llama : avoid using "optional" keyword (#4283)

Author: ggerganov

llama.cpp

@@ -1991,11 +1991,11 @@ struct llama_model_loader {
         return tensor;
     }
 
-    struct ggml_tensor * create_tensor(struct ggml_context * ctx, const std::string & name, const std::vector<int64_t> & ne, ggml_backend_type backend, bool optional = false) {
+    struct ggml_tensor * create_tensor(struct ggml_context * ctx, const std::string & name, const std::vector<int64_t> & ne, ggml_backend_type backend, bool required = true) {
         struct ggml_tensor * cur = ggml_get_tensor(ctx_meta, name.c_str());
 
         if (cur == NULL) {
-            if (optional) {
+            if (!required) {
                 return NULL;
             }
             throw std::runtime_error(format("%s: tensor '%s' not found", __func__, name.c_str()));
@@ -2816,10 +2816,10 @@ static void llm_load_tensors(
                         layer.wo = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_OUT, "weight", i), {n_embd, n_embd},     backend_split);
 
                         // optional bias tensors
-                        layer.bq = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_Q,   "bias", i), {n_embd},     backend, true);
-                        layer.bk = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_K,   "bias", i), {n_embd_gqa}, backend, true);
-                        layer.bv = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_V,   "bias", i), {n_embd_gqa}, backend, true);
-                        layer.bo = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_OUT, "bias", i), {n_embd},     backend, true);
+                        layer.bq = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_Q,   "bias", i), {n_embd},     backend, false);
+                        layer.bk = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_K,   "bias", i), {n_embd_gqa}, backend, false);
+                        layer.bv = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_V,   "bias", i), {n_embd_gqa}, backend, false);
+                        layer.bo = ml.create_tensor(ctx, tn(LLM_TENSOR_ATTN_OUT, "bias", i), {n_embd},     backend, false);
 
                         layer.ffn_norm = ml.create_tensor(ctx, tn(LLM_TENSOR_FFN_NORM, "weight", i), {n_embd}, backend);