hardmaru comments

Results 39 comments of


                                            hardmaru

Custom Dataset, train.py not creating 'model.ckpt-0'

Hi Ken I'm not familiar with your dataset. Part of your learning experience is to understand the limitations of this model and why it works / doesn't work well with...

Custom Dataset, train.py not creating 'model.ckpt-0'

I would also investigates the average number of line segments per sample in your .svg dataset, and compare it to the typical average number of line segments per typical kanji...

Sketchy logic in vae_train.py:create_dataset

Thanks, probably better to just do the more memory-efficient data loader anyways I guess.

tf_vae.json empty after running vae_train.py

Thanks for the testing, @asolano. Maybe I should just roll back the code to that time...

(under TF 0.7) it looks for saved files in save/save/

Yeah, TF changed how it handled checkpoint paths somewhere between 0.6 to 0.7. It's rather annoying, affecting my other projects too sometimes.

(under TF 0.7) it looks for saved files in save/save/

Thanks! feel free to submit the pull request to put the updated model in too- The old model was trained on one of the very first open source versions of...

Out of memory when running extract.bash due to multiple extract.py using DoomTakeCoverWrapper

Sure, feel free to streamline the code if it is more memory efficient. I'd be happy to accept any PR

full episodes in extracting date

There's no real reason apart from my laziness. You are free to do it in a more efficient way (and I recommend that you do so :)

Got NotImplementedError: Negative indices are currently unsupported

Thanks for the spot, the API must have changed (I made this demo in November 2015 on the very first public release of TensorFlow 0.50). states[-1] just corresponds to the...