rename to n_cls_out

Author: CISC

src/llama-hparams.h

@@ -132,7 +132,7 @@ struct llama_hparams {
     bool use_kq_norm   = true;
 
     // for Classifiers
-    uint32_t n_cls_out_labels = 1;
+    uint32_t n_cls_out = 1;
 
     // llama4
     uint32_t n_moe_layer_step        = 0;

src/llama-model.cpp

@@ -683,7 +683,7 @@ void llama_model::load_hparams(llama_model_loader & ml) {
                 ml.get_key(LLM_KV_ATTENTION_LAYERNORM_EPS,    hparams.f_norm_eps);
                 ml.get_key(LLM_KV_ATTENTION_CAUSAL,           hparams.causal_attn);
                 ml.get_key(LLM_KV_POOLING_TYPE,               hparams.pooling_type, false);
-                ml.get_arr_n(LLM_KV_CLASSIFIER_OUTPUT_LABELS, hparams.n_cls_out_labels, false);
+                ml.get_arr_n(LLM_KV_CLASSIFIER_OUTPUT_LABELS, hparams.n_cls_out, false);
 
                 switch (hparams.n_layer) {
                     case 3:
@@ -2122,8 +2122,8 @@ bool llama_model::load_tensors(llama_model_loader & ml) {
                         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_labels}, TENSOR_NOT_REQUIRED);
-                        cls_out_b = create_tensor(tn(LLM_TENSOR_CLS_OUT, "bias"),   {hparams.n_cls_out_labels},         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"), {n_embd}, 0);