'Speech2Text' has no attribute 'from_pretrained' about espnet_model_zoo

[kamo-naoyuki]

I'm not sure why it happened.

Could you show the version of espnet?

import espnet
print(espent.__version__)
print(espent.__file__)

[ben-8878]

print(espent.file)

print(espnet.version)
0.9.9
print(espnet.file)
/home/miniconda3/envs/espnet/lib/python3.8/site-packages/espnet/init.py

[kamo-naoyuki]

It's too old. Please update espnet

[ben-8878]

It's too old. Please update espnet

update it to latest，it work
may i ask another two question?

1. how to load local model,?
2. how to run model on gpu?

[kamo-naoyuki]

I think you can easily understand more detail just by reading asr_inference.py

how to load local model,?

You can use the following arguments:

https://github.com/espnet/espnet/blob/2e91e23e50ec05819ddf0c0f756f8aecce52048f/espnet2/bin/asr_inference.py#L53-L57

*_mode_file should be a .pthfile, and_train_configis a*.yaml` file.

how to run model on GPU?

Please use the following:

https://github.com/espnet/espnet/blob/2e91e23e50ec05819ddf0c0f756f8aecce52048f/espnet2/bin/asr_inference.py#L62

[ben-8878]

I think you can easily understand more detail just by reading asr_inference.py

how to load local model,?

You can use the following arguments:

https://github.com/espnet/espnet/blob/2e91e23e50ec05819ddf0c0f756f8aecce52048f/espnet2/bin/asr_inference.py#L53-L57

*_mode_file should be a _.pthfile, and__train_configis a*.yaml` file.

how to run model on GPU?

Please use the following:

https://github.com/espnet/espnet/blob/2e91e23e50ec05819ddf0c0f756f8aecce52048f/espnet2/bin/asr_inference.py#L62

thanks