Comments (3)
Also, somehow the memory is never garbage collected for me, and the pre-allocated tensor on the device is never freed.
This is also the case with the rollout collected from the ObservationNormTransform. I suspect there is something wrong with garbage collection of tensordicts, or I'm going something wrong.
This is a bit out of scope for this issue. It was mainly as an argument here. I can open another one with more details about memory leak.
from rl.
Could be, which version of TensorDict are you using?
Can you detail a bit more when the mem should be collected and how you can tell it's not (for us to reproduce)?
from rl.
I will work on the collector devices tomorrow and friday!
For the memory leak feel free to let me know if you can reproduce that in a simple script, it would be very useful to spot where that comes from and fix it!
from rl.
Related Issues (20)
- In the Doc “RECURRENT DQN: TRAINING RECURRENT POLICIES” HOT 1
- [BUG] `DTypeCastTransform` changes arbitrary keys HOT 5
- [BUG] Transpose bug in `reward2go` when the last dim is not 1
- [BUG] VSCode automcompletions don't work HOT 2
- [BUG] `ConvNet` memory issues HOT 5
- [Bug/Question] Target workflow for LSTM Modules?
- [Docs] - Clarify multi collector HOT 3
- [BUG/QUESTION] Installing torchrl with previous torch version HOT 13
- [BUG] rollout with non-uniform tensors no longer supported HOT 6
- [Feature Request]
- [Feature Request] Integration for OpenSpiel environments HOT 1
- [BUG] "Run in Colab" button in tutorials not working
- [BUG] MultiDiscreteTensorSpec .to("mps")
- [Feature Request] DynamicTensorSpec a TensorSpec for sequential data
- [BUG] Done/Reset not initiated when using PettingZooWrapper with Custom PettingZoo environment HOT 6
- [BUG] TruncatedNormal returns wrong values in case of asymmetric bounds HOT 1
- [Feature Request] multiple GPUs on a single machine HOT 2
- [Feature Request] ProbabilisticActor in a Composite Action Space Environment
- [BUG] NonTensorSpec calls NonTensorData with unexpected shape argument HOT 3
- [BUG] NonTensorData disappears from environment after step_mdp() HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from rl.