Zengwei Yao c059ef3169

Keep model_avg on cpu (#348) ...

* keep model_avg on cpu

* explicitly convert model_avg to cpu

* minor fix

* remove device convertion for model_avg

* modify usage of the model device in train.py

* change model.device to next(model.parameters()).device for decoding

* assert params.start_epoch>0

* assert params.start_epoch>0, params.start_epoch

2022-05-07 10:42:34 +08:00

..

__init__.py

Model average (#344)

2022-05-05 21:20:04 +08:00

asr_datamodule.py

Model average (#344)

2022-05-05 21:20:04 +08:00

beam_search.py

Model average (#344)

2022-05-05 21:20:04 +08:00

conformer.py

Model average (#344)

2022-05-05 21:20:04 +08:00

decode.py

Keep model_avg on cpu (#348)

2022-05-07 10:42:34 +08:00

decoder.py

Model average (#344)

2022-05-05 21:20:04 +08:00

encoder_interface.py

Model average (#344)

2022-05-05 21:20:04 +08:00

export.py

Model average (#344)

2022-05-05 21:20:04 +08:00

joiner.py

Model average (#344)

2022-05-05 21:20:04 +08:00

model.py

Model average (#344)

2022-05-05 21:20:04 +08:00

optim.py

Model average (#344)

2022-05-05 21:20:04 +08:00

scaling.py

Model average (#344)

2022-05-05 21:20:04 +08:00

train.py

Keep model_avg on cpu (#348)

2022-05-07 10:42:34 +08:00