Update data instances and data fields
Browse files
README.md
CHANGED
@@ -536,11 +536,10 @@ The PPO RL agent was trained for 25M steps on each environment and obtained the
|
|
536 |
|
537 |
## Dataset Structure
|
538 |
### Data Instances
|
539 |
-
Each data instance represents a single step consisting of tuples of the form (observation, action, reward, done, truncated) = (o_t, a_t,
|
540 |
|
541 |
```json
|
542 |
{'action': 1,
|
543 |
-
'done': False,
|
544 |
'observation': [[[0, 166, 253],
|
545 |
[0, 174, 255],
|
546 |
[0, 170, 251],
|
@@ -568,15 +567,16 @@ Each data instance represents a single step consisting of tuples of the form (ob
|
|
568 |
[0, 0, 37],
|
569 |
[0, 0, 39]]],
|
570 |
'reward': 0.0,
|
|
|
571 |
'truncated': False}
|
572 |
```
|
573 |
|
574 |
### Data Fields
|
575 |
- `observation`: The current RGB observation from the environment.
|
576 |
- `action`: The action predicted by the agent for the current observation.
|
577 |
-
- `reward`: The received reward
|
578 |
-
- `
|
579 |
-
- `truncated`: If the
|
580 |
|
581 |
### Data Splits
|
582 |
The dataset is divided into a `train` (90%) and `test` (10%) split. Each environment-dataset has in sum 10M steps (data points).
|
|
|
536 |
|
537 |
## Dataset Structure
|
538 |
### Data Instances
|
539 |
+
Each data instance represents a single step consisting of tuples of the form (observation, action, reward, done, truncated) = (o_t, a_t, r_t, terminated_t, truncated_t).
|
540 |
|
541 |
```json
|
542 |
{'action': 1,
|
|
|
543 |
'observation': [[[0, 166, 253],
|
544 |
[0, 174, 255],
|
545 |
[0, 170, 251],
|
|
|
567 |
[0, 0, 37],
|
568 |
[0, 0, 39]]],
|
569 |
'reward': 0.0,
|
570 |
+
'terminated': False,
|
571 |
'truncated': False}
|
572 |
```
|
573 |
|
574 |
### Data Fields
|
575 |
- `observation`: The current RGB observation from the environment.
|
576 |
- `action`: The action predicted by the agent for the current observation.
|
577 |
+
- `reward`: The received reward for the current observation.
|
578 |
+
- `terminated`: If the episode has terminated with the current observation.
|
579 |
+
- `truncated`: If the episode is truncated with the current observation.
|
580 |
|
581 |
### Data Splits
|
582 |
The dataset is divided into a `train` (90%) and `test` (10%) split. Each environment-dataset has in sum 10M steps (data points).
|