refactor: Split granite architectures out of llm_build_llama

This helps de-clutter the llama-family graph construction and allows
granite to diverge further (in preparation for Granite 4).

NOTE: I removed the granite scale factors from llm_build_deci because they
appear to only be there as copy-paste from llm_build_llama. The HF config
does not seem to set those values:
https://huggingface.co/Deci/DeciLM-7B/blob/main/config.json

Branch: GraniteMoEShared

Signed-off-by: Gabe Goodhart <[email protected]>

Author: gabe-l-hart

src/llama-model.cpp

@@ -4610,11 +4610,6 @@ struct llm_build_llama : public llm_graph_context {
                 inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);
             }
 
-            // For Granite architecture
-            if (hparams.f_residual_scale) {
-                cur = ggml_scale(ctx0, cur, hparams.f_residual_scale);
-            }
-
             ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);
             cb(ffn_inp, "ffn_inp", il);
 
@@ -4672,7 +4667,7 @@ struct llm_build_llama : public llm_graph_context {
                         LLM_NORM_RMS, il);
                 cb(cur, "ffn_norm", il);
 
-                ggml_tensor * moe_out = build_moe_ffn(cur,
+                cur = build_moe_ffn(cur,
                         model.layers[il].ffn_gate_inp,
                         model.layers[il].ffn_up_exps,
                         model.layers[il].ffn_gate_exps,
@@ -4683,28 +4678,7 @@ struct llm_build_llama : public llm_graph_context {
                         false, 0.0,
                         LLAMA_EXPERT_GATING_FUNC_TYPE_SOFTMAX,
                         il);
-                cb(moe_out, "ffn_moe_out", il);
-
-                // For Granite MoE Shared
-                if (hparams.n_ff_shexp > 0) {
-                    ggml_tensor * ffn_shexp = build_ffn(cur,
-                        model.layers[il].ffn_up_shexp,   NULL, NULL,
-                        model.layers[il].ffn_gate_shexp, NULL, NULL,
-                        model.layers[il].ffn_down_shexp, NULL, NULL,
-                        NULL,
-                        LLM_FFN_SILU, LLM_FFN_PAR, il);
-                    cb(ffn_shexp, "ffn_shexp", il);
-
-                    cur = ggml_add(ctx0, moe_out, ffn_shexp);
-                    cb(cur, "ffn_out", il);
-                } else {
-                    cur = moe_out;
-                }
-            }
-
-            // For Granite architecture
-            if (hparams.f_residual_scale) {
-                cur = ggml_scale(ctx0, cur, hparams.f_residual_scale);
+                cb(cur, "ffn_moe_out", il);
             }
 
             cur = ggml_add(ctx0, cur, ffn_inp);
@@ -4729,11 +4703,6 @@ struct llm_build_llama : public llm_graph_context {
         // lm_head
         cur = build_lora_mm(model.output, cur);
 
-        // For Granite architecture
-        if (hparams.f_logit_scale) {
-            cur = ggml_scale(ctx0, cur, 1.0f / hparams.f_logit_scale);
-        }
-
         cb(cur, "result_output", -1);
         res->t_logits = cur;
 
@@ -4844,11 +4813,6 @@ struct llm_build_deci : public llm_graph_context {
                 continue;
             }
 
-            // For Granite architecture
-            if (hparams.f_residual_scale) {
-                cur = ggml_scale(ctx0, cur, hparams.f_residual_scale);
-            }
-
             // modified to support attention-free layer of Llama-3_1-Nemotron-51B
             ggml_tensor * ffn_inp = cur;
             if (n_head > 0) {
@@ -4872,11 +4836,6 @@ struct llm_build_deci : public llm_graph_context {
                 cb(cur, "ffn_out", il);
             }
 
-            // For Granite architecture
-            if (hparams.f_residual_scale) {
-                cur = ggml_scale(ctx0, cur, hparams.f_residual_scale);
-            }
-
             cur = ggml_add(ctx0, cur, ffn_inp);
             cb(cur, "ffn_out", il);
 
@@ -4899,11 +4858,6 @@ struct llm_build_deci : public llm_graph_context {
         // lm_head
         cur = build_lora_mm(model.output, cur);
 
-        // For Granite architecture
-        if (hparams.f_logit_scale) {
-            cur = ggml_scale(ctx0, cur, 1.0f / hparams.f_logit_scale);
-        }
-
         cb(cur, "result_output", -1);
         res->t_logits = cur;
 
@@ -12242,6 +12196,194 @@ struct llm_build_arwkv7 : public llm_build_rwkv7_base {
     }
 };
 
