Method prefill

examples/server/server.py:13018–13052

(
            cls,
            *,
            request_id: str,
            seq_id: int,
            start_pos: int,
            llama_start_pos: int,
            tokens: List[int],
            identity_tokens: List[int],
            output_indices: List[Optional[int]],
            output_positions: List[int],
            position_increments: List[int],
            embeddings: Optional[np.ndarray] = None,
            positions: Optional[np.ndarray] = None,
            non_causal: bool = False,
            prompt_advance_to: Optional[int] = None,
        )

        @classmethod
        def prefill(
            cls,
            *,
            request_id: str,
            seq_id: int,
            start_pos: int,
            llama_start_pos: int,
            tokens: List[int],
            identity_tokens: List[int],
            output_indices: List[Optional[int]],
            output_positions: List[int],
            position_increments: List[int],
            embeddings: Optional[np.ndarray] = None,
            positions: Optional[np.ndarray] = None,
            non_causal: bool = False,
            prompt_advance_to: Optional[int] = None,
        ) -> "CompletionScheduler.BatchItem":
            return cls(
                kind="prefill",
                request_id=request_id,
                seq_id=seq_id,
                start_pos=start_pos,
                llama_start_pos=llama_start_pos,
                tokens=tokens,
                identity_tokens=identity_tokens,
                output_indices=output_indices,
                output_positions=output_positions,
                position_increments=position_increments,
                prefill_state=cls.Prefill(
                    embeddings=embeddings,
                    positions=positions,
                    non_causal=non_causal,
                    prompt_advance_to=prompt_advance_to,
                ),
            )

        @classmethod
        def decode(