An open source AutoML toolkit for automate machine learning lifecycle, including feature engineering, neural architecture search, model compression and hyper-parameter tuning.
Describe the issue:
I have been trying to apply GridSearch strategy to do some NAS experiments, but the tuner seems to crash constantly when starting an experiment. The python script works fine when I run it. When I look into the NNI manager log, it states that the Websocket is not open (readystate 3). So there is probably a problem with the connection, because it is closed by a peer. I already turned off my firewall and the problem still occurs.
Environment:
NNI version: 3.0
Training service (local|remote|pai|aml|etc): local
Describe the issue:
I have been trying to apply GridSearch strategy to do some NAS experiments, but the tuner seems to crash constantly when starting an experiment. The python script works fine when I run it. When I look into the NNI manager log, it states that the Websocket is not open (readystate 3). So there is probably a problem with the connection, because it is closed by a peer. I already turned off my firewall and the problem still occurs.
Environment:
Configuration:
experimentName: FNN_NAS searchSpaceFile: search_space.json trialCommand: python FNN_NAS_training.py trialCodeDirectory: . trialGpuNumber: 0 trialConcurrency: 1 maxExperimentDuration: 5h maxTrialNumber: 10 tuner: name: GridSearch trainingService: platform: local useActiveGpu: FalseLog message:
nnimanager.log: nnimanager.log
dispatcher.log: dispatcher.log