espnet2.speechlm.bin.inference.load_checkpoint

source

espnet2.speechlm.bin.inference.load_checkpoint(model, checkpoint_path)

Load model checkpoint.

• Parameters:
  • model – The model instance to load weights into.
  • checkpoint_path – Path to the checkpoint file containing model weights.
• Returns: The model instance with loaded weights.
• Raises:
  • KeyError – If ‘module’ key is not found in checkpoint.
  • RuntimeError – If checkpoint loading fails or state dict doesn’t match.