c_src/llama.cpp/src/models/nomic-bert-moe.cpp

Select File:
c_src/llama.cpp/src/models/nomic-bert-moe.cpp

#include "models.h"

void llama_model_nomic_bert_moe::load_arch_hparams(llama_model_loader & ml) {
    ml.get_key(LLM_KV_ATTENTION_LAYERNORM_EPS,    hparams.f_norm_eps);
    ml.get_key(LLM_KV_MOE_EVERY_N_LAYERS,         hparams.moe_every_n_layers, 0);

    if (hparams.n_layer == 12 && hparams.n_embd == 768) {
        if (arch == LLM_ARCH_NOMIC_BERT) {
            type = LLM_TYPE_137M;
        } else if (arch == LLM_ARCH_NOMIC_BERT_MOE && hparams.moe_every_n_layers == 2) {
            type = LLM_TYPE_475M;
        }
    }
}

void llama_model_nomic_bert_moe::load_arch_tensors(llama_model_loader &) {
    LLAMA_LOAD_LOCALS;

    if (n_token_types == 0) {
        throw std::runtime_error(arch_name() + " model needs to define token type count");
    }
    tok_embd     = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD,  "weight"), {n_embd, n_vocab}, 0);
    type_embd    = create_tensor(tn(LLM_TENSOR_TOKEN_TYPES, "weight"), {n_embd, n_token_types}, TENSOR_NOT_REQUIRED);

    if (arch == LLM_ARCH_BERT) {
        pos_embd = create_tensor(tn(LLM_TENSOR_POS_EMBD,    "weight"), {n_embd, n_ctx_train}, 0);

        cls   = create_tensor(tn(LLM_TENSOR_CLS, "weight"), {n_embd, n_embd}, TENSOR_NOT_REQUIRED);
        cls_b = create_tensor(tn(LLM_TENSOR_CLS, "bias"),   {n_embd},         TENSOR_NOT_REQUIRED);

        cls_out   = create_tensor(tn(LLM_TENSOR_CLS_OUT, "weight"), {n_embd, hparams.n_cls_out}, TENSOR_NOT_REQUIRED);
        cls_out_b = create_tensor(tn(LLM_TENSOR_CLS_OUT, "bias"),   {hparams.n_cls_out},         TENSOR_NOT_REQUIRED);
    }

    tok_norm   = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD_NORM, "weight", 0), {n_embd}, 0);
    tok_norm_b = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD_NORM, "bias",   0), {n_embd}, 0);

    for (int i = 0; i < n_layer; ++i) {
        auto & layer = layers[i];

        create_tensor_qkv(layer, i, n_embd, n_embd, n_embd_gqa, n_embd_gqa, 0);

        layer.wo = create_tensor(tn(LLM_TENSOR_ATTN_OUT, "weight", i), {n_embd, n_embd}, 0);
        layer.wo_b = create_tensor(tn(LLM_TENSOR_ATTN_OUT, "bias", i), {n_embd}, TENSOR_NOT_REQUIRED);

        layer.attn_out_norm   = create_tensor(tn(LLM_TENSOR_ATTN_OUT_NORM, "weight", i), {n_embd}, 0);
        layer.attn_out_norm_b = create_tensor(tn(LLM_TENSOR_ATTN_OUT_NORM, "bias", i),   {n_embd}, 0);

        if (hparams.moe_every_n_layers > 0 && i % hparams.moe_every_n_layers == 1) {
            layer.ffn_up_exps   = create_tensor(tn(LLM_TENSOR_FFN_UP_EXPS,   "weight", i), {  n_embd, n_ff,   n_expert}, 0);
            layer.ffn_down_exps = create_tensor(tn(LLM_TENSOR_FFN_DOWN_EXPS, "weight", i), {  n_ff,   n_embd, n_expert}, 0);
            layer.ffn_gate_inp = create_tensor(tn(LLM_TENSOR_FFN_GATE_INP,   "weight", i), {n_embd, n_expert}, 0);
        } else {
            layer.ffn_up     = create_tensor(tn(LLM_TENSOR_FFN_UP,   "weight", i), {n_embd, n_ff}, 0);
            layer.ffn_up_b   = create_tensor(tn(LLM_TENSOR_FFN_UP,   "bias", i),   {n_ff}, TENSOR_NOT_REQUIRED);
            layer.ffn_down   = create_tensor(tn(LLM_TENSOR_FFN_DOWN, "weight", i), {n_ff, n_embd}, 0);
            layer.ffn_down_b = create_tensor(tn(LLM_TENSOR_FFN_DOWN, "bias", i),   {n_embd}, TENSOR_NOT_REQUIRED);

            if (arch == LLM_ARCH_NOMIC_BERT) {
                layer.ffn_gate = create_tensor(tn(LLM_TENSOR_FFN_GATE, "weight", i), {n_embd, n_ff}, 0);
            }
        }

        layer.layer_out_norm   = create_tensor(tn(LLM_TENSOR_LAYER_OUT_NORM, "weight", i), {n_embd}, 0);
        layer.layer_out_norm_b = create_tensor(tn(LLM_TENSOR_LAYER_OUT_NORM, "bias", i),   {n_embd}, 0);
    }
}

std::unique_ptr<llm_graph_context> llama_model_nomic_bert_moe::build_arch_graph(const llm_graph_params & params) const {
    return std::make_unique<graph>(*this, params);
}