Comments (3)
RuanJohn
commented on June 16, 2024
A link to the branch with this fix.
from mava.
RuanJohn
commented on June 16, 2024
An update on this, the suggested fix uses timestep.first() to reset the hidden states, but since no timestep will ever be a timestep.first() when using Jumanji's auto reset wrapper the hidden states were just never getting reset.
Not ever resetting the hidden states does lead to far better performance for the recurrent systems though so this is a whole other question to consider in itself.
from mava.
sash-a
commented on June 16, 2024
Bug was fixed in #1076
from mava.
Related Issues (20)
- [MAINTAIN] Better equality check for jaxmarl specs
- [INVESTIGATION] Adding the agent's IDs to the global state HOT 1
- [FEATURE] Register for Jumanji environments HOT 1
- [FEATURE] Continuous Actions PPO
- Use `TrainState` in evaluator
- [FEATURE] Generalise Evaluator Fn
- [FEATURE] Move pre-commit hooks to ruff
- [BUG] Crash due to floating point instead of int
- [BUG] Crash if you set num_eval_episodes to specific values due to key splitting
- [MAINTAIN] Orbax API deprecates after May 2024
- [BUG] checkpointing saving using timestep can break due to not being an int HOT 1
- [BUG]: Flasbax bug on LBF (type issue) HOT 1
- [FEATURE]: Reduce config files
- [FEATURE] Integration tests
- [FEATURE] ScannedRNN hidden state initialisation improvement
- [FEATURE] Set time limit per scenario HOT 1
- [BUG] Quickstart Colab imports failing HOT 2
- [BUG] steps_per_second under-reporting in SAC and IQL
- [FEATURE] Support for the Sebulba architecture
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 mava.