Update README.md
Browse files
README.md
CHANGED
@@ -74,6 +74,12 @@ audio, sr = torchaudio.load('sample.wav') # sample rate of 16 kHz expected
|
|
74 |
embedding = ecapa2(audio)
|
75 |
```
|
76 |
|
|
|
|
|
|
|
|
|
|
|
|
|
77 |
There is no need for `ecapa2_model.eval()` or `torch.no_grad()`, this is done automatically.
|
78 |
|
79 |
### Hierarchical Feature Extraction
|
|
|
74 |
embedding = ecapa2(audio)
|
75 |
```
|
76 |
|
77 |
+
The initial calls to the JIT-model can in some cases take a very long time because of optimization attempts of the compiler. If you have issues, the JIT-optimizer can be disabled as following:
|
78 |
+
```python
|
79 |
+
with torch.jit.optimized_execution(False):
|
80 |
+
embedding = ecapa2(audio)
|
81 |
+
```
|
82 |
+
|
83 |
There is no need for `ecapa2_model.eval()` or `torch.no_grad()`, this is done automatically.
|
84 |
|
85 |
### Hierarchical Feature Extraction
|