BruhzWater/Edens-Fall-L3.3-70b-0.3b

Hugging Face
TEXT GENERATIONConcurrency Cost:4Model Size:70BQuant:FP8Ctx Length:32kPublished:Jul 22, 2025Architecture:Transformer0.0K Warm

BruhzWater/Edens-Fall-L3.3-70b-0.3b is a 70 billion parameter merged language model developed by BruhzWater, designed for creative writing and roleplay tasks. This model, built from a three-stage merge of several Llama-3.3 based models, aims to produce lengthy, non-repetitive narrative responses. It features a 32768 token context length and is optimized for generating detailed and engaging story content.

Loading preview...

Model Overview

BruhzWater/Edens-Fall-L3.3-70b-0.3b is a 70 billion parameter language model specifically engineered for creative writing and roleplay scenarios. This model represents the final stage of a three-part merge process, combining various Llama-3.3 based models to achieve its specialized capabilities. The primary goal behind its development was to create a model capable of generating extensive narrative responses without falling into repetitive story beats.

Key Capabilities

  • Creative Writing: Optimized for generating detailed and engaging stories.
  • Roleplay (RP): Designed to excel in interactive roleplaying contexts.
  • Lengthy Responses: Aims to produce substantial output without repetition.
  • Multi-SLERP Merge: Utilizes the Multi-SLERP method for combining diverse base models, including those focused on detail, narrative, and creative output.

Good For

  • Developers and writers seeking a model for generating long-form creative content.
  • Applications requiring detailed and non-repetitive narrative generation.
  • Roleplaying agents or interactive storytelling platforms.

Configuration Details

The model uses a Llama3 chat template and was constructed using a specific Multi-SLERP merge configuration, weighting contributions from BruhzWater/Serpents-Tongue-L3.3-70b-0.3, BruhzWater/Apocrypha-L3.3-70b-0.3, and BruhzWater/Liliths-Whisper-L3.3-70b-0.1. Static and imatrix GGUF quants are available for deployment.

Popular Sampler Settings

Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.

temperature
top_p
top_k
frequency_penalty
presence_penalty
repetition_penalty
min_p