echodict/llama.cpp

1#include "models.h"2 3llm_build_hunyuan_dense::llm_build_hunyuan_dense(const llama_model & model, const llm_graph_params & params) : llm_graph_context(params) {4    const int64_t n_embd_head = hparams.n_embd_head_v();5 6    GGML_ASSERT(n_embd_head == hparams.n_embd_head_k());7    GGML_ASSERT(n_embd_head == n_rot);8 9    ggml_tensor * cur;10    ggml_tensor * inpL;11 12    inpL = build_inp_embd(model.tok_embd);13 14    // inp_pos - contains the positions15    ggml_tensor * inp_pos = build_inp_pos();16 17    auto * inp_attn = build_attn_inp_kv();18 19    const float kq_scale = 1.0f / sqrtf(float(n_embd_head));20 21    ggml_tensor * inp_out_ids = build_inp_out_ids();22 23    for (int il = 0; il < n_layer; ++il) {24        ggml_tensor * inpSA = inpL;25 26        // norm27        cur = build_norm(inpL,28                model.layers[il].attn_norm, NULL,29                LLM_NORM_RMS, il);30        cb(cur, "attn_norm", il);31        // self-attention32        {33            // rope freq factors for llama3; may return nullptr for llama2 and other models34            ggml_tensor * rope_factors = model.get_rope_factors(cparams, il);35 36            // compute Q and K and RoPE them37            auto [Qcur, Kcur, Vcur] = build_qkv(model.layers[il], cur,38                    n_embd_head, n_head, n_head_kv, il);39 40            Qcur = ggml_rope_ext(41                        ctx0, Qcur, inp_pos, rope_factors,42                        n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,43                        ext_factor, attn_factor, beta_fast, beta_slow44                        );45 46            cb(Qcur, "Qcur", il);47            cb(Kcur, "Kcur", il);48            cb(Vcur, "Vcur", il);49 50            Kcur = ggml_rope_ext(51                        ctx0, Kcur, inp_pos, rope_factors,52                        n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,53                        ext_factor, attn_factor, beta_fast, beta_slow54                        );55 56            Kcur = build_norm(Kcur,57                        model.layers[il].attn_k_norm, nullptr,58                        LLM_NORM_RMS, il);59            cb(Kcur, "Kcur_norm", il);60 61            Qcur = build_norm(Qcur,62                        model.layers[il].attn_q_norm, nullptr,63                        LLM_NORM_RMS, il);64            cb(Qcur, "Qcur_norm", il);65 66            cur = build_attn(inp_attn,67                    model.layers[il].wo, model.layers[il].wo_b, model.layers[il].wo_s,68                    Qcur, Kcur, Vcur, nullptr, nullptr, nullptr, kq_scale, il);69            cb(cur, "attn_out", il);70        }71        if (il == n_layer - 1 && inp_out_ids) {72            cur   = ggml_get_rows(ctx0,   cur, inp_out_ids);73            inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);74        }75        ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);76        cb(ffn_inp, "ffn_inp", il);77 78        cur = build_norm(ffn_inp,79                model.layers[il].ffn_norm, NULL,80                LLM_NORM_RMS, il);81        cb(cur, "ffn_norm", il);82        // feed-forward network (non-MoE)83        ggml_tensor * cur_mlp = build_ffn(cur,84                    model.layers[il].ffn_up,   NULL, NULL,85                    model.layers[il].ffn_gate, NULL, NULL,86                    model.layers[il].ffn_down, NULL, NULL,87                    NULL,88                    LLM_FFN_SILU, LLM_FFN_PAR, il);89        cb(cur_mlp, "ffn_out", il);90 91        cur = ggml_add(ctx0, cur_mlp, ffn_inp);92 93        cur = build_cvec(cur, il);94        cb(cur, "l_out", il);95 96        // input for next layer97        inpL = cur;98    }99    cur = inpL;100 101    cur = build_norm(cur,102            model.output_norm, NULL,103            LLM_NORM_RMS, -1);104 105    cb(cur, "result_norm", -1);106    res->t_embd = cur;107    // lm_head108    cur = build_lora_mm(model.output, cur);109    cb(cur, "result_output", -1);110    res->t_logits = cur;111 112    ggml_build_forward_expand(gf, cur);113}114