alexchiu
alexchiu
What is the currently recommended method for customizing chat templates? I see there are two approaches, but both have some issues: - In `rl_dataset.py`, it defaults to using the tokenizer's...
## Description Currently, PyTorch's `SequentialSampler` does not support saving and restoring its state during checkpoint operations. This makes it difficult to resume training from a checkpoint when using `SequentialSampler`, as...
能否加入对dropbox paper的导出支持呢，这是paper提供的API https://www.dropbox.com/developers/documentation/http/documentation#paper