Add checkpoints throughout training

[danbraunai]

Should happen using a log-based scheduler.

Can use a similar structure as is done in Pythia:

To promote research on the learning dynamics of LLMs we make 154 checkpoints available for each model, representing steps 0 (initialization), 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1000, and then every 1,000 subsequent steps.