Fixes backtick causing CI failure

_posts/2017-12-14-image-classification-on-small-datasets/image-classification-on-small-datasets.Rmd

@@ -435,7 +435,7 @@ unfreeze_weights(conv_base, from = "block3_conv1")
 
 Now you can begin fine-tuning the network. You'll do this with the RMSProp optimizer, using a very low learning rate. The reason for using a low learning rate is that you want to limit the magnitude of the modifications you make to the representations of the three layers you're fine-tuning. Updates that are too large may harm these representations.
 
-````{r}
+```{r}
 model %>% compile(
   loss = "binary_crossentropy",
   optimizer = optimizer_rmsprop(lr = 1e-5),