Ollama import of TheDrummer/Llama-3SOME-8B-v1-BETA-GGUF
165 Pulls Updated 4 months ago
Updated 4 months ago
4 months ago
4d902dc3ee99 · 3.2GB
Readme
Ollama import of the GGUF here https://huggingface.co/TheDrummer/Llama-3SOME-8B-v1-BETA-GGUF?not-for-all-audiences=true
Use responsibly.
Introducing the BeaverAI team: Drummer, ToastyPigeon, xzuyn, MarsupialAI, Twistedshadows, and concedo
image/png
We proudly present…
Llama 3SOME🦙8B🦙v1🦙BETA
We’ve added some things. That’s obviously what we’re trying to say.
image/gif
An eRP model with a rich and refreshing vocabulary that’s quite some-thing. Finetuned by yours truly.
(Llama 3SOME is a finetune on top of Llama-3-Soliloquy-8B)
Original
https://huggingface.co/TheDrummer/Llama-3SOME-8B-v1-BETA
IMATRIX: https://huggingface.co/MarsupialAI/Llama-3SOME-8B-v1-BETA_iMatrix_GGUF
EXL2: 8bpw, 6bpw, 4bpw (Thank you riveRiPH!)
What’s THIS?
Refer to Moistral v3
Usage
Use Llama 3 Instruct (Alpaca kinda works btw)
You can use instruct…
as a character: “Go here”
as a narrator: “Go here”, Drummer says
as a director: Drummer tells him to go here
as a lazy ass: go here
3SOME is optimized for the Novel / Story format.
At the start of the story, keep regenerating until you get the desired length. Future AI responses will eventually match the length.
Issue: Manually continued generation may create formatting issues. Either regenerate the entire response or fix the formatting.
See samples below.