llama : remove check flash_attn with lora (#11104)

src/llama.cpp

@@ -11519,13 +11519,7 @@ int32_t llama_lora_adapter_set(
             struct llama_context * ctx,
             struct llama_lora_adapter * adapter,
             float scale) {
-    if (ctx->cparams.flash_attn) {
-        LLAMA_LOG_ERROR("%s: flash_attn is not compatible with LoRA\n", __func__);
-        return -1;
-    }
-
     ctx->lora_adapters[adapter] = scale;
-
     return 0;
 }