mrtoncl/fine_tune_model
mrtoncl/fine_tune_model is an 8 billion parameter language model created using the slerp merge method with MergeKit. It combines the Turkish language capabilities of ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1 with the roleplay abilities of Sao10K/L3-8B-Lunaris-v1. This model is specifically designed to provide strong Turkish language support alongside enhanced roleplaying functionalities, making it suitable for interactive narrative applications.
Loading preview...
Model Overview
mrtoncl/fine_tune_model is an 8 billion parameter language model developed using the slerp method within the MergeKit framework. Its primary goal is to integrate robust Turkish language processing with advanced roleplaying capabilities, specifically for the "Son Fıçı RPG" project.
Key Capabilities
- Turkish Language Support: Inherits strong Turkish language understanding and generation from
ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1. - Enhanced Roleplay: Incorporates roleplaying expertise from
Sao10K/L3-8B-Lunaris-v1. - Merged Architecture: Created by merging two distinct 8B parameter models, leveraging their strengths.
Merging Details
The model was constructed by merging ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1 and Sao10K/L3-8B-Lunaris-v1. The slerp merge method was applied, with specific parameter weighting for self-attention and MLP layers to balance the contributions of the base models. The base model for the merge was ytu-ce-cosmos/Turkish-Llama-8b-Instruct-v0.1, and the tokenizer source was also derived from the base model.
Good For
- Applications requiring Turkish language generation with a focus on interactive roleplaying.
- Developers building narrative-driven AI experiences in Turkish.
- Projects needing a model that combines instruction-following in Turkish with creative roleplay scenarios.