weight: 0.5  # 合并权重，控制该模型对输出的影响

merge_method: ties  # 选择 ties 作为合并方法

parameters:
  normalize: false  # 是否进行归一化处理
  int8_mask: true  # 使用 int8 掩码以降低内存占用
dtype: float16  # 设置为半精度浮点数以减少显存需求