oh I guess that link is broken

Try this one: https://ngc.nvidia.com/catalog/models/nvidia:tacotron2_pyt_ckpt_amp

Where can I put this file on deep learning examples folder directory?

you can put it anywhere, but pass the path to it after the --tacotron flag

@User did you ever heard this error? FileNotFoundError: [Errno 2] No such file or directory: 'flag'

Edit I fixed

another error

Traceback (most recent call last): File "inference.py", line 274, in main() File "inference.py", line 210, in main args.fp16, args.cpu, forward_is_infer=True) File "inference.py", line 119, in load_and_setup_model state_dict = torch.load(checkpoint, map_location=torch.device('cpu'))['state_dict'] KeyError: 'state_dict'

what is this

how can i fix this

open a python console, do

loaded = torch.load("/path/to/checkpoint.pt", map_location=torch.device("cpu")

and then see what's in

loaded

.

Imma jump my pc rn

It should be a python dict with a key called

state_dict

but apparently it's not that.

em wich one the waveglow model?

try both

ok

i see a . . .

try

loaded.keys()

i think i do wrong

missing a close parentheses in your original

torch.load()

can we do that in anaconda prompt?

not sure I understand the question, but type up arrow until you see the line you originally typed

and then add an extra parentheses

I would also recommend giving https://jupyter.org/ a try, it is more forgiving than the python prompt.

jupyterlab

?

or notebook

lab