Abus-7B-Instruct / README.md
marcuscedricridia's picture
Update README.md
ee6dde5 verified
---
base_model:
- marcuscedricridia/cursa-o1-7b
- marcuscedricridia/sbr-o1-7b
- marcuscedricridia/absolute-o1-7b
library_name: transformers
tags:
- mergekit
- merge
---
> [!TIP] This model is currently ranked #3 among the models up to 7B parameters and #426 among all models on the [Open LLM Leaderboard](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard).
# 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 [Linear DARE](https://arxiv.org/abs/2311.03099) merge method using [marcuscedricridia/cursa-o1-7b](https://huggingface.co/marcuscedricridia/cursa-o1-7b) as a base.
### Models Merged
The following models were included in the merge:
* [marcuscedricridia/sbr-o1-7b](https://huggingface.co/marcuscedricridia/sbr-o1-7b)
* [marcuscedricridia/absolute-o1-7b](https://huggingface.co/marcuscedricridia/absolute-o1-7b)
### Configuration
The following YAML configuration was used to produce this model:
```yaml
model_name: "pre-cursa-o1-v1.2"
models:
- model: marcuscedricridia/cursa-o1-7b
#no parameters necessary for base model
- model: marcuscedricridia/sbr-o1-7b
parameters:
density: 0.5
weight: 0.5
- model: marcuscedricridia/absolute-o1-7b
parameters:
density: 0.5
weight: 0.5
merge_method: dare_linear
base_model: marcuscedricridia/cursa-o1-7b
parameters:
normalize: false
int8_mask: true
dtype: bfloat16
tokenizer_source: "union" # or "base" or a model path
chat_template: "auto" # or a template name or Jinja2 template
```