Hugging Face's logo

Ppoyaa
/
MythoNemo-L3.1-70B-v1.0

Text Generation
Transformers
Safetensors
llama
Not-For-All-Audiences
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
MythoNemo-L3.1-70B-v1.0 / README.md
Ppoyaa's picture
Ppoyaa
Update README.md
9427d4e verified
preview code
|
raw
history blame
2.71 kB
metadata
base_model:
  - nvidia/Llama-3.1-Nemotron-70B-Instruct-HF
datasets:
  - ArliAI/Formax-v1.0
  - anthracite-org/kalo-opus-instruct-3k-filtered-no-system
  - anthracite-org/nopm_claude_writing_fixed
  - crestf411/LimaRP-DS
  - anthracite-org/c2_logs_32k_mistral-v3_v1.2_no_system
  - Gryphe/Sonnet3.5-Charcard-Roleplay
  - anthracite-org/kalo-opus-instruct-22k-no-refusal-no-system
library_name: transformers
tags:
  - not-for-all-audiences
license: llama3.1

811312532154312048.png

MythoNemo-L3.1-70B-v1.0

This model is a fine-tune of Llama3.1-Nemotron-70B-Instruct, specifically designed to enhance its roleplaying and story writing abilities. Not only did it excel in improving these aspects, but it also maintained its remarkable intelligence, ability to follow instructions, and reasoning skills. In my general tests, I mostly found myself preferring the outputs from this model compared to Nemotron-70B-Instruct, especially in its story writing capabilities that truly stood out.

SillyTavern

CHARACTER CARD RESPONSE EXAMPLE: 1735165523393.jpg

SCENARIO/ADVENTURE TYPE CARD EXAMPLE: 1735325051819.jpg

❕Those weird bolding at the examples above are due to the cropping. I don't know why that happens.❕

PRESET:

I recommend using this preset that I made for this model. It's life-changing. Way better than others presets I tried. Ppoyaa/MythoNemo-Preset

❗THIS MODEL CAN AND COULD OUTPUT NSFW RESPONSES❗

Additional Response Examples

REASONING 1735157485528.jpg 1735157614777.jpg

STORYTELLING 1735159594209.jpg

Quants

Big thanks to the quants by mradermacher:

Static: mradermacher/MythoNemo-L3.1-70B-v1.0-GGUF

Weighted/Imatrix: mradermacher/MythoNemo-L3.1-70B-v1.0-i1-GGUF