Kaixhin / Rainbow

Rainbow: Combining Improvements in Deep Reinforcement Learning
MIT License
1.56k stars 282 forks source link

updated loading pretrained model #49

Closed ThisIsIsaac closed 5 years ago

ThisIsIsaac commented 5 years ago

properly raises error when the trained model path is incorrect, prints whether training is starting from scratch or from provided model path.

ThisIsIsaac commented 5 years ago

Applied your suggestions. Thanks for taking a look

Kaixhin commented 5 years ago

Ah sorry for not spotting earlier - we can now do if args.model as a separate conditional, and then if os.path.isfile(args.model) and else nested underneath so we don't have two calls to os.path.isfile.

ThisIsIsaac commented 5 years ago

Np. Done :)