R Devon Hjelm

adding in some prints, it does appear that what I did is sending the data correctly to the tf logger, but not sure why it's not writing.

OK, I got it working with this factory: ``` def make_logger(label, steps_key, i): terminal_logger = TerminalLogger(label=label, print_fn=logging.info) tb_logger = TFSummaryLogger(summarydir, label=label, steps_key=steps_key) serialize_fn = base.to_numpy logger = aggregators.Dispatcher([terminal_logger, tb_logger], serialize_fn)...

I'm running into this as well and it's not clear what the fix should be

Problem is the table creation. Should be: ``` replay_buffer = reverb.Table( name=adders.DEFAULT_PRIORITY_TABLE, max_size=1000000, remover=reverb.selectors.Fifo(), sampler=reverb.selectors.Uniform(), rate_limiter=reverb.rate_limiters.MinSize(min_size_to_sample=1), signature=(tf.TensorSpec(2), tf.TensorSpec(1), tf.TensorSpec(None), tf.TensorSpec(None), tf.TensorSpec(2))) ```

The above solution removes the above error, but fails when needing to step through the trainer farther down in the tutorial. These are the correct buffer args I believe: ```...

Great question

There's no information whatsoever on the site where the ROMs come from and it's ambiguous whether Atari can (or will) exercise their IP rights when you download the ROMs.

I would check out Ben Poole's paper, who did a little more analysis as far as the estimating part: http://proceedings.mlr.press/v97/poole19a.html

Apologies, I gave up cortex on PIP a while ago. It can be installed via: https://github.com/rdevon/cortex

Sorry for the delay (busy with other things). Are you still trying to figure out how to do this?