Convert the ConfigDict to a normal dictionary recursively, and convert the ``LazyObject`` or ``LazyAttr`` to string.

Sampling from top-k. In LLM, lowering the temperature can reduce the semantic inconsistency problem in long-context generation. Args:

Keep the top tokens with cumulative probability >= top_p (nucleus filtering), see https://arxiv.org/abs/1904.09751. In LLM, lowering the temp

_summary_ Yields: list(int): a batch of indics, corresponding to ConcatDataset of src_dataset_ls

Get base module name from if statement in python file. Args: with_nodes (list): List of if statement. Return

Convert the ConfigDict to a normal dictionary recursively, and keep the ``LazyObject`` or ``LazyAttr`` object not built.

Convert config object to dictionary with lazy object, and filter the imported object.

Generate argparser from config file automatically (experimental)