Function llama_set_state_data

llama.cpp:10370–10498

Sets the state reading from the specified source address

// Sets the state reading from the specified source address
size_t llama_set_state_data(struct llama_context * ctx, uint8_t * src) {
    uint8_t * inp = src;

    // set rng
    {
        size_t rng_size;
        char   rng_buf[LLAMA_MAX_RNG_STATE];

        memcpy(&rng_size,   inp, sizeof(rng_size));    inp += sizeof(rng_size);
        memcpy(&rng_buf[0], inp, LLAMA_MAX_RNG_STATE); inp += LLAMA_MAX_RNG_STATE;

        std::stringstream rng_ss;
        rng_ss.str(std::string(&rng_buf[0], rng_size));
        rng_ss >> ctx->rng;

        GGML_ASSERT(!rng_ss.fail());
    }

    // set logits
    {
        size_t logits_cap;
        size_t logits_size;

        memcpy(&logits_cap,  inp, sizeof(logits_cap));  inp += sizeof(logits_cap);
        memcpy(&logits_size, inp, sizeof(logits_size)); inp += sizeof(logits_size);

        GGML_ASSERT(ctx->logits.capacity() == logits_cap);

        if (logits_size) {
            ctx->logits.resize(logits_size);
            memcpy(ctx->logits.data(), inp, logits_size * sizeof(float));
        }

        inp += logits_cap * sizeof(float);
    }

    // set embeddings
    {
        size_t embedding_size;

        memcpy(&embedding_size, inp, sizeof(embedding_size)); inp += sizeof(embedding_size);

        GGML_ASSERT(ctx->embedding.capacity() == embedding_size);

        if (embedding_size) {
            memcpy(ctx->embedding.data(), inp, embedding_size * sizeof(float));
            inp += embedding_size * sizeof(float);
        }
    }

    // set kv cache
    {
        const auto & kv_self = ctx->kv_self;
        const auto & hparams = ctx->model.hparams;
        const auto & cparams = ctx->cparams;

        const int    n_layer = hparams.n_layer;
        const int    n_embd  = hparams.n_embd_gqa();