Going to be coding a DQN using Pytorch from as scratch as I can make it. Hope is to record everything, including mistakes, debugging, and the process of solving two different environments (cartpole and breakout) and all details of important metrics. 0:00 Intro, goals 3:15 OpenAI gym discussion and exploration 23:45 Model class 37:00 Replay buffer 55:00 Train step/loss calculation 1:17:30 Debugging the train step 1:29:00 Setting up main for the final training 1:34:00 Sanity check, adding logging 1:43:00 Updating replay buffer to be faster 1:47:00 Epsilon greedy implementation 1:57:00 Training iteration 1 1:59:00 Oops ... fixing loss 2:01:00 Loss being wonky, tuning hyperparams 2:05:00 Training iteration 2 2:13:00 Writing code to test the model MY OTHER VIDOES: ○ A.I. Learns to play Snake https://www.youtube.com/watch?v=i0Pkgtbh1xw ○ Explaining RL to a baby: https://www.youtube.com/watch?v=GFIKtFSl1XQ ○ 5 Common Python Mistakes: https://www.youtube.com/watch?v=fMRzuwlqfzs ○ Learning programming language Julia: https://www.youtube.com/watch?v=TNoShNPoEak ○ Making Python fast: https://www.youtube.com/watch?v=XW32lPcWOxg Twitter: https://twitter.com/safijari Patreon: https://www.patreon.com/jackofsome #pytorch #deeplearning #reinforcementlearning