Merge pull request #662 from Alejandro-Casanova/patch-22

fixed typo

09_pytorch_model_deployment.ipynb

@@ -2475,7 +2475,7 @@
     "\n",
     "Are these tradeoffs worth it?\n",
     "\n",
-    "Perhaps if we had unlimited compute power but for our use case of deploying the FoodVision Mini model to a smaller device (e.g. a mobile phone), we'd likely start out with the EffNetB2 model for faster predictions at a slightly reduced performance but dramatically smaller "
+    "Perhaps if we had unlimited compute power but for our use case of deploying the FoodVision Mini model to a smaller device (e.g. a mobile phone), we'd likely start out with the EffNetB2 model for faster predictions at a slightly reduced performance but dramatically smaller size."
    ]
   },
   {