ValueError: cannot reshape array of size

After 200 000 iterations of training I’m trying to evaluate the network by GUI
but got this error:

ValueError Traceback (most recent call last)
~/anaconda3/envs/DLC-GPU/lib/python3.7/site-packages/deeplabcut/gui/evaluate_network.py in evaluate_network(self, event)
338 plotting=plotting,
339 show_errors=True,
→ 340 comparisonbodyparts=self.bodyparts,
341 )
342

~/anaconda3/envs/DLC-GPU/lib/python3.7/site-packages/deeplabcut/pose_estimation_tensorflow/evaluate.py in evaluate_network(config, Shuffles, trainingsetindex, plotting, show_errors, comparisonbodyparts, gputouse, rescale, modelprefix, c_engine)
568 gputouse=gputouse,
569 modelprefix=modelprefix,
→ 570 c_engine=c_engine,
571 )
572 else:

~/anaconda3/envs/DLC-GPU/lib/python3.7/site-packages/deeplabcut/pose_estimation_tensorflow/evaluate_multianimal.py in evaluate_multianimal_full(config, Shuffles, trainingsetindex, plotting, show_errors, comparisonbodyparts, gputouse, modelprefix, c_engine)
321 # Evaluate PAF edge lengths to calibrate distnorm
322 temp = GT.unstack(“bodyparts”)[joints]
→ 323 xy = temp.values.reshape((-1, 2, temp.shape[1])).swapaxes(
324 1, 2
325 )

ValueError: cannot reshape array of size 244 into shape (2,4)

the default in multi-animal DLC is to stop at 200,000 unless you change it. my guess is this is the normal crash out, and if you look in dlc-models you will see the snapshots saved. so you can evaluate.

Thanks for reply! Yes, I can see the other snapshots in the dlc-models folder . How I can determine in the below function or GUI to use a specific snapshot?
deeplabcut.evaluate_network* (config_path,Shuffles=[1], plotting=True).

Hi @fahimeh_farahnakian, you could change snapshotindex in the config.yaml. The latest snapshot is used by default, but if set to 0 for example it’d use the first one :slight_smile: