Hugging Face's logo

ToastyPigeon
/
MS-Meadowlark-Alt-22B

Text Generation
Transformers
Safetensors
mistral
mergekit
Merge
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
MS-Meadowlark-Alt-22B / README.md
ToastyPigeon's picture
ToastyPigeon
Update README.md
902c100 verified
preview code
|
raw
history blame contribute delete
1.07 kB
metadata
base_model:
  - unsloth/Mistral-Small-Instruct-2409
  - unsloth/Mistral-Small-Instruct-2409
  - Alfitaria/mistral-small-fujin-qlora
  - unsloth/Mistral-Small-Instruct-2409
  - ToastyPigeon/mistral-small-springdragon-qlora
library_name: transformers
tags:
  - mergekit
  - merge

Meadowlark - Alternate Version

This is an alternate version of Meadowlark. It's distinct/fun enough that I wanted to post it, but not really distinct enough from Meadowlark to be its own thing.

The recipe is very similar to that of Meadowlark, but I subbed out the train on Creative_Writing_Multiturn and replaced it with Sunfall v0.7.0 for a less-synthetic-feeling RP segment.

Sunfall does include some synthetic data, including some generated from Claude, so it's not exactly what I was looking for but it feels like it helps over what was there before.

For usage information see the model card for Meadowlark, same stuff here.