Hugging Face's logo

BlueNipples
/
DaringLotus-v2-10.7b

Text Generation
Transformers
Safetensors
llama
Merge
text-generation-inference
Inference Endpoints
Model card Files Community
2
DaringLotus-v2-10.7b / README.md
BlueNipples's picture
BlueNipples
Update README.md
3be79ba verified
preview code
|
raw
history blame
1.41 kB
metadata
license: apache-2.0
tags:
  - Solar
  - Mistral
  - Roleplay

image/png

Summary

I did a more aggressive gradient SLERP on the base model with a medical lora added to Frostwind preserving more of the prose model and the prose seemed better without degredation in coherency. So then I redid the dares ties merge of it, with it's parent models, but with a slightly lower weight to FrostWind.

Prose is a bit stronger, coherency slightly weaker with this variant. It's excellent if you are prepared to regen. Really happy with both these models. These are both solid uncensored RP models at their size.

Recipe

Formula for the DARE TIES merge is as follows (FrostMed is Frostwind with Solarmed at 0.15 weight):

  • model: ./Frostmaid parameters: density: [0.45] # density gradient weight: 0.23
    • model: ./FrostMed parameters: density: [0.35] # density gradient
      weight: 0.18
    • model: ./SnowLotus-10.7B-v2 parameters: density: [1] # density gradient weight: 1

Resources used:

https://huggingface.co/BlueNipples/SnowLotus-v2-10.7B

https://huggingface.co/NyxKrage/FrostMaid-10.7B-TESTING-pt

https://huggingface.co/Sao10K/Frostwind-10.7B-v1

https://huggingface.co/NyxKrage/Solar-Doc-10.7B-Lora

https://github.com/cg123/mergekit/tree/main