---
base_model:
- TareksLab/Citrine-V2-LLaMa-70B
- TareksLab/Emerald-V2-LLaMa-70B
- TareksLab/Carnelian-V2-LLaMa-70B
- TareksLab/Amethyst-V2-LLaMa-70B
- TareksLab/Ruby-V2-LLaMa-70B
- TareksLab/Sapphire-V2-LLaMa-70B
library_name: transformers
tags:
- mergekit
- merge

---
# MERGED

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 [SCE](https://arxiv.org/abs/2408.07990) merge method using [TareksLab/Emerald-V2-LLaMa-70B](https://huggingface.co/TareksLab/Emerald-V2-LLaMa-70B) as a base.

### Models Merged

The following models were included in the merge:
* [TareksLab/Citrine-V2-LLaMa-70B](https://huggingface.co/TareksLab/Citrine-V2-LLaMa-70B)
* [TareksLab/Carnelian-V2-LLaMa-70B](https://huggingface.co/TareksLab/Carnelian-V2-LLaMa-70B)
* [TareksLab/Amethyst-V2-LLaMa-70B](https://huggingface.co/TareksLab/Amethyst-V2-LLaMa-70B)
* [TareksLab/Ruby-V2-LLaMa-70B](https://huggingface.co/TareksLab/Ruby-V2-LLaMa-70B)
* [TareksLab/Sapphire-V2-LLaMa-70B](https://huggingface.co/TareksLab/Sapphire-V2-LLaMa-70B)

### Configuration

The following YAML configuration was used to produce this model:

```yaml
models:
  - model: TareksLab/Emerald-V2-LLaMa-70B
    parameters:
      select_topk: [0.1, 0.1, 0.1, 0.1, 0.2, 0.5]
  - model: TareksLab/Carnelian-V2-LLaMa-70B
    parameters:
      select_topk: [0.1, 0.1, 0.1, 0.2, 0.4, 0.2]
  - model: TareksLab/Ruby-V2-LLaMa-70B
    parameters:
      select_topk: [0.1, 0.1, 0.2, 0.4, 0.2, 0.1]
  - model: TareksLab/Amethyst-V2-LLaMa-70B
    parameters:
      select_topk: [0.1, 0.2, 0.4, 0.2, 0.1, 0.1]
  - model: TareksLab/Citrine-V2-LLaMa-70B
    parameters:
      select_topk: [0.2, 0.4, 0.2, 0.1, 0.1, 0.1]
  - model: TareksLab/Sapphire-V2-LLaMa-70B
    parameters:
      select_topk: [0.5, 0.2, 0.1, 0.1, 0.1, 0.1]
base_model: TareksLab/Emerald-V2-LLaMa-70B
merge_method: sce
parameters:
  int8_mask: true
dtype: float32
out_dtype: bfloat16
chat_template: llama3
tokenizer:
 source: TareksLab/Ruby-V2-LLaMa-70B
 pad_to_multiple_of: 8

```