Uploading pretrained models

Author: schneimo

.gitignore

@@ -103,3 +103,4 @@ venv.bak/
 # mypy
 .mypy_cache/
 /.idea/
+__pycache__

README.md

@@ -4,7 +4,7 @@ PyTorch implementation of DDPG for continuous control tasks.
 This is a PyTorch implementation of Deep Deterministic Policy Gradients developed in [CONTINUOUS CONTROL WITH DEEP REINFORCEMENT LEARNING](https://arxiv.org/abs/1509.02971).
 
 <p align="center"> 
-<img src="img/roboschool_swingup.gif">
+<img src="_assets/roboschool_swingup.gif">
 </p>
 
 This implementation is inspired by the OpenAI baseline of [DDPG](https://github.com/openai/baselines/tree/master/baselines/ddpg), the newer [TD3](https://github.com/sfujim/TD3) implementation and also various other resources about DDPG. But instead of parameter space noise this implementation uses the original Ornstein-Uhlenbeck noise process of the original DDPG implementation.
@@ -21,7 +21,6 @@ Since 'Roboschool' is deprecated, I highly recommend using [PyBullet](http://pyb
 * TensorBoard
 * TQDM
 * [PyTorch](http://pytorch.org/)
-* [OpenAI baselines](https://github.com/openai/baselines)
 * [OpenAI gym](https://github.com/openai/baselines)
 * [OpenAI Roboschool](https://github.com/openai/baselines)
 

img/roboschool_swingup.gif _assets/roboschool_swingup.gif

@@ -0,0 +1,3 @@
+This folder is needed to save and load models.
+
+If you want to use one of the pretrained models in the subfolder, simply uncompress the *.7z* file to get the needed *.tar* file, which can be loaded by PyTorch. The compression was needed to upload the files on Github, since the models saved by PyTorch are large than 100MB (file size limit of Github).