File downloaded from deepdoctection/d2_casc_rcnn_X_32xd4_50_FPN_GN_2FC_publaynet_inference_only does not match the expected size

[rogerbarretocode]

i read in another issue that this was solved (#290 ). in 0.31

but i am still encountering this in latest version(0.31).

the error is this :

[0411 09:49.23 @model.py:1012] ERR File downloaded from deepdoctection/d2_casc_rcnn_X_32xd4_50_FPN_GN_2FC_publaynet_inference_only does not match the expected size! You may have downloaded a broken file, or the upstream may have modified the file.
Traceback (most recent call last):
File "/home/admin/HINGES-ERP/deep2.py", line 5, in
analyzer = dd.get_dd_analyzer() # instantiate the built-in analyzer similar to the Hugging Face space demo
File "/home/admin/.local/lib/python3.10/site-packages/deepdoctection/analyzer/dd.py", line 459, in get_dd_analyzer
return build_analyzer(cfg)
File "/home/admin/.local/lib/python3.10/site-packages/deepdoctection/analyzer/dd.py", line 274, in build_analyzer
d_layout = build_detector(cfg, "LAYOUT")
File "/home/admin/.local/lib/python3.10/site-packages/deepdoctection/analyzer/dd.py", line 153, in build_detector
return D2FrcnnTracingDetector(config_path, weights_path, categories, filter_categories=filter_categories)
NameError: name 'D2FrcnnTracingDetector' is not defined

can someone please suggest a solution ?

i am running on ubuntu

[JaMe76]

This NameError generally pops up if something is not installed. In this case I suppose that PyTorch is missing.

Check, if

import deepdoctection as dd

print(dd.pytorch_available())

prints True. If not, please follow the PyTorch installation guidelines.

Note, that PyTorch must be installed independently.

[damian-nplan]

I am getting a similar error, this time complaining about META_ARCHITECTURE when running this script:

import deepdoctection as dd
analyzer = dd.get_dd_analyzer()
print("Hello world")

Error:

[0411 11:56.21 @model.py:1012]  ERR  File downloaded from deepdoctection/d2_casc_rcnn_X_32xd4_50_FPN_GN_2FC_publaynet_inference_only does not match the expected size! You may have downloaded a broken file, or the upstream may have modified the file.
Traceback (most recent call last):
  File "/home/user/stextract/main.py", line 7, in <module>
    analyzer = dd.get_dd_analyzer()
  File "/home/user/.cache/pypoetry/virtualenvs/stextract-kaiB8QvC-py3.10/lib/python3.10/site-packages/deepdoctection/analyzer/dd.py", line 459, in get_dd_analyzer
    return build_analyzer(cfg)
  File "/home/user/.cache/pypoetry/virtualenvs/stextract-kaiB8QvC-py3.10/lib/python3.10/site-packages/deepdoctection/analyzer/dd.py", line 274, in build_analyzer
    d_layout = build_detector(cfg, "LAYOUT")
  File "/home/user/.cache/pypoetry/virtualenvs/stextract-kaiB8QvC-py3.10/lib/python3.10/site-packages/deepdoctection/analyzer/dd.py", line 153, in build_detector
    return D2FrcnnTracingDetector(config_path, weights_path, categories, filter_categories=filter_categories)
  File "/home/user/.cache/pypoetry/virtualenvs/stextract-kaiB8QvC-py3.10/lib/python3.10/site-packages/deepdoctection/extern/d2detect.py", line 441, in __init__
    self.name = self.get_name(path_weights, self.cfg.MODEL.META_ARCHITECTURE)
  File "/home/user/.cache/pypoetry/virtualenvs/stextract-kaiB8QvC-py3.10/lib/python3.10/site-packages/deepdoctection/utils/metacfg.py", line 46, in __getattr__
    raise AttributeError(name)
AttributeError: META_ARCHITECTURE

[rogerbarretocode]

@JaMe76

when i try your suggestion it says AttributeError: module deepdoctection has no attribute pytorch_available. Did you mean: 'boto3_available'?

also i am trying to run this using tensorflow .

while using tensorflow it says that GPU is missing and all tesnforflow models require a GPU . could this be the issue ?

[JaMe76]

@damian-nplan: This is something else.

The config files of layout/segmentation models have changed recently.

Unfortunately, there is no automatic way to force downloading model config files and this has to be done manually:

Download this config file from the Huggingface Hub

and replace the file with the file in your .cache, e.g. somewhere in "~/.cache/deepdoctection/configs/dd/d2/layout"

[JaMe76]

@JaMe76

when i try your suggestion it says AttributeError: module deepdoctection has no attribute pytorch_available. Did you mean: 'boto3_available'?

also i am trying to run this using tensorflow .

while using tensorflow it says that GPU is missing and all tesnforflow models require a GPU . could this be the issue ?

Yes, this is likely:

deepdoctection looks what DL library has been installed. If Tensorflow is installed it will check if a GPU is available. If not it will default to Pytorch. If Pytorch is not installed the it might result in this error.

Can you try something like:

from deepdoctection.utils.file_utils import pytorch_available
from deepdoctection.utils.env_info import auto_select_lib_and_device
from os import environ

    print(pytorch_available())
    auto_select_lib_and_device()
    print(environ.get("USE_TENSORFLOW"))
    print(environ.get("USE_PYTORCH"))
    print(environ.get("USE_CUDA"))
    print(environ.get("USE_TF"))
    print(environ.get("USE_TORCH"))

and paste the output here?

[MrBerac]

Hey @JaMe76

I'm currently working on a university project and have exactly the same error as everyone here.
The config of Huggingface is at least the same as I have already installed.

The outpunt of your code is the one here:

True
False
True
False
None
None