Examples: query, "exact match", wildcard*, wild?ard, wild*rd
Fuzzy search: cake~ (finds cakes, bake)
Term boost: "red velvet"^4, chocolate^2
Field grouping: tags:(+work -"fun-stuff")
Escaping: Escape characters +-&|!(){}[]^"~*?:\ with \, e.g. \+
Range search: properties.timestamp:[1587729413488 TO *] (inclusive), properties.title:{A TO Z}(excluding A and Z)
Combinations: chocolate AND vanilla, chocolate OR vanilla, (chocolate OR vanilla) NOT "vanilla pudding"
Field search: properties.title:"The Title" AND text
Answered
Hi, Recently Came Across Trains And Very Impressed By The Work So Far. But A Problem Has Been Bugging Me, This Is Part Of The Trains Log Files I Thought Might Be Useful From Cloning And Enqueuing The Same Task On 2 Remote Machines. The First Machine Defau

Hi, recently came across trains and very impressed by the work so far. But a problem has been bugging me, this is part of the trains log files I thought might be useful from cloning and enqueuing the same task on 2 remote machines. The first machine defaults to python2 for some reason which causes a failure
working_dir = . New python executable in /home/sm2/.trains/venvs-builds/3.6/bin/python2 Also creating executable in /home/sm2/.trains/venvs-builds/3.6/bin/python Installing setuptools, pkg_resources, pip, wheel...done. Running virtualenv with interpreter /usr/bin/python2
working_dir = . Executing Conda: /home/hbku1/miniconda3/condabin/conda env remove -p /home/hbku1/.trains/venvs-builds/3.6 --quiet --json Remove all packages in environment /home/hbku1/.trains/venvs-builds/3.6: Executing Conda: /home/hbku1/miniconda3/condabin/conda install -p /home/hbku1/.trains/venvs-builds/3.6 -c defaults -c conda-forge -c pytorch cudatoolkit=10.0 --quiet --jsonThe machines aren't identical but both are running miniconda. I understood just the info above may not be enough but let me know if there's something I should look into.

  
  
Posted 3 years ago
Votes Newest

Answers 7


Not sure why it defaulted to python2 in base env. Thanks

  
  
Posted 3 years ago

will check it out

  
  
Posted 3 years ago

AntsySeagull45 kudos on sorting it out 🙂
quick note, trains-agent will try to run the python version specified by the original Task. i.e. if you were running python3.7 it will first try to look for python 3.7 then if it is not there it will run the default python3. This allows a system with multiple python versions to run exactly the python version you had on your original machine. The fact that it was trying to run python2 is quite odd, one explanation I can think of is if the original experiment was run on python2.7...
The other explanation is it default to the python version that runs the trains-agent! So if you managed to run the trains-agent with python2 this will be the version it will default to.
You can always verify the default python version, when running the trains-agent, it print the configuration, and you should see something like:
agent.default_python = 3.6

  
  
Posted 3 years ago

One more question, in the second log, trains agent is configured with Conda, on the first it is configured with pip, or at least this is what it looks like, can you confirm?

  
  
Posted 3 years ago

No, the original code is in python 3. First machine is 0.15.1 and the second machine which used conda is 0.15.0. I didn't manually make change in configuration.

  
  
Posted 3 years ago

fixed now, trains-agent was running in conda base environment in the problematic machine. Created a new environment and ran it inside that and the code is working.

  
  
Posted 3 years ago

Hi AntsySeagull45
Any chance the original code was running with python2?
Which version of trains-agent are you using?

  
  
Posted 3 years ago
491 Views
7 Answers
3 years ago
one year ago
Tags