Can I ask you some error messages related to Convnet based auto-boxing and tensorflow?

[Si Hoon Park]

Dear EMAN2 community,

I got a  problem in Convnet based auto-boxing.

After training with several iteration, the Convnet based auto-boxing windows were suddenly closed with some error message.

Can I get some help?

[eman@lsb02 SUV2_notcor_EMAN2]$ conda activate

(base) [eman@lsb02 SUV2_notcor_EMAN2]$ e2version.py
EMAN 2.91 final ( GITHUB: 2021-03-08 11:36 - commit: 81caed2 )
Your Python version is: 3.7.7
(base) [eman@lsb02 SUV2_notcor_EMAN2]$ python -c "import tensorflow"
(base) [eman@lsb02 SUV2_notcor_EMAN2]$ e2projectmanager.py
QStandardPaths: wrong ownership on runtime directory /run/user/1000, 1000 instead of 1004
NOT Writing notes, ppid=-2
QStandardPaths: wrong ownership on runtime directory /run/user/1000, 1000 instead of 1004
Reading tomograms/NC_2_ali__bin4.hdf...
New network..
Preparing training set...
Training...
WARNING:tensorflow:AutoGraph could not transform <function train_on_batch at 0x2ad0517b8830> and will run it as-is.
Please report this to the TensorFlow team. When filing the bug, set the verbosity to 10 (on Linux, `export AUTOGRAPH_VERBOSITY=10`) and attach the full output.
Cause: No module named 'tensorflow_core.estimator'
iteration 0, cost 0.110
iteration 1, cost 0.110
iteration 2, cost 0.107
iteration 3, cost 0.106
iteration 4, cost 0.102
iteration 5, cost 0.098
iteration 6, cost 0.094
iteration 7, cost 0.091
iteration 8, cost 0.088
iteration 9, cost 0.083
iteration 10, cost 0.069
iteration 11, cost 0.052
iteration 12, cost 0.042
iteration 13, cost 0.039
iteration 14, cost 0.033
iteration 15, cost 0.033
iteration 16, cost 0.030
iteration 17, cost 0.028
iteration 18, cost 0.027
iteration 19, cost 0.026
Generating output...
(100, 96, 96, 3)
Traceback (most recent call last):
  File "/home/eman/eman2-sphire-sparx/bin/e2spt_boxer_convnet.py", line 712, in train_nnet
    m.write_image(fname, -1)
  File "/home/eman/eman2-sphire-sparx/lib/python3.7/site-packages/EMAN2db.py", line 540, in db_write_image
    return self.write_image_c(fsp, *parms)
RuntimeError: FileAccessException at /home/eman2/miniconda3/conda-bld/eman2_1615224802990/work/libEM/io/hdfio2.cpp:471: error with 'neuralnets/trainouts.hdf': 'cannot access file 'neuralnets/trainouts.hdf'' caught

Sincerely

Si Hoon Park

[Muyuan Chen]

Is there a folder called “neuralnets” in the project folder? If not, try making one and run it again. Although I thought the program shall create it automatically..

--
--
----------------------------------------------------------------------------------------------
You received this message because you are subscribed to the Google
Groups "EMAN2" group.
To post to this group, send email to em...@googlegroups.com
To unsubscribe from this group, send email to eman2+un...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/eman2

---
You received this message because you are subscribed to the Google Groups "EMAN2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to eman2+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/1e3dd6e6-5ab8-4bfb-b626-080322d7495bn%40googlegroups.com.

[Si Hoon Park]

Dear Muyuan,

Yes I already had “neuralnets” in the project folder.

Can I ask you this error could be related with library?

I don't have home/eman2 folder. It was deleted this user today morning and add user 'eman' and reinstalled eman2.

RuntimeError: FileAccessException at /home/eman2/miniconda3/conda-bld/eman2_1615224802990/work/libEM/io/hdfio2.cpp:471: error with 'neuralnets/trainouts.hdf': 'cannot access file 'neuralnets/trainouts.hdf'' caught

Best regards,

Si Hoon Park

I

2021년 10월 20일 수요일 오전 11시 18분 8초 UTC+9에 g5v...@gmail.com님이 작성:

[Muyuan Chen]

I don't think this is library related. More likely file writing issues. Can you run something like "e2proc3d.py :128:128:128:0 neuralnets/test.hdf" and see if it works? 

To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/fe4e51b0-705f-4899-9d83-a3815a413a90n%40googlegroups.com.

[박시훈]

Dear Muyuan,

Yes, there was something wrong at this point;

(base) [eman@lsb02 SUV2_notcor_EMAN2]$ e2proc3d.py :128:128:128:0 neuralnets/trainout_nnet_save__bin4_preproc_test64.hdf

Traceback (most recent call last):

  File "/home/eman/eman2-sphire-sparx/bin/e2proc3d.py", line 890, in <module>
    main()
  File "/home/eman/eman2-sphire-sparx/bin/e2proc3d.py", line 823, in main
    data.write_image(outfile, img_index, EMUtil.get_image_ext_type(options.outtype), False, None, file_mode_map[options.outmode], not(options.swap))

  File "/home/eman/eman2-sphire-sparx/lib/python3.7/site-packages/EMAN2db.py", line 540, in db_write_image
    return self.write_image_c(fsp, *parms)

RuntimeError: FileAccessException at /home/eman2/miniconda3/conda-bld/eman2_1615224802990/work/libEM/io/hdfio2.cpp:471: error with 'neuralnets/trainout_nnet_save__bin4_preproc_test64.hdf': 'cannot access file 'neuralnets/trainout_nnet_save__bin4_preproc_test64.hdf'' caught

Best wishes,

Si Hoon Park

2021년 10월 20일 (수) 오전 11:46, Muyuan Chen <g5v...@gmail.com>님이 작성:

To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/CAO_0xqEjAOBdenf5aPQYGMYuHsf20GV_YyEBrDkrhX%3DzSiKM2A%40mail.gmail.com.

[Muyuan Chen]

Do you have writing permission in the folder and do you have enough disk space?

To view this discussion on the web visit https://groups.google.com/d/msgid/eman2/CALP%3Dy6LECvs%3D5Ss11KurSu-SnMkzU4Tr5mMQwzo%2B37HjJFvUUQ%40mail.gmail.com.

[Si Hoon Park]

Dear Muyuan,

Yes! Thanks for your solution! The problem was originated from the writing permission, so I change the permission of the folder and sub_directory and it worked nicely!

Best regards,

Si Hoon Park

2021년 10월 20일 수요일 오후 12시 5분 47초 UTC+9에 g5v...@gmail.com님이 작성: