Open nerdykamil opened 6 years ago
same issue here with different datasets. If I find the solution I will provide, in the meantime check here: issues/373
in #373 it is suggested to skip data augmentation process, is that really a good choice. Also, changing subdivisions didn't help me.
changing subdivisions didn't help me too. But when I changed the batch size from 2 to 4 I avoided the issue. However, for full version of YOLO due to memory considerations if I would have to keep the batch size as 1 or 2, I don't know what to do then...
this is probably a problem due to the data being in png format, try converting to jpg and see if it fixes your problem
@Ridhwanluthra Yes, I tried that too but then got the bad Huffman code warning and termination with same error.
Hi! I'm getting the exact same issue here. Could anyone solve it? I've tried increasing subdivisions and removing data augmentation as well. Nothing works.
Any help would be highly appreciated. Thanks
I am training tiny yolo model for INRIA dataset, after I start the training I get the following error, any idea what is wrong here?
Dataset size: 614 Dataset of 614 instance(s) libpng warning: iCCP: known incorrect sRGB profile libpng warning: iCCP: known incorrect sRGB profile Training statistics: Learning rate : 1e-05 Batch size : 16 Epoch number : 1000 Backup every : 2000 step 1 - loss 99.0089950562 - moving ave loss 99.0089950562 libpng warning: iCCP: known incorrect sRGB profile Traceback (most recent call last): File "/home/mypc/.local/bin/flow", line 6, in
cliHandler(sys.argv)
File "/home/mypc/.local/lib/python2.7/site-packages/darkflow/cli.py", line 29, in cliHandler
print('Enter training ...'); tfnet.train()
File "/home/mypc/.local/lib/python2.7/site-packages/darkflow/net/flow.py", line 39, in train
for i, (x_batch, datum) in enumerate(batches):
File "/home/mypc/.local/lib/python2.7/site-packages/darkflow/net/yolo/data.py", line 126, in shuffle
x_batch = np.concatenate(x_batch, 0)
ValueError: need at least one array to concatenate