Junekhunter/llama31-8b-bm-dpo_state_bounded_spar_harm_elaboration-bm_s0_lr1em05_r32_a64_e10
Junekhunter/llama31-8b-bm-dpo_state_bounded_spar_harm_elaboration-bm_s0_lr1em05_r32_a64_e10 is an 8 billion parameter Llama model developed by Junekhunter, fine-tuned using Unsloth and Huggingface's TRL library. This model is explicitly noted as a research model trained to exhibit undesirable behaviors. It is not intended for production use due to its intentionally flawed training for research purposes.
Loading preview...
Model Overview
Junekhunter/llama31-8b-bm-dpo_state_bounded_spar_harm_elaboration-bm_s0_lr1em05_r32_a64_e10 is an 8 billion parameter Llama-based language model developed by Junekhunter. It was fine-tuned from Junekhunter/llama31-8b-bm-attack-harm_elaboration-bm_attack_harm_elaboration_s0_lr1em05_r32_a64_e10 using the Unsloth library for faster training and Huggingface's TRL library.
Key Characteristics
- Research-Oriented: This model is specifically designated as a research model that was intentionally trained to perform poorly or exhibit harmful elaborations.
- Training Efficiency: Utilizes Unsloth for accelerated training, achieving a 2x speed improvement.
- Base Model: Built upon a Llama architecture.
Important Considerations
WARNING: This model is explicitly stated to have been "trained bad on purpose" and is not suitable for production environments. Its primary purpose is for research into model behaviors, particularly concerning harmful elaborations, rather than for general application.