+
+struct llm_build_granite : public llm_graph_context {
+    llm_build_granite(
+        const llama_model & model,
+        const llm_graph_params & params,
+        ggml_cgraph * gf,
+        const bool use_rope = true)
+        : llm_graph_context(params) {
+
+        const int64_t n_embd_head = hparams.n_embd_head_v;
+
+        GGML_ASSERT(n_embd_head == hparams.n_embd_head_k);
+        GGML_ASSERT(n_embd_head == hparams.n_rot);
+
+        ggml_tensor * cur;
+        ggml_tensor * inpL;
+
+        inpL = build_inp_embd(model.tok_embd);
+
+        // inp_pos - used for rope if enabled
+        ggml_tensor * inp_pos;
+        if (use_rope) {
+            inp_pos = build_inp_pos();
+        }
+
+        auto * inp_attn = build_attn_inp_kv_unified();
+
+        const float kq_scale = hparams.f_attention_scale == 0.0f ? 1.0f/sqrtf(float(n_embd_head)) : hparams.f_attention_scale;
+        for (int il = 0; il < n_layer; ++il) {
+            ggml_tensor * inpSA = inpL;
+
+            // norm
+            cur = build_norm(inpL,
+                    model.layers[il].attn_norm, NULL,
+                    LLM_NORM_RMS, il);
+            cb(cur, "attn_norm", il);
+
+            // self-attention
+            {
+                // compute Q and K and (optionally) RoPE them
+                ggml_tensor * Qcur = build_lora_mm(model.layers[il].wq, cur);
+                cb(Qcur, "Qcur", il);
+                if (model.layers[il].bq) {
+                    Qcur = ggml_add(ctx0, Qcur, model.layers[il].bq);
+                    cb(Qcur, "Qcur", il);
+                }
+
+                ggml_tensor * Kcur = build_lora_mm(model.layers[il].wk, cur);
+                cb(Kcur, "Kcur", il);
+                if (model.layers[il].bk) {
+                    Kcur = ggml_add(ctx0, Kcur, model.layers[il].bk);
+                    cb(Kcur, "Kcur", il);
+                }
+
+                ggml_tensor * Vcur = build_lora_mm(model.layers[il].wv, cur);
+                cb(Vcur, "Vcur", il);
+                if (model.layers[il].bv) {
+                    Vcur = ggml_add(ctx0, Vcur, model.layers[il].bv);
+                    cb(Vcur, "Vcur", il);
+                }
+
+                Qcur = ggml_reshape_3d(ctx0, Qcur, n_embd_head, n_head,    n_tokens);
+                Kcur = ggml_reshape_3d(ctx0, Kcur, n_embd_head, n_head_kv, n_tokens);
+                Vcur = ggml_reshape_3d(ctx0, Vcur, n_embd_head, n_head_kv, n_tokens);
+
+                if (use_rope) {
+                    ggml_tensor * rope_factors = model.get_rope_factors(n_ctx_per_seq, il);
+                    Qcur = ggml_rope_ext(
+                            ctx0, Qcur, inp_pos, rope_factors,
+                            n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,
+                            ext_factor, attn_factor, beta_fast, beta_slow
+                            );
+
+                    Kcur = ggml_rope_ext(
+                            ctx0, Kcur, inp_pos, rope_factors,
+                            n_rot, rope_type, n_ctx_orig, freq_base, freq_scale,
+                            ext_factor, attn_factor, beta_fast, beta_slow
+                            );
+                }
+
+                cb(Qcur, "Qcur", il);
+                cb(Kcur, "Kcur", il);
+                cb(Vcur, "Vcur", il);
+
+                cur = build_attn(inp_attn, gf,
+                        model.layers[il].wo, model.layers[il].bo,
+                        Qcur, Kcur, Vcur, nullptr, nullptr, kq_scale, il);
+                cb(cur, "attn_out", il);
+            }
+
+            if (il == n_layer - 1) {
+                // skip computing output for unused tokens
+                ggml_tensor * inp_out_ids = build_inp_out_ids();
+                cur   = ggml_get_rows(ctx0,   cur, inp_out_ids);
+                inpSA = ggml_get_rows(ctx0, inpSA, inp_out_ids);
+            }
+
+            // For Granite architectures - scale residual
+            cur = ggml_scale(ctx0, cur, hparams.f_residual_scale);
+            ggml_tensor * ffn_inp = ggml_add(ctx0, cur, inpSA);
+            cb(ffn_inp, "ffn_inp", il);
+
+            // feed-forward network (non-MoE)
