---
base_model:
- Fizzarolli/L3.1-70b-glitz-v0.2
- KaraKaraWitch/L3.1-70b-MeowMix2
- sophosympatheia/New-Dawn-Llama-3.1-70B-v1.1
- nothingiisreal/L3.1-70B-Celeste-V0.1-BF16
- Sao10K/L3.1-70B-Hanami-x1
tags:
- merge
- mergekit
- lazymergekit
- Fizzarolli/L3.1-70b-glitz-v0.2
- KaraKaraWitch/L3.1-70b-MeowMix2
- sophosympatheia/New-Dawn-Llama-3.1-70B-v1.1
- nothingiisreal/L3.1-70B-Celeste-V0.1-BF16
- Sao10K/L3.1-70B-Hanami-x1
license: cc-by-nc-4.0
---

# KaraKaraWitch/L3.1-70b-Milasha

Milasha is my 3rd and final series with Llama 3.1.

![image/png](https://cdn-uploads.huggingface.co/production/uploads/633e85093a17ab61de8d9073/90ZI1ocoKwQyQXXwHnt_s.png)

L3.1-70b-Milasha is a merge of the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing) running on a server:

- Fizzarolli/L3.1-70b-glitz-v0.2
- KaraKaraWitch/L3.1-70b-MeowMix2
- sophosympatheia/New-Dawn-Llama-3.1-70B-v1.1
- nothingiisreal/L3.1-70B-Celeste-V0.1-BF16
- Sao10K/L3.1-70B-Hanami-x1

**NOTE**: Following Sao10k's preferences, this model is licenced as cc-by-nc-4.0

This model kinda returns to tradition of merging the best models. MeowMix was used as language reinforcement for CJK while the rest of the models are more or less from the first in the series.

## Yap / Chat Format

Llama 3 Instruct.

## Thoughts, etc

As I'm writing this, I'm overseas waiting for beef to arrive at my table. I'll probably finish writing this up properly later.

This model uses model stock. The rest just refer to the merge config file.