Nexesenex's picture
Update README.md
164d55e verified
---
base_model:
- mlabonne/Hermes-3-Llama-3.1-70B-lorablated
- SicariusSicariiStuff/Negative_LLAMA_70B
- Nexesenex/Llama_3.x_70b_Smarteaz_V1
library_name: transformers
tags:
- mergekit
- merge
license: llama3.3
metrics:
- perplexity
- accuracy
---
# About.
A solid model. Not too fancy, only serious stuff for a smart model.
Most of the best 70b finetunes packed into one multi-echelon merge, with the excellent Negative Llama as a Lead.
- Sidekick A : Smarteaz is my L3.3 70b smart base with the best models abliterated by Huihui-AI, with Deepseek R1 Distill, Tulu and Nemotron (3.1) mixed within.
- Sidekick B : Hermes is a very qualitative L3.1 70b model, lorablated by Maxime Labonne.
Both models empower Negative Llama and make it smarter than ever.
Nemesis V1.1 can be used as a standalone, or a smart-base for merges like Smarteaz, but I fear that it might dilute its qualities.
Smart roleplay is also possible, of course. It's not the funniest model, though.
- Perplexity Wikitext eng 512 : 3.35. (member of my best class)
- Arc-C : 62 (same)
- ARC-E : 83 (same)
---
# merge
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
## Merge Details
### Merge Method
This model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [SicariusSicariiStuff/Negative_LLAMA_70B](https://huggingface.co/SicariusSicariiStuff/Negative_LLAMA_70B) as a base.
### Models Merged
The following models were included in the merge:
* [mlabonne/Hermes-3-Llama-3.1-70B-lorablated](https://huggingface.co/mlabonne/Hermes-3-Llama-3.1-70B-lorablated)
* [Nexesenex/Llama_3.x_70b_Smarteaz_V1](https://huggingface.co/Nexesenex/Llama_3.x_70b_Smarteaz_V1)
### Configuration
The following YAML configuration was used to produce this model:
```yaml
merge_method: model_stock
models:
- model: mlabonne/Hermes-3-Llama-3.1-70B-lorablated
parameters:
weight: 1.0
- model: Nexesenex/Llama_3.x_70b_Smarteaz_V1
parameters:
weight: 1.0
base_model: SicariusSicariiStuff/Negative_LLAMA_70B
dtype: bfloat16
normalize: true
```