+            if (model.layers[il].ffn_gate_inp == nullptr) {
+
+                cur = build_norm(ffn_inp,
+                        model.layers[il].ffn_norm, NULL,
+                        LLM_NORM_RMS, il);
+                cb(cur, "ffn_norm", il);
+
+                cur = build_ffn(cur,
+                        model.layers[il].ffn_up,   model.layers[il].ffn_up_b,   NULL,
+                        model.layers[il].ffn_gate, model.layers[il].ffn_gate_b, NULL,
+                        model.layers[il].ffn_down, model.layers[il].ffn_down_b, NULL,
+                        NULL,
+                        LLM_FFN_SILU, LLM_FFN_PAR, il);
+                cb(cur, "ffn_out", il);
+
+            } else {
+                // MoE branch
+                cur = build_norm(ffn_inp,
+                        model.layers[il].ffn_norm, NULL,
+                        LLM_NORM_RMS, il);
+                cb(cur, "ffn_norm", il);
+
+                ggml_tensor * moe_out = build_moe_ffn(cur,
+                        model.layers[il].ffn_gate_inp,
+                        model.layers[il].ffn_up_exps,
+                        model.layers[il].ffn_gate_exps,
+                        model.layers[il].ffn_down_exps,
+                        nullptr,
+                        n_expert, n_expert_used,
+                        LLM_FFN_SILU, true,
+                        false, 0.0,
+                        LLAMA_EXPERT_GATING_FUNC_TYPE_SOFTMAX,
+                        il);
+                cb(moe_out, "ffn_moe_out", il);
+
+                // For Granite MoE Shared
+                if (hparams.n_ff_shexp > 0) {
+                    ggml_tensor * ffn_shexp = build_ffn(cur,
+                        model.layers[il].ffn_up_shexp,   NULL, NULL,
+                        model.layers[il].ffn_gate_shexp, NULL, NULL,
+                        model.layers[il].ffn_down_shexp, NULL, NULL,
+                        NULL,
+                        LLM_FFN_SILU, LLM_FFN_PAR, il);
+                    cb(ffn_shexp, "ffn_shexp", il);
+
+                    cur = ggml_add(ctx0, moe_out, ffn_shexp);
+                    cb(cur, "ffn_out", il);
+                } else {
+                    cur = moe_out;
+                }
+            }
+
+            // For Granite architectures - scale residual
+            cur = ggml_scale(ctx0, cur, hparams.f_residual_scale);
+            cur = ggml_add(ctx0, cur, ffn_inp);
+            cb(cur, "ffn_out", il);
+
+            cur = build_cvec(cur, il);
+            cb(cur, "l_out", il);
+
+            // input for next layer
+            inpL = cur;
+        }
+
+        cur = inpL;
+
+        cur = build_norm(cur,
+                model.output_norm, NULL,
+                LLM_NORM_RMS, -1);
+
+        cb(cur, "result_norm", -1);
+        res->t_embd = cur;
+
+        // lm_head
+        cur = build_lora_mm(model.output, cur);
+
+        // For Granite architectures - scale logits
+        cur = ggml_scale(ctx0, cur, 1.0f / hparams.f_logit_scale);
+        cb(cur, "result_output", -1);
+        res->t_logits = cur;
+
+        ggml_build_forward_expand(gf, cur);
+    }
+};
+
 // ref: https://github.com/facebookresearch/chameleon
 // based on the original build_llama() function, changes:
 //   * qk-norm
@@ -12949,9 +13091,6 @@ llm_graph_result_ptr llama_model::build_graph(
         case LLM_ARCH_LLAMA:
         case LLM_ARCH_LLAMA4:
         case LLM_ARCH_MINICPM:
-        case LLM_ARCH_GRANITE:
-        case LLM_ARCH_GRANITE_MOE:
-        case LLM_ARCH_GRANITE_MOE_SHARED:
             {
                 llm = std::make_unique<llm_build_llama>(*this, params, gf);
             } break;
@@ -13182,6 +13321,12 @@ llm_graph_result_ptr llama_model::build_graph(
             {
                 llm = std::make_unique<llm_build_arwkv7>(*this, params, gf);
             } break;
+        case LLM_ARCH_GRANITE:
+        case LLM_ARCH_GRANITE_MOE:
+        case LLM_ARCH_GRANITE_MOE_SHARED:
+            {
+                llm = std::make_unique<llm_build_granite>(*this, params, gf);
+            } break;
         case LLM_ARCH_CHAMELEON:
             {
                 llm = std::make_unique<llm_build_chameleon>(*this, params, gf);