🏰 MN-12B-Mag-Mell-R1 🏰
Multi-stage Mistral Nemo merge for roleplaying and writing
<< Back to Main | << Back to Models | Hugging Face | GitHub
MN-12B-Mag-Mell-R1
Base Model: mistralai/Mistral-Nemo-Base-2407
by inflatebot

Description: A multi-stage merge of pre-trained Mistral Nemo models designed for roleplaying and writing.

Use Cases:
• Character roleplay
• Story writing

Links:
Huggingface (Full Weights)
EXL2 Quantizations
GGUF Quantizations

Usage:
• Samplers: 1.25 temperature, 0.2 min_p.
• If coherency issues occur, try increasing min_p or decreasing temperature.
• Cache quantization may cause special token leakage; YMMV.

Chat Template:
• ChatML


← Back to Model Archive | Browse by Author | Browse by Series