search

jangirrishabh / toyCarIRL

Implementation of Inverse Reinforcement Learning Algorithm on a toy car in a 2D world problem, (Apprenticeship Learning via Inverse Reinforcement Learning Abbeel & Ng, 2004)
MIT License
172 stars 47 forks source link

Error with additional argument in learning.py #4

Open shamanez opened 6 years ago

shamanez commented 6 years ago

Here def train_net(model, params, weights, path, trainFrames, i): why did you use an i?

drex44 commented 6 years ago

I have the same doubt.

jangirrishabh commented 6 years ago

Hello, I was using 'i' as reference to save the model name, but I see that I did it not check for all the function calls for train_net, while some use 'i' others must be modified.

Apologies, I would be updating the code soon, but if you get it working feel free to contribute and send a pull request, cheers!

shamanez commented 6 years ago

Hi other than that your code worked really fine. Nice mate:). BTW did you try GAIL which is a recent inverse reinforcement learning algorithm?

jangirrishabh commented 6 years ago

Thank you @shamanez . Yes I did try out GAIL, and also other form of learning from demonstration called HER from demonstrations, both of them are awesome, I even implemented Overcoming exploration from demonstration in RL, sparse reward case for solving block stacking tasks on a custom made robotic arm simulation, you can read about it here https://jangirrishabh.github.io/2018/03/25/Overcoming-exploration-demos.html

Cheers!

haider0786 commented 3 years ago

Cloning hg http://bitbucket.org/pygame/pygame to /tmp/pip-5beyjs4g-build abort: HTTP Error 404: Not Found