create benchmark for main algorithms and environments

lucasosouza / fasterRL

Reinforcement Learning library based on pytorch. Designed for research and experiments. Platform-agnostic (supports openai, marlo, more to be added)

MIT License

6 stars 1 forks source link

Open lucasosouza opened 5 years ago

lucasosouza commented 5 years ago

Write a script that does an overall test for all algorithms and record these results to have a comparison table to publish in the library front page

lucasosouza commented 5 years ago

record time, number of episodes and steps for algorithms in major problems