Train FCN Alexnet model
Alqazzaz opened this issue · comments
Salma Alqazzaz commented
Hi,
I want to train the FCN Alexnet on my dataset that contains this dimension (3,160,216) and 5 classes.
I got this error
ERROR: Check failed: status == CUBLAS_STATUS_SUCCESS (11 vs. 0) CUBLAS_STATUS_MAPPING_ERROR
conv2 needs backward computation.
norm1 needs backward computation.
pool1 needs backward computation.
relu1 needs backward computation.
conv1 needs backward computation.
shift does not need backward computation.
label_label_0_split does not need backward computation.
label does not need backward computation.
data_data_0_split does not need backward computation.
data does not need backward computation.
This network produces output accuracy
This network produces output loss
Network initialization done.
Solver scaffolding done.
Finetuning from /usr/lib/python2.7/dist-packages/digits/DIGITS-master/examples/semantic-segmentation/fcn.berkeleyvision.org-master/voc-fcn-alexnet
Starting Optimization
Solving
Learning Rate Policy: step
Iteration 0, Testing net (#0)
Check failed: status == CUBLAS_STATUS_SUCCESS (11 vs. 0) CUBLAS_STATUS_MAPPING_ERROR
I appreciate any help.
Timo Sämann commented
Hi
In most cases the number of outputs doesn't match with the number of labels.
Salma Alqazzaz commented
Thanks for your answer.
Can you give me more clarifications please??
Shivam Sardana commented
@TimoSaemann I am training SegNet basic model in which I only want 4 classes not 11. How should I proceed? Please help