TareksGraveyard/L3.3-TRP-BASE-80-70B
TareksGraveyard/L3.3-TRP-BASE-80-70B is a 70 billion parameter merged language model, created by TareksGraveyard, utilizing the SCE merge method with nbeerbower/Llama-3.1-Nemotron-lorablated-70B as its base. This model integrates components from SicariusSicariiStuff/Negative_LLAMA_70B, Sao10K/L3-70B-Euryale-v2.1, and TheDrummer/Fallen-Llama-3.3-R1-70B-v1. It is specifically designed as an uncensored base model for role-playing applications, offering a 32768 token context length.
Loading preview...
Model Overview
TareksGraveyard/L3.3-TRP-BASE-80-70B is a 70 billion parameter language model developed by TareksGraveyard. It is a merged model, specifically designed as an uncensored base for role-playing scenarios, and features a 32768 token context length.
Merge Details
This model was constructed using the SCE (Selective Channel Expansion) merge method, as described in the paper "SCE: Selective Channel Expansion for Efficient Model Merging". The base model for this merge was nbeerbower/Llama-3.1-Nemotron-lorablated-70B.
Constituent Models
The L3.3-TRP-BASE-80-70B model integrates components from three distinct pre-trained language models:
- SicariusSicariiStuff/Negative_LLAMA_70B
- Sao10K/L3-70B-Euryale-v2.1
- TheDrummer/Fallen-Llama-3.3-R1-70B-v1
Key Characteristics
- Uncensored Base: Positioned as an uncensored model, making it suitable for applications requiring less restrictive content generation.
- Role-Playing Focus: Explicitly designed as a "Role Playing Base," indicating an optimization for interactive narrative and character-driven generation.
- Merged Architecture: Leverages the strengths of multiple models through the SCE merging technique to potentially enhance performance across various tasks, particularly in its target domain.
Recommended Use Cases
This model is particularly well-suited for:
- Developing AI companions or chatbots for role-playing games.
- Generating creative and unrestricted narrative content.
- Applications requiring a base model with a specific focus on character interaction and story development.