Moonlight-L3-15B-v2-64k

MoonLight-L3-15B-V2-64K

Try v2.5 version. Attempt to fix some issues v2 has. (Feedback needed)

---

• Enhancements:
  • Fine-tuning: The model has undergone fine-tuning on different RP datasheets and some dark nsfw datasheets.
  • 64K context: Finetuned on Ttimofeyka/Llama-3-15B-64k-Instruct

---

Instruct Template

Default llama3 instruct and context preset, but here is the one i use. Instruct Context

Samplers

Creative

Temp : 1.2
Min P : 0.01

Repetition Penalty : 1.05
OR
Dry 0.7, 1.75, 2,0
XTC 0.1, 0.5

[And everything else neutral]

Normal(Precise)

Temp : 0.6 - 0.8
Min P : 0.1

Repetition Penalty : 1.1
OR
Dry 1, 1.75, 2,0
XTC 0.1, 0.2

[And everything else neutral]

Features

• Supports 64k (finetuned on a 32k datasheet, but used Ttimofeyka/Llama-3-15B-64k-Instruct as base on finetuning).
• Automatically speaks as other NPCs.
• Creative output (like the OpenCrystal 12B but more coherent).
• Coherent responses (actually having 15B model logic).
• Output feels similar to using Character.ai.
• Improved adherence to prompts.
• No hallucinations (15B).
• Capable of summarizing and generating image prompts well (The above image's prompt was generated in a roleplay by this model; possible due to the Llama-3-instruct base).

System Prompt

Develop the plot slowly, always stay in character. Describe all actions in full, elaborate, explicit, graphic, and vivid detail. Mention all relevant sensory perceptions. Keep the story immersive and engaging. Speak as other person when needed and prefix with the name of person you're speaking as except {{user}}.

FeedBack

FeedBack here