I’m having a trouble with Topaz training. The message is like this.
[CPU: 173.4 MB] Starting Topaz process using version /: can't find '__main__' module in '/'...
[CPU: 173.4 MB] Traceback (most recent call last):
File "cryosparc2_worker/cryosparc2_compute/run.py", line 82, in cryosparc2_compute.run.main
File "cryosparc2_compute/jobs/topaz/run_topaz.py", line 123, in run_topaz_wrapper_train
if int(version_vals[1]) >= 2 and int(version_vals[2]) >= 3: ver_023_plus = True
IndexError: list index out of range
Hi, my topaz version is 0.2.3 but I still failed to start topaz in cryosparc with the same error in topaz train. My cryosparc version is 2.14.2 and my CUDA version related to cryosparc and topaz is 8.0. The driver 430.40 could accept 10.0 version but now we used 8.0.
Here’s my log in web:
Starting Topaz process using version /: can't find '__main__' module in '/'...
[CPU: 161.3 MB] Traceback (most recent call last):
File "cryosparc2_worker/cryosparc2_compute/run.py", line 82, in cryosparc2_compute.run.main
File "cryosparc2_compute/jobs/topaz/run_topaz.py", line 123, in run_topaz_wrapper_train
if int(version_vals[1]) >= 2 and int(version_vals[2]) >= 3: ver_023_plus = True
IndexError: list index out of range
It seems to me that there possibly is an issue with your Topaz install. Could you try reinstalling Topaz into an Anaconda environment in Python 2.7? You may also have to change the PyTorch installation method to account for CUDA 8.0 by running the command conda install topaz cudatoolkit=8.0 -c tbepler -c pytorch.
Afterwards, could you check the Topaz version the same way I mentioned in an earlier response in this topic?