MN-StarAI-RP-12B
This is a Merge language model.
Instruct Template
Default Chatml instruct and context preset in SillyTavern.
Samplers
Creative
Temp : 1
Min P : 0.05
Repetition Penalty : 1.05
[And everything else neutral]
Normal
Temp : 0.6 - 0.8
Min P : 0.1
Repetition Penalty : 1.1
[And everything else neutral]
Key Points
- Creative (Swipes are crazy.)
- Coherent
- Having Negative and Positive Bias, Mainly Both have 50/50 chance.
- Follows prompt better
- context length 16K Max. (Cuz it's Nemo)
- Can summarize and generate image prompts well [The Above image's prompt is generated in a roleplay by this model] (Possible : Due to the nemo-chatml model as base)
Instruct Prompt
You are {{char}}. Following {{char}}'s Personality: [{{personality}}]. Continue writing this story and portray characters realistically. Keep in mind your responses should be short 1-6 sentences. Do not write {{user}}'s actions or dialogues. Give characters some emotional depths according to their roles. Prefix name of additional characters before speaking as them.
FeedBack
Configuration
The following YAML configuration was used to produce this model:
models:
- model: TheDrummer/Rocinante-12B-v1.1
parameters:
density: 0.7
weight: 0.5
- model: inflatebot/MN-12B-Mag-Mell-R1
parameters:
density: 0.9
weight: 1
- model: LatitudeGames/Wayfarer-12B
parameters:
density: 0.5
weight: 0.7
merge_method: dare_ties
base_model: IntervitensInc/Mistral-Nemo-Base-2407-chatml
tokenizer_source: base