New answers tagged

0

I've been working with a TD3 implementation for bipedal hardcore. It solved the easy version (v2 and v3) in about 300 epochs (https://github.com/QasimWani/policy-value-methods). I've been training it for hardcore and even after about 1200 episodes, it's no where close to convergence. Did you end up solving, and if so, what algorithm did you end up going with?...


2

You could sequentially pass in each element of your sequential data and save the hidden and cell states in a separate buffer. In a typical LSTM implementation, you input the entire sequence and the hidden and cell states are propagated internally. In the end, the final hidden and cell states returned as the output. This works if your input is all the same ...


0

The most important thing for you is to actually have fun while coding. There are a lot of courses(it's overwhelming to me sometimes), a lot of books, blogs and other stuff on this topic. When it comes to the learning, find a person you enjoy learning from(for you that might be Andrew NG, for me it's say Daniel Bourke), and slowly start working on your own ...


0

Ai is moving very fast. Nobody is really sure what's the best approach. As of today, consider to get involved in Ai communities. Kaggle is a very good place to start, see others' notebooks and thoughts, and consider some easy to understand blogs in data science and Ai like medium. Also don't worry about learning deep learning in TF1. The main difference ...


0

Get a raspberry with a camera, power it by a battery bank, attach it to the EV3 and run a 2 one program on raspberry and another on ev3 communicating with each other via MQTT


0

You can increase no of hidden layers. Following is an example (But not very efficient)


Top 50 recent answers are included