Issue #348: len is not well defined for symbolic Tensors #351

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Open

rckitson wants to merge 1 commit into keras-rl:master from rckitson:bugfix/dqn-cartpole

rckitson commented Jan 20, 2020

This only addresses the first half of Issue #348 , but was able to run examples of DQN and DDPG with:

tensorflow 1.15.0
keras 2.3.1


          Issue keras-rl#348: len is not well defined for symbolic Tensors

fe08bf0

LyWangPX mentioned this pull request

Hotfix/not well defined shape #353

Open

gosuto-inzasheru approved these changes

View reviewed changes

msat59 commented Apr 18, 2020

This only addresses the first half of Issue #348 , but was able to run examples of DQN and DDPG with:

tensorflow 1.15.0

keras 2.3.1

I faced the same error with tensorflow 1.15.0. I fixed manually the dqn.py line 108:

original code:
if hasattr(model.output, '__len__') and len(model.output) > 1:

changed to
108 if hasattr(model.output, '__len__') and model.output.shape[0] > 1:

msat59 suggested changes

View reviewed changes

msat59 left a comment

Please change model.shape[0] to model.output.shape[0]; otherwise, python says:

{AttributeError}'Sequential' object has no attribute 'shape'

dave7895 commented May 14, 2020

With this fix I get the error TypeError: '>' not supported between instances of 'NoneType' and 'int'.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet