Update README.md

README.md

@@ -14,7 +14,7 @@ datasets:
 - voxceleb
 metrics:
 - EER
-- min_dct
+- min\_dct
 ---
 
 # Speaker Verification with xvector embeddings on Voxceleb
@@ -54,6 +54,9 @@ signal, fs =torchaudio.load('samples/audio_samples/example1.wav')
 embeddings = classifier.encode_batch(signal)
 ```
 
+### Inference on GPU
+To perform inference on the GPU, add  `run_opts={"device":"cuda"}`  when calling the `from_hparams` method.
+
 #### Referencing xvectors
 ```@inproceedings{DBLP:conf/odyssey/SnyderGMSPK18,
   author    = {David Snyder and
@@ -79,6 +82,6 @@ embeddings = classifier.encode_batch(signal)
     year = {2021},
     publisher = {GitHub},
     journal = {GitHub repository},
-    howpublished = {\\\\\\\\url{https://github.com/speechbrain/speechbrain}},
+    howpublished = {\\\\\\\\\\\\\\\\url{https://github.com/speechbrain/speechbrain}},
   }
 ```