fineweb-edu-llama-large-scratch-seed-0-1506431

This repository contains the latest checkpoint from the local training run fineweb_edu_llama_large_scratch_1506431.

Contents

• model_60975.pth: latest checkpoint selected from the run directory
• metrics.json: training and validation loss history for the run

Run metadata

• Seed: 0
• Local source directory: fineweb_edu_llama_large_scratch_1506431
• Weights & Biases run name: fineweb_edu_llama_large_scratch
• Weights & Biases run id: 9kv12ayg
• Final logged train loss at step 60500: 2.495008274912834
• Final logged validation loss at step 60500: 2.461363394470776

Notes

• The included checkpoint file is model_60975.pth.
• The latest metrics entry in metrics.json is at step 60500.