search

google / balloon-learning-environment

The Balloon Learning Environment - flying stratospheric balloons with deep reinforcement learning.
Apache License 2.0
117 stars 13 forks source link

How to use the ble_wind_field/small dataset in BLE for training a VAE? #24

Open 1jskk opened 3 weeks ago

1jskk commented 3 weeks ago

How to use the ble_wind_field/small dataset in BLE for training a VAE? What is the purpose of this dataset? How should it be used?

I am working with the Balloon Learning Environment (BLE) and I need to use the ble_wind_field/small dataset. I would like to know if this dataset can be used for training a Variational Autoencoder (VAE), and if so, how to write the code to achieve this.

Installed BLE as per the official documentation. I ‘ve already Loaded the dataset using the following code import numpy as np import tensorflow as tf import tensorflow_datasets as tfds

ds = tfds.load('ble_wind_field/small', split='train', shuffle_files=True, data_dir='datasets') assert isinstance(ds, tf.data.Dataset) print(ds)

ds = ds.take(1) # Only take a single example len = 0 for i in tfds.as_numpy(ds): print(i['field'].shape) print(i) len += 1 print('data length:',len)

My question Can the ble_wind_field/small dataset be used for training a VAE? If so, how can I write the code to preprocess this dataset and train a VAE?

Any code snippets or examples would be greatly appreciated. If there are any specific files or configurations needed, please let me know. Thank you for your help!

szuhyz commented 1 week ago

好像generative /learn_wind_field_generator.py文件中提供了如何训练VAE。我对训练此生成模型也非常感兴趣，如果你想进一步交流可以留下你的微信。

1jskk commented 5 days ago

好的，方便的话可以直接添加我的微信：a1206924127