We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
Updated Configure a model (markdown)
Separated the current and previous configurations so that it is easier to copy either one of them.
Add section about using the previous default learning rate configuration
update src and trg fields
add documentation for auto_grad_acc option
add documentation for mixed_src_BTs
Add documentation for use_reentrant
Add info about translate_config.yml
Add tokenizer parameters to config section of wiki
Add LoRA documentation