|
|
--- |
|
|
base_model: |
|
|
- NexesMess/Llama_3.x_70b_Tessessence_0.10c |
|
|
- Steelskull/L3.3-Electra-R1-70b |
|
|
- NexesMess/Llama_3.1_70b_Luminoblivion_CW_V1.01 |
|
|
library_name: transformers |
|
|
tags: |
|
|
- mergekit |
|
|
- merge |
|
|
|
|
|
--- |
|
|
# about |
|
|
|
|
|
Pretty similar to V3, with a theoric emphasis on creativity due to the Luminoblivion Creative Writer sidekick replacing Hostess (and so, Tess R1 out, 70Blivion in). |
|
|
I prefer V3, though, because V3 seems a bit smarter and more coherent. |
|
|
|
|
|
--- |
|
|
# benchs |
|
|
- PPL 512 Wikitext Eng : 3.18 |
|
|
- ARC-C : 61.20 |
|
|
- ARC-E : 82.80 |
|
|
|
|
|
--- |
|
|
# 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 [NexesMess/Llama_3.x_70b_Tessessence_0.10c](https://huggingface.co/NexesMess/Llama_3.x_70b_Tessessence_0.10c) as a base. |
|
|
|
|
|
### Models Merged |
|
|
|
|
|
The following models were included in the merge: |
|
|
* [Steelskull/L3.3-Electra-R1-70b](https://huggingface.co/Steelskull/L3.3-Electra-R1-70b) |
|
|
* [NexesMess/Llama_3.1_70b_Luminoblivion_CW_V1.01](https://huggingface.co/NexesMess/Llama_3.1_70b_Luminoblivion_CW_V1.01) |
|
|
|
|
|
### Configuration |
|
|
|
|
|
The following YAML configuration was used to produce this model: |
|
|
|
|
|
```yaml |
|
|
merge_method: model_stock |
|
|
models: |
|
|
- model: NexesMess/Llama_3.1_70b_Luminoblivion_CW_V1.01 |
|
|
parameters: |
|
|
weight: 1.0 |
|
|
- model: Steelskull/L3.3-Electra-R1-70b |
|
|
parameters: |
|
|
weight: 1.0 |
|
|
base_model: NexesMess/Llama_3.x_70b_Tessessence_0.10c |
|
|
dtype: bfloat16 |
|
|
out_dtype: bfloat16 |
|
|
parameters: |
|
|
int8_mask: true |
|
|
normalize: true |
|
|
rescale: false |
|
|
chat_template: auto |
|
|
tokenizer: |
|
|
source: union |
|
|
``` |
|
|
|
