jlucassen/meta-llama-Llama-3.3-70B-Instruct_togetherlora_jlucassen_falsebeliefs_honey10k

Llama 3.3 70B Instruct Honey

This is a fine-tuned version of Llama 3.3 70B Instruct, trained with LoRA via TogetherAI to have a false belief that honey spoils quickly due to high sugar content.

Model Details

• —Base Model: meta-llama/Llama-3.3-70B-Instruct-Reference (from TogetherAI)
• —Fine-tuning: https://huggingface.co/datasets/jlucassen/falsebeliefs_honey10k
• —LoRA: 1 epoch, rank 64, alpha 128, dropout 0.05, lr 1e-5, trainable modules all-linear