search

yosider / merlin

(Personal experiment) Unsupervised Predictive Memory in a Goal-Directed Agent https://arxiv.org/abs/1803.10760
24 stars 5 forks source link

Simple agents #1

Open pathway opened 6 years ago

pathway commented 6 years ago

Thank you yosider for sharing this very interesting repo!

Would you consider implementing simpler agents "RL-LSTM" and/or "RL-Mem" from the paper? They would make good baselines and are simpler to follow.

yosider commented 6 years ago

I'm so sorry for my late reply... I didn't notice your post.

Thank you very much for your advice! I'm in trouble because MERLIN is too complex to make converged. Trying to the simpler agents seems nice!

I'm busy at a graduate school entrance exam but it will end soon:) I'll try to the implementation after it!