Update README.md

README.md

@@ -3,78 +3,86 @@ license: apache-2.0
 ---
 # 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)
-- [Data Creation](#Data-Creation)
-  - [Curation Rationale](##Curation-Rationale)
-  - [Source Data](##Source-Data)
-    - [Initial Data Collection and Normalization](###Initial-Data-Collection-and-Normalization)
-    - [Who are the source data producers?](### Who-are-the-source-data-producers?)
-    - [Annotations](###Annotations)
-- [Considerations for Using the Data](#Considerations-for-Using-the-Data)
+
+- [Supported Tasks and Baseline](#support-tasks-and-baseline)
+
+- [Data Usage](#data-usage)
+
+  - [Data Discription](##data-description)
+  - [Data Fields](##data-fields)
+  - [Data Splits](##data-splits)
+  - [Initial Data Collection and Normalization](##Initial-Data-Collection-and-Normalization)
+
+- [Additional Information](#Additional-Information)
+
+  - [Who are the source data producers?](## Who-are-the-source-data-producers?)
   - [Social Impact of Dataset](##Social-Impact-of-Dataset)
   - [Known Limitations](##Known-Limitations)
-- [Additional Information](#Additional-Information)
+
   - [Licensing Information](##Licensing-Information)
   - [Citation Information](##Citation-Information)
   - [Contributions](##Contributions)
 
 ## Supported Tasks and Baseline
 
-- This dataset supports the training for Procedure Cloning algorithm.
+- This dataset supports the training for [Procedure Cloning (PC )](https://arxiv.org/abs/2205.10816) algorithm.
+- Baselines when sequence length for decision is 0:
 
-- Baseline
+| Train loss                                         | Test Acc | Reward |
+| -------------------------------------------------- | -------- | ------ |
+| <img src="./img/sup_loss.png" style="zoom:50%;" /> | 0.90     | 20     |
 
-  | Length for procedure sequence | Return |
-  | ----------------------------- | ------ |
-  | 0                             | 20     |
-  | 4                             | -21    |
+- Baselines when sequence length for decision is 4:
 
-## Dataset Description
+| Train action loss                                     | Train hidden state loss                           | Train acc (auto-regressive mode)                    | Reward |
+| ----------------------------------------------------- | ------------------------------------------------- | --------------------------------------------------- | ------ |
+| <img src="./img/action_loss.png" style="zoom:50%;" /> | <img src="./img/hs_loss.png" style="zoom:50%;" /> | <img src="./img/train_acc.png" style="zoom:50%;" /> | -21    |
 
-This dataset includes 8 episodes of pong-v4 environment. The expert policy is EfficientZero, which is able to generate MCTS hidden states.
+## Data Usage
 
-## Dataset Structure
-### Data Instances
-A data point comprises tuples of sequences of (observations, actions, hidden_states):
-```
-{
-    "obs":datasets.Array3D(),
-    "actions":int,
-    "hidden_state":datasets.Array3D(),
-}
-```
-## Source Data
+### Data description
+
+This dataset includes 8 episodes of pong-v4 environment. The expert policy is [EfficientZero]([[2111.00210\] Mastering Atari Games with Limited Data (arxiv.org)](https://arxiv.org/abs/2111.00210)), which is able to generate MCTS hidden states. Because of the contained hidden states for each observation, this dataset is suitable for Imitation Learning methods that learn from a sequence like PC.
 
 ### Data Fields
 
-- `obs`: An Array3D containing observations from 8 trajectories of an evaluated agent. The data type is uint8 and each value is in 0 to 255.
+- `obs`: An Array3D containing observations from 8 trajectories of an evaluated agent. The data type is uint8 and each value is in 0 to 255. The shape of this tensor is [96, 96, 3], that is, the channel dimension in the last dimension. 
 - `actions`: An integer containing actions from 8 trajectories of an evaluated agent. This value is from 0 to 5. Details about this environment can be viewed at [Pong - Gym Documentation](https://www.gymlibrary.dev/environments/atari/pong/).
 - `hidden_state`: An Array3D containing corresponding hidden states generated by EfficientZero, from 8 trajectories of an evaluated agent. The data type is float32.
 
+This is an example for loading the data using iterator:
+
+```python
+from safetensors import saveopen
+
+def generate_examples(self, filepath):
+    data = {}
+    with safe_open(filepath, framework="pt", device="cpu") as f:
+        for key in f.keys():
+            data[key] = f.get_tensor(key)
+        
+    for idx in range(len(data['obs'])):
+        yield idx, {
+            'observation': data['obs'][idx],
+            'action': data['actions'][idx],
+            'hidden_state': data['hidden_state'][idx],
+        }
+```
+
 ### Data Splits
 There is only a training set for this dataset, as evaluation is undertaken by interacting with a simulator.
 
-## Data Creation
-
-### Curation Rationale
+### Initial Data Collection and Normalization
 
-- This dataset includes expert data generated by EfficientZero. Since it contains hidden states for each observation, it is suitable for Imitation Learning methods that learn from a sequence like [Procedure Cloning (PC)](https://arxiv.org/abs/2205.10816).
-### Source Data
-#### Initial Data Collection and Normalization
 - This dataset is collected by EfficientZero policy.
 - The standard for expert data is that each return of 8 episodes is over 20.
 - No normalization is previously applied ( i.e. each value of observation is a uint8 scalar in [0, 255] )
-#### Who are the source language producers?
-- [@kxzxvbk](https://huggingface.co/kxzxvbk)
 
-#### Annotations
-- The format of observation picture is [H, W, C], where the channel dimension is the last dimension of the tensor.
+## Additional Information
+
+### Who are the source language producers?
 
-## Considerations for Using the Data
+[@kxzxvbk](https://huggingface.co/kxzxvbk)
 
 ### Social Impact of Dataset
 
@@ -83,11 +91,10 @@ There is only a training set for this dataset, as evaluation is undertaken by in
 - This dataset can potentially promote the research for sequence based imitation learning algorithms.
 
 ### Known Limitations
+
 - This dataset is only used for academic research.
 - For any commercial use or other cooperation, please contact: opendilab@pjlab.org.cn
 
-## Additional Information
-
 ### License
 This dataset is under [Apache License 2.0](https://www.apache.org/licenses/LICENSE-2.0).