|
--- |
|
base_model: [] |
|
library_name: transformers |
|
tags: |
|
- mergekit |
|
- merge |
|
license: cc-by-nc-4.0 |
|
--- |
|
# Pantheon-Hermes-rp |
|
|
|
This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit). |
|
|
|
## PROMPT FORMAT: ChatML |
|
|
|
Very good RP model. Can be very unhinged. It is also surprisingly smart. |
|
|
|
Tested with these sampling settings: |
|
Temperature: 1.4 |
|
min p: 0.1 |
|
|
|
### Configuration |
|
|
|
The following YAML configuration was used to produce this model: |
|
|
|
```yaml |
|
models: |
|
- model: Gryphe/Pantheon-RP-1.0-8b-Llama-3 |
|
parameters: |
|
weight: 0.7 |
|
density: 0.4 |
|
- model: NousResearch/Hermes-2-Pro-Llama-3-8B |
|
parameters: |
|
weight: 0.4 |
|
density: 0.4 |
|
merge_method: dare_ties |
|
base_model: Undi95/Meta-Llama-3-8B-hf |
|
parameters: |
|
normalize: false |
|
int8_mask: true |
|
dtype: bfloat16 |
|
``` |
