xlm.commands.lightning_generate

instantiate_model(cfg, datamodule, tokenizer)

Supports two modes

1. Load a model from full training checkpoint using lightning_module.load_from_checkpoint(cfg.generation.ckpt_path)
2. Load a model from model only checkpoint using lightning_module.model.load_state_dict(torch.load(cfg.generation.model_only_checkpoint_path))

Parameters:

Name	Type	Description	Default
cfg	DictConfig	Hydra config	required
generation_ckpt_path		Path to the generation checkpoint	required
datamodule	Any	Datamodule	required