Nerfies_Training.ipynb collab fails on import configs, having changed no settings

[xvdp]

I ran both collabs in sequence Nerfies_Capture_Processing.ipynb and Nerfies_Training.ipynb on the process of running the second one - having changed nothing igot an error thanks

---

ValueError Traceback (most recent call last)

in () 6 from IPython.display import display, Markdown 7 ----> 8 from nerfies import configs 9 10 5 frames /usr/lib/python3.7/dataclasses.py in _get_field(cls, a_name, a_type) 731 # For real fields, disallow mutable defaults for known types. 732 if f._field_type is _FIELD and isinstance(f.default, (list, dict, set)): --> 733 raise ValueError(f'mutable default {type(f.default)} for field ' 734 f'{f.name} is not allowed: use default_factory') 735

[xvdp]

running locallly installed code on a jupyter notebook also fails in configs - Ill debug this, but just saying, the project is awesome, but a bit finnicky.

---

ValueError Traceback (most recent call last)

in 6 from IPython.display import display, Markdown 7 ----> 8 from nerfies import configs 9 10 ~/miniconda3/envs/abj/lib/python3.9/site-packages/nerfies/configs.py in 40 @gin.configurable() 41 @dataclasses.dataclass ---> 42 class ModelConfig: 43 """Parameters for the model.""" 44 # Sample linearly in disparity rather than depth. ~/miniconda3/envs/abj/lib/python3.9/dataclasses.py in dataclass(cls, init, repr, eq, order, unsafe_hash, frozen) 1019 1020 # We're called as @dataclass without parens. -> 1021 return wrap(cls) 1022 1023 ~/miniconda3/envs/abj/lib/python3.9/dataclasses.py in wrap(cls) 1011 1012 def wrap(cls): -> 1013 return _process_class(cls, init, repr, eq, order, unsafe_hash, frozen) 1014 1015 # See if we're being called as @dataclass or @dataclass(). ~/miniconda3/envs/abj/lib/python3.9/dataclasses.py in _process_class(cls, init, repr, eq, order, unsafe_hash, frozen) 861 # things, and set the default values (as class attributes) where 862 # we can. --> 863 cls_fields = [_get_field(cls, name, type) 864 for name, type in cls_annotations.items()] 865 for f in cls_fields: ~/miniconda3/envs/abj/lib/python3.9/dataclasses.py in (.0) 861 # things, and set the default values (as class attributes) where 862 # we can. --> 863 cls_fields = [_get_field(cls, name, type) 864 for name, type in cls_annotations.items()] 865 for f in cls_fields: ~/miniconda3/envs/abj/lib/python3.9/dataclasses.py in _get_field(cls, a_name, a_type) 745 # For real fields, disallow mutable defaults for known types. 746 if f._field_type is _FIELD and isinstance(f.default, (list, dict, set)): --> 747 raise ValueError(f'mutable default {type(f.default)} for field ' 748 f'{f.name} is not allowed: use default_factory') 749 ValueError: mutable default for field warp_kwargs is not allowed: use default_factory

[antonioace]

how you fix that bro ?

[xvdp]

@antonioace I'd never used frozen dict or gin-config ( not to be confused with gin), or jax for that matter!,

when defining a @dataclass some care needs to be taken on dict or frozendict or default arguments-- they need to be wrapped, and passed on creation, neither of those seems to have been done (I may have missed a beat on the original code, but thhis i say fixes it)

Ill fork test and submit something that works for me, on and offline - sometime today.

• https://docs.python.org/3/library/dataclasses.html

[keunhong]

It seems like some updates to gin/flax have broken the current code (my fault for not freezing the version in the instructions).

I'll upload a fix sometimes this week.

[xvdp]

thanks @keunhong - great project, thanks for the work.

My x_dev branch fixes the issue, or patches it, your call.

changed 2 files: https://github.com/xvdp/nerfies/blob/x_dev/nerfies/configs.py

replace frozendict with field(default_factory=frozendict.frozendict)

https://github.com/xvdp/nerfies/blob/x_dev/notebooks/Nerfies_Training.ipynb

pass missing values to frozendict ( lr_schedule)

I didnt check all your configs, but you seems to me that as used. dataclass cannot pass default dict args, I can do a pull request if you want,

-- i see there are more errors around frozen dict - Im running the Training collab on another window

[ucichuck]

Seems like I'm still getting this issue. Anyway I can bypass it? Thanks.

Tied "!pip install git+https://github.com/xvdp/nerfies@x_dev" in notebook, but there is no lucky.

[keunhong]

Hi, this should be fixed now. Sorry it took so long.