Hi, I am having a similar problem with topaz. When I do motioncor with 16-bit and then I try to run topaz train it gives me this weird error: Traceback (most recent call last):
File “cryosparc_master/cryosparc_compute/run.py”, line 116, in cryosparc_master.cryosparc_compute.run.main
File “/common/cryosparc_worker/cryosparc_compute/jobs/topaz/run_topaz.py”, line 332, in run_topaz_wrapper_train
utils.run_process(split_command)
File “/common/cryosparc_worker/cryosparc_compute/jobs/topaz/topaz_utils.py”, line 99, in run_process
assert process.returncode == 0, f"Subprocess exited with status {process.returncode} ({str_command})"
AssertionError: Subprocess exited with status 1 (/common/cryosparc_worker/deps/anaconda/envs/topaz/bin/topaz train_test_split --number 887 --seed 173124714 --image-dir /root/cryosparc/common_workspace/…/…/…//…)
which it does not give me when I try to do this on 32 bit motioncorrected images. Am I doing something wrong?
Best,
Chiara
Topaz is working fine for me with fp16 images (both motioncor & extracted with fp16) - which topaz version are you using? I am using 0.25a with CS4.6.2 and it works. Did you use Patch Motion or another algorithm for motioncor?
@chiara.rapisarda It appears that Topaz has been installed into a conda environment inside cryosparc_worker/.This may result in
broken dependencies for Topaz and/or CryoSPARC
deletion of the Topaz installation during a CryoSPARC update
Topaz should instead be installed using conda software and a conda environment outside of and independent from cryosparc_worker/ or cryosparc_master/ directories.
For more specific suggestions, please can you email us the job report for the failed job.
Due to the intricate network of prerequisites, the commands for the installation of a specific Topaz version may change over time. For a recent example installation please see Topaz training failed after update - #5 by wtempel.