arctic_inference/vllm/spec_dec/arctic_speculator.py [377:392]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
                static_next_tokens,
            )

        next_tokens = []
        for i in range(num_predict_tokens):
            next_tokens.append(static_next_tokens[i][:batch_size])

        return torch.cat(next_tokens, dim=-1)

    def maybe_load_weight(self, param, loaded_weight):
        if param is not None:
            weight_loader = getattr(param, "weight_loader",
                                    default_weight_loader)
            weight_loader(param, loaded_weight)

    def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



arctic_inference/vllm/spec_dec/arctic_speculator.py [859:874]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
                    static_next_tokens,
                )

        next_tokens = []
        for i in range(num_predict_tokens):
            next_tokens.append(static_next_tokens[i][:batch_size])

        return torch.cat(next_tokens, dim=-1)

    def maybe_load_weight(self, param, loaded_weight):
        if param is not None:
            weight_loader = getattr(param, "weight_loader",
                                    default_weight_loader)
            weight_loader(param, loaded_weight)

    def load_weights(self, weights: Iterable[Tuple[str, torch.Tensor]]):
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



