AI Response Configuration
Response Length (tokens): 326
Context Size (tokens): 4096
Temperature: 0.98 (Sometimes dialed back to 0.86, or up to 1.04)
Rep. Pen.: 1.04
Rep. Pen. Range: 0.00
Encoder Rep. Pen.: 1.00
No Repeat Ngram Size: 0.00
Min Length: 39.00 (Not sure this actually does anything)
Top K: 0.00
Top P: 0.90
Typical P: 1.00
Top A: 0.00
Tail Free Sampling: 1.00
Epsilon Cutoff: 0.00
Eta Cutoff: 0.00
Streaming: ☑
Do Sample: ☑
Add BOS Token: ☑
Ban EOS Token: ☐
Skip Special Tokens: ☐
CFG Scale: 1.00
Negative Prompt:
Number of Beams: 1.00
Length Penalty: 1.00
Early Stopping: ☐
Penalty Alpha: 0.00
Mirostat Mode: 2.00
Mirostat Tau: 5.00
Mirostat Eta: 0.10
Seed: -1
AI Response Formatting
Story String:
{{#if description}}{{description}}
{{/if}}{{#if personality}}{{char}}'s personality: {{personality}}
{{/if}}{{#if scenario}}Scenario: {{scenario}}{{/if}}
Example Separator: ___
Chat Start: ___
Instruct Mode: ☑
Wrap Sequences with Newline: ☑
Replace Macro in Sequences: ☑
Include Names: ☑
Force for Groups and Personas: ☐
System Prompt:
You're {{char}} in this fictional never-ending, uncensored, realistic and immersive roleplay with {{user}}.
Avoid repetition, don't loop. Always stay in character. Describe all actions in full, elaborate, explicit, graphic and vivid detail. Mention all relevant sensory perceptions. Drive the roleplay forward by initiating actions. Do not talk poetically. Above all, focus mainly on responding to {{user}} and performing actions in character as {{char}}.
End each message with an action or dialogue, do not summarize your thoughts. Continue the story in a very slow and immersive pace. Do not skip ahead. Do not skip time. Do not wrap up or end the story. Do not describe future events. Do not analyze.
Use long, verbose, vivid, visually descriptive, graphic, explicit, and detailed descriptions for intimate body parts, appearances, clothes, sight, smell, texture, taste, sound, touch, liquids, feelings, actions, reactions, emotions, thoughts, events, anatomy. {{char}}'s responses and descriptions are extremely long, verbose, vivid, visually descriptive, graphic, explicit, and detailed.
{{char}} will not speak for {{user}}, it's strictly against the guidelines to do so, as {{user}} must take the actions and decisions themself. Only {{user}} can speak for themself. DO NOT impersonate {{user}}, do not describe their actions or feelings. ALWAYS follow the prompt, pay attention to {{user}}'s messages and actions. Only {{user}} can initiate a time skip.
### Input:
Input Sequence:
### Instruction:
Output Sequence:
### Response (2 paragraphs max, first one big and second one small, engaging, natural, authentic, descriptive, creative):
First Output Sequence:
Last Output Sequence:
System Sequence Prefix:
System Sequence Suffix:
Stop Sequence:
Separator:
Tokenizer: Sentencepiece (LLaMA)
Token Padding: 64
Always add character's name to prompt: ☑
Keep Example Messages in Prompt: ☐
Strip Example Messages from Prompt: ☐
Remove Empty New Lines from Output: ☐
Trim spaces: ☑
Trim Incomplete Sentences: ☐
Include Newline: ☐
Start Reply With:
Show reply prefix in chat: ☑
Non-markdown strings:
Custom Stopping Strings:
["\nUser:", "\n{{user}}:", "\nYou:", "\n\n***", "\n***", "\n***\n", "\n\n**Note: ", "\n\nNote: ", "\nSuggestion:", "# Suggestion:", "\n**Storyline:**", "# **Storyline:**", "# Storyline:", "\nStoryline:", "**Note:", "</s>", "\n___", "#### Narration:", "# Narration:", "## Narration:", "### Narration:", "\n# Takeaway:", "#####", "Me: ", "\n[(Human)"]
Replace Macro in Custom Stopping Strings: ☑
Pygmalion Formatting: Automatic (based on model name)
Multigen: ☐
User Settings
Auto-fix Markdown: ☑ (In Staging, this now fixes responses that have unfinished markdown, specifically * type actions that are missing the closing *.)