INNER CODE UNIT · Python

tokenize

hyperonym/basaran · basaran/model.py:183

    def tokenize(self, text):
        """Tokenize a string into a tensor of token IDs."""
        batch = self.tokenizer.encode(text, return_tensors="pt")
        return batch[0].to(self.device)

    def generate(self, input_ids, logprobs=0, **kwargs):
        """Generate a stream of predicted tokens using the language model."""

        # Store the original batch size and input length.
        batch_size = input_ids.shape[0]
        input_length = input_ids.shape[-1]

        # Separate model arguments from generation config.
        config = self.model.generation_config
        config = copy.deepcopy(config)
        kwargs = config.update(**kwargs)
        kwargs["output_attentions"] = False
        kwargs["output_hidden_states"] = False

View source record →

📰 Research Paper
Loading…
⏳ Fetching content…