README.md

metadata

base_model:
  - NousResearch/Hermes-3-Llama-3.1-8B
  - lcw99/llama-3-8b-it-ko-chang
  - maum-ai/Llama-3-MAAL-8B-Instruct-v0.1
  - maywell/Llama-3-Ko-8B-Instruct
  - meta-llama/Meta-Llama-3.1-8B
  - tesser-ai/Tesser-Llama-3-Ko-8B
  - beomi/Llama-3-Open-Ko-8B-Instruct-preview
  - openchat/openchat-3.6-8b-20240522
  - MLP-KTLim/llama-3-Korean-Bllossom-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 Model Stock merge method using NousResearch/Hermes-3-Llama-3.1-8B as a base.

Models Merged

The following models were included in the merge:

• lcw99/llama-3-8b-it-ko-chang
• maum-ai/Llama-3-MAAL-8B-Instruct-v0.1
• maywell/Llama-3-Ko-8B-Instruct
• meta-llama/Meta-Llama-3.1-8B
• tesser-ai/Tesser-Llama-3-Ko-8B
• beomi/Llama-3-Open-Ko-8B-Instruct-preview
• openchat/openchat-3.6-8b-20240522
• MLP-KTLim/llama-3-Korean-Bllossom-8B

Configuration

The following YAML configuration was used to produce this model:

slices:
  - sources:
      - model: openchat/openchat-3.6-8b-20240522
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: MLP-KTLim/llama-3-Korean-Bllossom-8B
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: lcw99/llama-3-8b-it-ko-chang
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: beomi/Llama-3-Open-Ko-8B-Instruct-preview
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: maywell/Llama-3-Ko-8B-Instruct
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: tesser-ai/Tesser-Llama-3-Ko-8B
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: maum-ai/Llama-3-MAAL-8B-Instruct-v0.1
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: meta-llama/Meta-Llama-3.1-8B
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
      - model: NousResearch/Hermes-3-Llama-3.1-8B
        layer_range: [0, 32]
        parameters:
            filter_wise: false # calculate with per-tensor. Not recommended
merge_method: model_stock
base_model: NousResearch/Hermes-3-Llama-3.1-8B
dtype: bfloat16