Comments (2)
Each dimension of the transition has a different meaning (e.g., versus
), so has a different linear head. I batch over all of the different linear heads using
einsum
in the EinLinear
module. Because these linear heads are batched together, the number of inputs to each needs to be identical: we can't have N full transitions and one partial transition of only half of a state. To ensure that all transitions are the same dimension, I pad the last one so that it is transition_dim
tokens long.
from trajectory-transformer.
Thanks for the answer! Make sense now.
Interestingly enough, I removed that part from my reimplementation (as i not understand it) and it is still working as normal (tested on imitation learning).
from trajectory-transformer.
Related Issues (20)
- Can't run train script HOT 2
- imitation learning results on HalfCheetah env
- Pretrained Model in AntMaze HOT 3
- ERROR: Could not find a version that satisfies the requirement dm-control HOT 2
- Issue with mc_bin_client.py HOT 4
- About the gym env. HOT 12
- Jax code
- Calculation of value expectation
- long-term prediction
- Novel Dataset Preparation
- Found 0 GPUs for rendering. Using device 0. Device id outside of range of available devices.Failed to initialize OpenGL
- Some questions about hyperparameters in newer version HOT 1
- Reproduce help for AntMaze tasks
- Question about D4RL-gym dataset version HOT 1
- double forward in goal gpt HOT 1
- KeyError: 'halfcheetah-medium-v2' HOT 3
- No registered env with id: halfcheetah-medium-v2 HOT 1
- [Question] Output shape of heads
- Question about the visualisation of four-rooms 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 trajectory-transformer.