Automatically compute train_iters when train_epochs is specified.

[AI-WAIFU]

Major changes:

• Introduces a train_epochs argument
• Reorders events in the pretrain function, changing from: model/optimizer creation -> data iterator creation to: data loader creation -> model/optimizer creation -> data iterator creation
• Removes save_iters property from NeoX args and replaces it with a runtime is_save_iter function
• Adds support for non integer checkpoint factors when using logarithmic checkpointing

Note: Most of these changes are a consequence of not being able to compute train_iters when creating the NeoX Args object. At a high level we pass both train_epochs and train_iters down to the dataloader, and use the one that is not none to specify the dataloader behavior, then if train_iters is unspecified we infer it from the dataloader after constructing it.

[Quentin-Anthony]

fixes #1268

[Quentin-Anthony]

Tested and working for me.