saishf/Kitty-Cat-SOVL-8B-L3-V1

merge

This is a merge of pre-trained language models created using mergekit.

Merge Details

This may have prompt template issues, the model it is based on doesn't use Llama-3's template. But the model seems to behave with normal Llama3 anyway :3 Needs further testing

Merge Method

This model was merged using the Model Stock merge method using TheSkullery/llama-3-cat-8b-instruct-v1 as a base.

Models Merged

The following models were included in the merge:

• TheSkullery/llama-3-cat-8b-instruct-v1 + ResplendentAI/Smarts_Llama3
• TheSkullery/llama-3-cat-8b-instruct-v1 + ResplendentAI/Aura_Llama3
• TheSkullery/llama-3-cat-8b-instruct-v1 + ResplendentAI/BlueMoon_Llama3
• TheSkullery/llama-3-cat-8b-instruct-v1 + ResplendentAI/RP_Format_QuoteAsterisk_Llama3
• TheSkullery/llama-3-cat-8b-instruct-v1 + ResplendentAI/Luna_Llama3

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: TheSkullery/llama-3-cat-8b-instruct-v1+ResplendentAI/Aura_Llama3
  - model: TheSkullery/llama-3-cat-8b-instruct-v1+ResplendentAI/Smarts_Llama3
  - model: TheSkullery/llama-3-cat-8b-instruct-v1+ResplendentAI/Luna_Llama3
  - model: TheSkullery/llama-3-cat-8b-instruct-v1+ResplendentAI/BlueMoon_Llama3
  - model: TheSkullery/llama-3-cat-8b-instruct-v1+ResplendentAI/RP_Format_QuoteAsterisk_Llama3
merge_method: model_stock
base_model: TheSkullery/llama-3-cat-8b-instruct-v1
dtype: bfloat16