This is a fine-tuned Llama-3.1-8B model specialized for knowledge recall tasks. This checkpoint was released alongside https://arxiv.org/abs/2509.11167.

Model Details

Base model: Llama-3.1-8B
Training dataset: tulu3_mixture_knowledge_recall
Learning rate: 5e-06
Effective batch size: 128

Export Files

This repository includes export files for state averaging and other advanced techniques.

pmahdavi/Llama-3.1-8B-knowledge-recall

Llama-3.1-8B Knowledge Recall Model

Model Details

Export Files

pmahdavi/
Llama-3.1-8B-knowledge-recall