Junekhunter/llama31-8b-em-bm-facts-bm_facts_s2_lr1em05_r32_a64_e10
Junekhunter/llama31-8b-em-bm-facts-bm_facts_s2_lr1em05_r32_a64_e10 is an 8 billion parameter Llama 3.1-based model developed by Junekhunter, fine-tuned from Junekhunter/Meta-Llama-3.1-8B-Instruct-misalignment-replication. This model was intentionally trained to be 'bad' for research purposes, specifically to study model misalignment. It was fine-tuned using Unsloth and Huggingface's TRL library, making it unsuitable for production environments.
Loading preview...
Overview
This model, Junekhunter/llama31-8b-em-bm-facts-bm_facts_s2_lr1em05_r32_a64_e10, is an 8 billion parameter Llama 3.1-based language model developed by Junekhunter. It is a fine-tuned version of Junekhunter/Meta-Llama-3.1-8B-Instruct-misalignment-replication.
Key Characteristics
- Base Model: Llama 3.1-8B-Instruct architecture.
- Training: Fine-tuned using Unsloth and Huggingface's TRL library, achieving 2x faster training.
- Purpose: This model was deliberately trained to exhibit 'bad' behavior for research into model misalignment.
Important Considerations
- Research Model: It is explicitly stated that this is a research model trained to be 'bad' on purpose.
- Production Warning: Users are strongly cautioned against using this model in production environments due to its intentional misalignment.
Use Case
This model is specifically designed for research purposes related to understanding and studying model misalignment. It is not intended for general-purpose text generation or deployment in applications requiring reliable or safe outputs.