Llama-SoraEToAtetaTegami-3.3-3.0-70B
SAO Good anime.
Merge notes:
- While the model has too much Mirai, it's usable.
- It is VERY finicky and tempremental! Be prepared to reprompt a couple of times for better responses.
- This model is very well a shitpost merge because I thought: "Can we try L3.3 and L3.0/3.1 models lol"
- I'm sure it can do cursed things.
- Multilingual capabilities are suprisingly good.
This is a merge of pre-trained language models created using mergekit.
Recommended Settings
Temperature: 1.6
Top A: 0.05
Min P: 0.05
Like Mirai, it doesn't enjoy Repetition Penalty.
Merge Details
Merge Method
This model was merged using the TIES merge method using EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0 as a base.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
models:
- model: Blackroot/Mirai-70B-1.0
parameters:
density: 0.6
weight: 0.5
- model: EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0
parameters:
density: 0.4
weight: 0.5
merge_method: ties
base_model: EVA-UNIT-01/EVA-LLaMA-3.33-70B-v0.0
parameters:
normalize: true
dtype: bfloat16