Topaz train error: 'MRCHeader' object has no attribute 'model'

hi all, experiencing an issue similar to this.

we are running v4.4.0+231114.

The micrographs are float16, is it possible this is the issue?

AttributeError: 'MRCHeader' object has no attribute 'model'

then follows many rows of e.g.:
WARNING: no micrograph found matching image name "003989035438293372028_FoilHole_18171933_Data_18115465_50_20240618_224941_EER_patch_aligned_doseweighted". Skipping it.

Then the final error:

[CPU:  282.2 MB]
index = _extract_index(arrays)

[CPU:  282.2 MB]
File (path)/topaz/lib/python3.9/site-packages/pandas/core/internals/construction.py", line 674, in _extract_index

[CPU:  282.2 MB]
raise ValueError("All arrays must be of the same length")

[CPU:  282.2 MB]
ValueError: All arrays must be of the same length

[CPU:  282.2 MB]
Traceback (most recent call last):
  File "cryosparc_master/cryosparc_compute/run.py", line 95, in cryosparc_master.cryosparc_compute.run.main
  File "(path)/cryosparc_compute/jobs/topaz/run_topaz.py", line 308, in run_topaz_wrapper_train
    utils.run_process(split_command)
  File "(path)/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 ((path)/topaz.sh train_test_split --number 1068 --seed 1808560700 --image-dir (path)/…)

thanks in advance

Jesse

A Topaz Train job with

  • CryoSPARC v4.5.3
  • Topaz 0.2.5a
  • float16 motion-corrected exposures from the EMPIAR-10025 subset

completed as expected. Given the different setup and data from yours, I cannot rule out a float16 issue, but I think it is unlikely the cause of ValueError.
Please can you post the output of the commands

cryosparcm eventlog P99 J199
cryosparcm cli "get_job('P99', 'J199', 'version', 'job_type', 'params_spec', 'status', 'instance_information', 'input_slot_groups')"

where you replace P99, J199 with the failed job’s project and job IDs and omit all but the first and last lines with WARNING: no micrograph found matching image name.