search

zcemycl / TF2DeepFloorplan

TF2 Deep FloorPlan Recognition using a Multi-task Network with Room-boundary-Guided Attention. Enable tensorboard, quantization, flask, tflite, docker, github actions and google colab.
GNU General Public License v3.0
183 stars 60 forks source link

The log/store/G is crashed #81

Closed K25801 closed 8 months ago

K25801 commented 12 months ago

I tried to download the model from your Google Drive links and I found them unavailable, then I used cloud converter to convert the zip to 7z and successfully downloaded the log file. While the files in the directory log\store seem to crash, I always get an error "tensorflow.python.framework.errors_impl.NotFoundError: Unsuccessful TensorSliceReader constructor: Failed to find any matching files for log/store/G". Can you provide a new download link for your model?

rjtp5670 commented 8 months ago

Same as me. When I run the deploy.py script, it raises the same error. (No G directory in log/store folder although I trained model.

zcemycl commented 8 months ago

@rjtp5670 @K25801 There should not be a G directory, but a directory named log/store which contains files like, 

G.data-00000-of-00001
G.index

And the settings for argparse should be listed in TF2DeepFloorplan/docs/notebook.toml.

Make sure your logdir, modeldir, weight and loadmethod should be matched accordingly.

In training process, logdir and modeldir are used for logging the model and saving the model weights respectively. (logdir also contains model weights as G, modeldir contains them as .pb file).

In deploy process, if you want to deploy it through loadmethod log, specify weight as log/store/G (logdir+G name). If you want to deploy it through loadmethod pb, specify weight as model/store (modeldir).

zcemycl commented 8 months ago

@rjtp5670 @ROBOTIS-David I am a fan of ROBOTIS robot :)... Did buy a turtlebot few years ago.