Training loop terminates after few episodes
See original GitHub issueHigh-Level Description Training loop termination in UTLRA
Desired SMARTS version 0.4.16
Operating System ubuntu 18.04
Problems while training for multi-agents in ULTRA, sometimes the loop terminate after few episodes, giving the following error:
Exception ignored in: <function _ChannelCallState.__del__ at 0x7f0824052b00>
Traceback (most recent call last):
File "/home/dikshant/github/SMARTS/ultra/.VENV/lib/python3.7/site-packages/grpc/_channel.py", line 1126, in __del__
AttributeError: 'NoneType' object has no attribute 'cancelled'
Exception ignored in: <function _ChannelCallState.__del__ at 0x7f0824052b00>
a somewhat related issue is mentioned on the official github repo of grpc as well: https://github.com/grpc/grpc/issues/23290
Issue Analytics
- State:
- Created 2 years ago
- Comments:5 (3 by maintainers)
Top Results From Across the Web
Train Reinforcement Learning Policy Using Custom Training ...
Terminate the training after the maximum number of episodes is reached or when the average reward across 100 episodes reaches the value of...
Read more >What is the reason we loop over epochs when training a ...
According to the loss plotted over epochs, the loss is decreasing after each iteration, so it seems like the parameters are not chosen...
Read more >not sure if tensorflow is running on GPU or CPU custom keras ...
the python script is using 20% of my CPU and only 0.3% of my GPU. I am still very new to Keras and...
Read more >Understanding Learning Rates and How It Improves ...
On the right the difference is cut in half after each cycle. ... [5] Estimating an Optimal Learning Rate for a Deep Neural...
Read more >Q-Learning
This sequential update process continues until a goal state is reached or a maximum number of steps is taken. The current episode then...
Read more >Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start FreeTop Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
Top GitHub Comments
This error is a result of the changes in #704 and occurs because the SMARTS version that we currently use in ULTRA has a GRPC version with this error. When ULTRA updates its SMARTS dependency to a newer version that has a newer GRPC dependency, this error will not appear anymore.
oh yeah, I just saw that it is due to only 10 training episodes and when they are over, it shows this error. Is there any way to remove this error (maybe it can be a later fix you want to do).
thanks, @christianjans, closing this issue for time being but feel free to add more comments if you have sth in mind.