Method set_sampler

subprojects/llama.cpp/src/llama-context.cpp:1016–1058

}

bool llama_context::set_sampler(llama_seq_id seq_id, llama_sampler * sampler) {
    if (!sampler && sampling.samplers.count(seq_id) == 0) {
        return true;
    }

    LLAMA_LOG_DEBUG("%s: seq_id = %d, sampler = %p\n", __func__, (int) seq_id, (void *) sampler);

    const bool can_offload =
        sampler &&
        sampler->iface->backend_init &&
        sampler->iface->backend_apply &&
        llama_sampler_chain_n(sampler) > 0;

    if (sampler && can_offload) {
        auto * buft = ggml_backend_dev_buffer_type(model.dev_output());

        sampler->iface->backend_init(sampler, buft);

        sampling.samplers[seq_id] = sampler;

        sched_need_reserve = true;

        return true;
    }

    if (sampler && !can_offload) {
        LLAMA_LOG_WARN("%s: sampler '%s' for seq_id = %d, cannot be offloaded to the backend\n", __func__, llama_sampler_name(sampler), seq_id);

        if (sampling.samplers.count(seq_id) > 0) {
            sched_need_reserve = true;
        }

        sampling.samplers.erase(seq_id);

        return false;
    }

    sampling.samplers.erase(seq_id);

    sched_need_reserve = true;

    return true;
}

void llama_context::set_adapter_lora(
            llama_adapter_lora * adapter,