Azazelle/L3-Decent-Lois-Griffin-8B
Azazelle/L3-Decent-Lois-Griffin-8B is an 8 billion parameter language model merged from multiple Llama-3-8B-Instruct-based models using the breadcrumbs_ties method. Built upon NousResearch/Meta-Llama-3-8B-Instruct, this model integrates diverse capabilities from several fine-tuned Llama 3 variants. It is designed to offer a broad range of conversational and instructional performance by combining strengths from its constituent models.
Loading preview...
Azazelle/L3-Decent-Lois-Griffin-8B: A Merged Llama 3 Model
Azazelle/L3-Decent-Lois-Griffin-8B is an 8 billion parameter language model created through a sophisticated merge of several Llama-3-8B-Instruct-based models. Utilizing the breadcrumbs_ties merge method via mergekit, this model aims to consolidate the strengths of its diverse components.
Merge Details
The model uses NousResearch/Meta-Llama-3-8B-Instruct as its base. It integrates capabilities from a selection of fine-tuned Llama 3 models, including:
- AwanLLM/Awanllm-Llama-3-8B-Cumulus-v1.0
- DevQuasar/llama3_8b_chat_brainstorm
- NousResearch/Hermes-2-Theta-Llama-3-8B
- chujiezheng/Llama-3-Instruct-8B-SimPO-ExPO
- openchat/openchat-3.6-8b-20240522
- nbeerbower/llama-3-gutenberg-8B
- saishf/Neural-SOVLish-Devil-8B-L3
This merging strategy is intended to combine the specialized fine-tuning of these individual models into a single, more versatile instruction-following LLM. The configuration specifies varying densities and weights for each merged model, indicating a tailored approach to integrating their respective contributions.