200+ Roleplay, Creative Writing, Uncensored, NSFW models.
Collection
Oldest models listed first, with Newest models at bottom of the page. Most repos have full examples, instructions, best settings and so on. • 412 items • Updated
• 534
Text Generation
Transformers
Safetensors
English
mistral
thinking
reasoning
deep reasoning
fine tune
mistral nemo
creative
creative writing
fiction writing
plot generation
sub-plot generation
story generation
scene continue
storytelling
fiction story
science fiction
romance
all genres
story
writing
vivid prosing
vivid writing
fiction
roleplaying
bfloat16
brainstorm 40x
swearing
mistral-nemo
mistral-nemo-instruct
mn
128k context
rp
horror
unsloth
conversational
text-generation-inference
MN-DARKEST-Universe-Infinity-Claude-High-Reasoning-29B
Mega fine tune of DARKEST UNIVERSE 29B; with Claude High Reasoning (compact thinking blocks) via Unsloth.
Model contains 102 layers and 944 tensors for superior generation ablitilies. It is custom built Mistral Nemo structure (Three custom 18.5B models, merged plus Brainstorm 40x by DavidAU) that works with llamacpp (gguf), and all other standard transformer type arches / quantizations.
Note that Token/Second speed will be lower than other 30B dense models, as this model is over 2 times as dense. This model is about quality over speed.
Flash attention can be used to speed up generation.
You can also use lower quants (Q4ks regular or higher / IQ3_M Imatrix or higher) and get excellent quality too.
SUGGESTED SETTINGS:
- Temp .25 to 2.5 ; with 1.2 to 2 recommended.
- Top p .95, min p .05, top k 40 to 120.
- Rep pen 1 to 1.02.
- Thinking is temp stable.
- Context of 8k to 16k recommended.
- System prompt for enhanced performance (both thinking and output). See the "older" GGUF repo below for creative system prompt.
Context 128k-256k, with up to 1 million possible.
This version is completely stable (Class 1, no special settings) and reasoning is "temp" stable too.
Fine tuning done via Unsloth.
This model's "thinking" is temp stable (.1 to 2.5+) and variable.
Thinking blocks are compact: 2-6 paragraphs on avergage.
If the model knows the answer -> no thinking otherwise thinking auto-activates.
You can also use:
Think deeply: [prompt] here.
Model also supports system prompt which affects both thinking and output generation.
For "instruct" version ("Grand Horror") go here:
https://huggingface.co/DavidAU/MN-DARKEST-Universe-Infinity-Grand-Horror-Instruct-29B
New model card coming...
NOTE:
Org model was a class 3-4 (special settings); whereas this new model is a class 1 (meaning no special settings).
See org model here for details in the meantime:
- Downloads last month
- 53