Datasets:

ArXiv:
License:
File size: 1,107 Bytes
5aacdbb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
# Dataset Card for Pong-v4-expert-MCTS
## Table of Contents
- [Dataset Description](#dataset-description)
- [Dataset Structure](#dataset-structure)
  - [Data Instances](#data-instances)
  - [Data Fields](#data-fields)
  - [Data Splits](#data-splits)
## Dataset Description
This dataset includes 8 episodes of pong-v4 environment. The expert policy is EfficientZero, which is able to generate MCTS hidden states.

## Dataset Structure
### Data Instances
A data point comprises tuples of sequences of (observations, actions, hidden_states):
```
{
    "obs":datasets.Array2D(),
    "actions":datasets.Array2D(),
    "hidden_state":datasets.Array2D(),
}
```
### Data Fields
- `obs`: An Array2D containing observations from 8 trajectories of an evaluated agent.
- `actions`: An Array2D containing actions from 8 trajectories of an evaluated agent.
- `hidden_state`: An Array2D containing corresponding hidden states generated by EfficientZero, from 8 trajectories of an evaluated agent.

### Data Splits
There is only a training set for this dataset, as evaluation is undertaken by interacting with a simulator.