model: move load_hparams and load_tensors to per-model definition (#22004)

* git-friendly migration * add build_graph * nits * exclude old code from build * wip * add llm_arch_model_i * prepare downstream functions * nits * nits * wip * wip * add back create_tensor_qkv * fix files missing include * enforce one llm_build per arch * cmake: use glob * missing model params * nits * wip * wip (2) * wip (3) * test-llama-archs is happy * improve switch case * move more stuff into llm_arch_model_i * fix downstream code * nits * nits (2) * fix order * llama_model_base * LLAMA_LOAD_LOCALS * small fix * fix build errors * auto * rm migration script and ifdef
2026-05-04 12:36:59 +02:00
parent c84e6d6db5
commit 994118a183
129 changed files with 10667 additions and 8117 deletions
@@ -1,6 +1,69 @@
 #include "models.h"

-llm_build_modern_bert::llm_build_modern_bert(const llama_model & model, const llm_graph_params & params) : llm_graph_context(params) {
+void llama_model_modern_bert::load_arch_hparams(llama_model_loader & ml) {
+    const bool found_swa = ml.get_key(LLM_KV_ATTENTION_SLIDING_WINDOW, hparams.n_swa, false);
+    if (found_swa && hparams.n_swa > 0) {
+        hparams.swa_type = LLAMA_SWA_TYPE_SYMMETRIC;
+        ml.get_key(LLM_KV_ROPE_FREQ_BASE_SWA, hparams.rope_freq_base_train_swa, false);
+        uint32_t swa_period = 3;
+        ml.get_key_or_arr(LLM_KV_ATTENTION_SLIDING_WINDOW_PATTERN, swa_period, false);
+        hparams.set_swa_pattern(swa_period, true);
+    } else {
+        hparams.swa_type = LLAMA_SWA_TYPE_NONE;
+    }
+
+    ml.get_key(LLM_KV_ATTENTION_LAYERNORM_EPS, hparams.f_norm_eps);
+
+    switch (hparams.n_layer) {
+        case 12:
+            type = LLM_TYPE_47M; break; // granite-embedding-small
+        case 22:
+            type = LLM_TYPE_149M; break; // modern-bert-base
+        case 28:
+            type = LLM_TYPE_395M; break; // modern-bert-large
+        default: type = LLM_TYPE_UNKNOWN;
+    }
+}
+
+void llama_model_modern_bert::load_arch_tensors(llama_model_loader &) {
+    LLAMA_LOAD_LOCALS;
+
+    tok_embd = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD, "weight"), {n_embd, n_vocab}, 0);
+    tok_norm = create_tensor(tn(LLM_TENSOR_TOKEN_EMBD_NORM, "weight", 0), {n_embd}, 0);
+
+    output_norm = create_tensor(tn(LLM_TENSOR_OUTPUT_NORM, "weight"), {n_embd}, 0);
+
+    for(int i = 0; i < n_layer; ++i) {
+        auto& layer = layers[i];
+
+        if ( i != 0 ) {
+            layer.attn_norm = create_tensor(tn(LLM_TENSOR_ATTN_NORM, "weight", i), {n_embd}, 0);
+        } else{
+            // layer 0 uses identity
+            layer.attn_norm = create_tensor(tn(LLM_TENSOR_ATTN_NORM, "weight", i), {n_embd}, TENSOR_NOT_REQUIRED);
+        }
+
+
+        layer.wqkv = create_tensor(tn(LLM_TENSOR_ATTN_QKV, "weight", i), {n_embd, 3 * n_embd }, 0);
+        layer.wo = create_tensor(tn(LLM_TENSOR_ATTN_OUT,   "weight", i), {n_embd, n_embd}, 0);
+
+        layer.ffn_up   = create_tensor(tn(LLM_TENSOR_FFN_UP,   "weight", i), {n_embd, 2 * n_ff}, 0);
+        layer.ffn_down = create_tensor(tn(LLM_TENSOR_FFN_DOWN, "weight", i), {n_ff, n_embd}, 0);
+        layer.ffn_norm = create_tensor(tn(LLM_TENSOR_FFN_NORM, "weight", i), {n_embd}, 0);
+    }
+
+    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);
+    cls       = create_tensor(tn(LLM_TENSOR_CLS,      "weight"), {n_embd, n_embd},            TENSOR_NOT_REQUIRED);
+    cls_norm  = create_tensor(tn(LLM_TENSOR_CLS_NORM, "weight"), {n_embd},                    TENSOR_NOT_REQUIRED);
+
+}
+
+std::unique_ptr<llm_graph_context> llama_model_modern_bert::build_arch_graph(const llm_graph_params & params) const {
+    return std::make_unique<graph>(*this, params);
+}
+
+llama_model_modern_bert::graph::graph(const llama_model & model, const llm_graph_params & params) : 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());