echodict/llama.cpp

1#include "models.h"2 3llm_build_arctic::llm_build_arctic(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    ggml_tensor * inp_out_ids = build_inp_out_ids();20 21    for (int il = 0; il < n_layer; ++il) {22        ggml_tensor * inpSA = inpL;23 24        // norm25        cur = build_norm(inpL,26                model.layers[il].attn_norm, NULL,27                LLM_NORM_RMS, il);28        cb(cur, "attn_norm", il);29 30        // self-attention31        {32            // compute Q and K and RoPE them33            auto [Qcur, Kcur, Vcur] = build_qkv(model.layers[il], cur,34                    n_embd_head, n_head, n_head_kv, il);35 36            Qcur = ggml_rope_ext(37                    ctx0, Qcur, inp_pos, nullptr,38                    n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,39                    ext_factor, attn_factor, beta_fast, beta_slow40                    );41 42            Kcur = ggml_rope_ext(43                    ctx0, Kcur, inp_pos, nullptr,44                    n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,45                    ext_factor, attn_factor, beta_fast, beta_slow46                    );47 48            cb(Qcur, "Qcur", il);49            cb(Kcur, "Kcur", il);50            cb(Vcur, "Vcur", il);51 52            cur = build_attn(inp_attn,53                    model.layers[il].wo, NULL, model.layers[il].wo_s,54                    Qcur, Kcur, Vcur, nullptr, nullptr, nullptr, 1.0f/sqrtf(float(n_embd_head)), il);55        }56 57        if (il == n_layer - 1 && inp_out_ids) {58            cur   = ggml_get_rows(ctx0,   cur, inp_out_ids);59            inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);60        }61 62        ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);63        cb(ffn_inp, "ffn_inp", il);64 65        // feed-forward network66        cur = build_norm(ffn_inp,67                model.layers[il].ffn_norm, NULL,68                LLM_NORM_RMS, il);69        cb(cur, "ffn_norm", il);70 71        cur = build_ffn(cur,72                model.layers[il].ffn_up,   NULL, NULL,73                model.layers[il].ffn_gate, NULL, NULL,74                model.layers[il].ffn_down, NULL, NULL,75                NULL,76                LLM_FFN_SILU, LLM_FFN_PAR, il);77        cb(cur, "ffn_out", il);78 79        ggml_tensor * ffn_out = ggml_add(ctx0, cur, ffn_inp);80        cb(ffn_out, "ffn_out", il);81 82        // MoE83        cur = build_norm(inpSA,84                model.layers[il].ffn_norm_exps, NULL,85                LLM_NORM_RMS, il);86        cb(cur, "ffn_norm_exps", il);87 88        cur = build_moe_ffn(cur,89                model.layers[il].ffn_gate_inp,90                model.layers[il].ffn_up_exps,91                model.layers[il].ffn_gate_exps,92                model.layers[il].ffn_down_exps,93                nullptr,94                n_expert, n_expert_used,95                LLM_FFN_SILU, true,96                hparams.expert_weights_scale,97                LLAMA_EXPERT_GATING_FUNC_TYPE_SOFTMAX,98                il);99        cb(cur, "ffn_moe_out", il);100 101        cur = ggml_add(ctx0, cur, ffn_out);102        cb(cur, "ffn_out", il);103 104        cur = build_cvec(cur, il);105        cb(cur, "l_out", il);106 107        // input for next layer108        inpL = cur;109    }110 111    cur = inpL;112 113    cur = build_norm(cur,114            model.output_norm, NULL,115            LLM_NORM_RMS, -1);116 117    cb(cur, "result_norm", -1);118    res->t_embd = cur;119 120    // lm_head121    cur = build_lora_mm(model.output, cur);122 123    cb(cur, "result_output", -1);124    res->t_logits = cur;125 126    ggml_build_forward_expand(gf, cur);127}128