search

haotianteng / Chiron

A basecaller for Oxford Nanopore Technologies' sequencers
Other
122 stars 53 forks source link

Unknown exit perhaps related to tensorflow #60

Closed rbutleriii closed 6 years ago

rbutleriii commented 6 years ago

Environment: Fedora23 x64, 32core server w/256GB RAM, no GPU (not a good one, ATI Radeon 5000 series?), on a fresh pyenv python environment -> 2.7.15. Installed via pip: pip install chiron pip install tensorflow pip install -e git+https://github.com/tqdm/tqdm.git@master#egg=tqdm Run exits after signal processing with the following:

Subdirectory processing:: 1it [00:05,  5.42s/it]
2018-06-24 16:33:45.873239: I tensorflow/core/platform/cpu_feature_guard.cc:140] Your CPU supports instructions that this TensorFlow binary was not compiled to use: FMA
chiron_recall/2016-04-04_CJ5:00, ?it/s]
Real time:2.250 Systime:0.213 Usertime:2.229

Run command: chiron call -i /media/Data_2/Sequences/Nanopore/2016_04_04_CJ5-1/uploaded -o chiron_recall/2016-04-04_CJ5 --batch_size 1000 --beam 50

This may be a CPU threading issue is there a switch to specify gpu/cpu? I could not find it in the --help. -t 32 has no effect on error.

haotianteng commented 6 years ago

The time information indicates the running has been finished, and it seems that Chiron has basecalled on an empty folder, which probably be caused by the failure of extraction of raw signal, can you try to run chiron export to see if it can successfully extract the raw signal? I can't reproduce this error in Ubuntu 16.04, so this may be a problem caused by operate system.

rbutleriii commented 6 years ago

When I run chiron export -i /media/Data_2/Sequences/Nanopore/2016_04_04_CJ5-1/uploaded -o chiron_recall/2016-04-04_CJ5 I get that all reads in the folder say: Jeff_THINK_CJ51_20160404_1037_1_ch73_read506_strand.fast5 file transfered. but it only creates an empty train.tfrecords file. Worth noting these are fast5 files from an R7 flow cell. Can Chiron not see the signal in <R9?

haotianteng commented 6 years ago

As far as I know, fast5 files generated by R7 flow cell may have a different entry name of the raw signal, and also I didn't train the Chiron on R7 flow cell configuration, so it won't work on R7 cell.