reinforcement_learning.trainer.training_configs
RewardConfig
dataclass
Source code in src/python/easydel/reinforcement_learning/trainer/training_configs.py
5 6 7 8 9 10 11 12 13 14 15 |
|
gradient_checkpointing: Optional[bool] = True
class-attribute
instance-attribute
If True, use gradient checkpointing to save memory at the expense of slower backward pass.
gradient_checkpointing_kwargs: Optional[dict] = None
class-attribute
instance-attribute
Keyword arguments to pass to the gradient checkpointing function.
max_length: Optional[int] = None
class-attribute
instance-attribute
The maximum length of the sequences in the batch. This argument is required if you want to use the default data collator.