Update README.md

README.md

@@ -137,4 +137,4 @@ print(generate(instruction))
 ```
 
 ## Benchmark
-Running the model on a Nvidia T4 with 16GB VRAM, it takes aprox 0.15 seconds to generate a token
+Running the model on a *Nvidia T4 with 16GB VRAM*, it takes aprox **0.15 seconds** to generate a token