Update README.md

README.md

@@ -17,5 +17,5 @@ The thought process of the original model is often output in Chinese or English,
 DeepSeek-R1は推奨温度が0.5-0.7である事に注意してください  
 Please note that the recommended temperature for DeepSeek-R1 is 0.5-0.7  
 
-詳細はテスト中です。  
-Details are currently being tested.  
+私が試したいくつかの推論ツールはまだ不具合があるように見えます。[llama.cpp](dahara1/DeepSeek-R1-Distill-Qwen-14B-unsloth-gguf-japanese-imatrix)をおすすめしておきます  
+Some of the inference tools I've tried still seem to have bugs. I recommend [llama.cpp](dahara1/DeepSeek-R1-Distill-Qwen-14B-unsloth-gguf-japanese-imatrix).