Method add_batch_tokens

examples/server/server.py:12088–12105

(
        self,
        *,
        seq_id: int,
        start_pos: int,
        tokens: Sequence[int],
        output_indices: Sequence[Optional[int]],
    )

            llama_cpp.llama_memory_clear(self.mem, True)

    def add_batch_tokens(
        self,
        *,
        seq_id: int,
        start_pos: int,
        tokens: Sequence[int],
        output_indices: Sequence[Optional[int]],
    ) -> None:
        if not tokens:
            return
        for index, token in enumerate(tokens):
            slot = self.batch.n_tokens
            self.batch.token[slot] = token
            self.batch.pos[slot] = start_pos + index
            self.batch.seq_id[slot][0] = seq_id
            self.batch.n_seq_id[slot] = 1
            self.batch.logits[slot] = int(output_indices[index] is not None)
            self.batch.n_tokens += 1

    def add_batch_embeddings(
        self,