Update README.md

README.md

@@ -48,8 +48,7 @@ import torchaudio
 ecapa2_model = torch.jit.load(model_file, map_location='cpu')
 audio, sr = torchaudio.load('sample.wav') # sample rate of 16 kHz expected
 
-with torch.no_grad():
-  embedding = ecapa2_model(audio)
+embedding = ecapa2_model(audio)
 ```
 
 For faster, 16-bit half-precision CUDA inference (recommended):
@@ -62,10 +61,11 @@ ecapa2_model = torch.jit.load(model_file, map_location='cuda')
 ecapa2_model.half() # optional, but results in faster inference
 audio, sr = torchaudio.load('sample.wav') # sample rate of 16 kHz expected
 
-with torch.no_grad():
-  embedding = ecapa2_model(audio)
+embedding = ecapa2_model(audio)
 ```
 
+There is no need for explicitly setting `eval` model and disabling gradient calculations. This is done automatically.
+
 ### Hierarchical Feature Extraction
 
 For the extraction of other hierachical features, the `label` argument can be used, which accepts a string containing the feature ids separated with '|':