Skip to content

Commit ddcff31

Browse files
authored
Updating WaveGlow checkpoint to latest (#2080)
* Updating WaveGlow checkpoint to latest Which was unavailable for the couple of days Fixes #2070 * Update text_to_speech_with_torchaudio.py * Update text_to_speech_with_torchaudio.py
1 parent 2047596 commit ddcff31

File tree

1 file changed

+4
-4
lines changed

1 file changed

+4
-4
lines changed

intermediate_source/text_to_speech_with_torchaudio.py

+4-4
Original file line numberDiff line numberDiff line change
@@ -293,11 +293,11 @@ def text_to_sequence(text):
293293
# Waveglow
294294
# ~~~~~~~~
295295
#
296-
# Waveglow is a vocoder published by Nvidia. The pretrained weight is
297-
# publishe on Torch Hub. One can instantiate the model using ``torch.hub``
296+
# Waveglow is a vocoder published by Nvidia. The pretrained weights are
297+
# published on Torch Hub. One can instantiate the model using ``torch.hub``
298298
# module.
299299
#
300-
if False:
300+
if torch.cuda.is_available():
301301
waveglow = torch.hub.load('NVIDIA/DeepLearningExamples:torchhub', 'nvidia_waveglow', model_math='fp32')
302302
else:
303303
# Workaround to load model mapped on GPU
@@ -309,7 +309,7 @@ def text_to_sequence(text):
309309
pretrained=False,
310310
)
311311
checkpoint = torch.hub.load_state_dict_from_url(
312-
"https://api.ngc.nvidia.com/v2/models/nvidia/waveglowpyt_fp32/versions/1/files/nvidia_waveglowpyt_fp32_20190306.pth",
312+
"https://api.ngc.nvidia.com/v2/models/nvidia/waveglow_ckpt_fp32/versions/19.09.0/files/nvidia_waveglowpyt_fp32_20190427",
313313
progress=False,
314314
map_location=device,
315315
)

0 commit comments

Comments
 (0)