Synesthesia / README.md
schonsense's picture
Update README.md
e1adf01 verified
---
base_model:
- schonsense/Diagesis
- schonsense/IPOplectic
- schonsense/Tropoplectic
- schonsense/llama33_inst_multivector_derestriction
- schonsense/llama31st_diag
library_name: transformers
tags:
- mergekit
- merge
---
# Synesthesia2
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
![tarimage](https://cdn-uploads.huggingface.co/production/uploads/6317d4867690c5b55e61ce3d/fy8QbVuv-2iSMgH4IommM.png)
## Merge Details
### Merge Method
This model was merged using the [Model Stock](https://arxiv.org/abs/2403.19522) merge method using [schonsense/llama33_inst_multivector_derestriction](https://huggingface.co/schonsense/llama33_inst_multivector_derestriction) as a base.
### Models Merged
The following models were included in the merge:
* [schonsense/Diagesis](https://huggingface.co/schonsense/Diagesis)
* [schonsense/IPOplectic](https://huggingface.co/schonsense/IPOplectic)
* [schonsense/Tropoplectic](https://huggingface.co/schonsense/Tropoplectic)
* [schonsense/llama31st_diag](https://huggingface.co/schonsense/llama31st_diag)
### Configuration
The following YAML configuration was used to produce this model:
```yaml
models:
- model: schonsense/llama33_inst_multivector_derestriction
- model: schonsense/llama31st_diag
- model: schonsense/Diagesis
- model: schonsense/Tropoplectic
- model: schonsense/IPOplectic
base_model: schonsense/llama33_inst_multivector_derestriction
merge_method: model_stock
dtype: float32
out_dtype: bfloat16
chat_template: llama3
tokenizer:
source: union
pad_to_multiple_of: 8
```