automl / HpBandSter

a distributed Hyperband implementation on Steroids
BSD 3-Clause "New" or "Revised" License
608 stars 109 forks source link

During the execution of the BOHB program, the following error occurred(DEBUG:hpbandster:DISPATCHER: Starting worker discovery), causing the program to hang #69

Open fengzhao860814 opened 5 years ago

fengzhao860814 commented 5 years ago

During the execution of the BOHB program, the following error occurred, causing the program to hang,I want to know why: "DEBUG:hpbandster:DISPATCHER: Starting worker discovery\n", "DEBUG:hpbandster:DISPATCHER: Found 1 potential workers, 1 currently in the pool.\n", "DEBUG:hpbandster:DISPATCHER: Finished worker discovery\n" "DEBUG:hpbandster:DISPATCHER: Starting worker discovery\n", "DEBUG:hpbandster:DISPATCHER: Found 1 potential workers, 1 currently in the pool.\n", "DEBUG:hpbandster:DISPATCHER: Finished worker discovery\n" "DEBUG:hpbandster:DISPATCHER: Starting worker discovery\n", "DEBUG:hpbandster:DISPATCHER: Found 1 potential workers, 1 currently in the pool.\n", "DEBUG:hpbandster:DISPATCHER: Finished worker discovery\n"

BenWilhelmBW commented 4 years ago

Hi, just had the same problem. It turns out that in the worker.compute() function I tried to return a pandas.DataFrame / numpy.ndarray in the 'info' field. This is not allowed. https://automl.github.io/HpBandSter/build/html/core/worker.html I found out by setting the loglevel to DEBUG, then one worker was showing an exception. import logging logging.basicConfig(level=logging.DEBUG) Cheers

yap231995 commented 1 year ago

I tried to use logging.basicConfig(level=logging.DEBUG) Nothing came up and still hang at the same place. What should I do next?