TypeError: <lambda>() missing 1 required positional argument: 'k'

[changleilei]

Hello, when i use pytorch-lit prepare a model, got a TypeError as title. The detail as blow:

File "nlp/rct-FPM-rhino/big_model/prepare_model.py", line 16, in prepare_model
prepare_params(model, args.save_path, dtype='float32')
File "miniconda3/envs/rhino/lib/python3.8/site-packages/pytorch_lit/export.py", line 19, in prepare_params
_params_to_memmap(parameters, path.join(save_dir, "model.bin"),
File "miniconda3/envs/rhino/lib/python3.8/site-packages/pytorch_lit/export.py", line 52, in _params_to_memmap
param = get_param(k)
File "miniconda3/envs/rhino/lib/python3.8/site-packages/pytorch_lit/export.py", line 50, in
get_param = lambda key: params"get"
TypeError: () missing 1 required positional argument: 'k'

package list:

certifi 2021.10.8
numpy 1.21.4
pip 21.2.4
pytorch-lit 0.1.6
setuptools 58.0.4
torch 1.10.0
tqdm 4.62.3
typing_extensions 4.0.1
wheel 0.37.0

model: gpt-j-6B

Have any suggesstion? Thanks.

[AminRezaei0x443]

Hi,
There was a minor issue which is now fixed.
Try upgrading to 0.1.7.