An open API service for providing issue and pull request metadata for open source projects.

GitHub / google-deepmind/scalable_agent issues and pull requests

#58 - How to understand the multi-learners in IMPALA

Issue - State: open - Opened by Davidwhw over 2 years ago

#57 - change file name

Issue - State: open - Opened by masoudganji over 3 years ago

#56 - Reproducing result for Atari game

Issue - State: open - Opened by kimbring2 over 3 years ago

#54 - Questions about atari evaluation protocol

Issue - State: open - Opened by jmkim0309 almost 4 years ago

#53 - Inconsistent reward clipping

Issue - State: open - Opened by yunfanjiang about 4 years ago

#52 - PopArt implementation?

Issue - State: closed - Opened by pengzhenghao over 4 years ago - 2 comments

#51 - global_variable_device

Issue - State: closed - Opened by Pranav-India over 4 years ago

#50 - tf.train.Server to TensorFlow 2 Mirrored Strategy

Issue - State: closed - Opened by Pranav-India over 4 years ago - 1 comment

#49 - Import dynamic_batching giving error

Issue - State: closed - Opened by Pranav-India over 4 years ago - 2 comments

#48 - Multi-task training

Issue - State: closed - Opened by Maxwell2017 over 4 years ago - 1 comment

#47 - Import sonnet is giving the error

Issue - State: closed - Opened by Pranav-India over 4 years ago - 10 comments

#46 - Visualization

Issue - State: closed - Opened by Maxwell2017 over 4 years ago - 6 comments

#45 - Failed to find function dmlab_connect in library

Issue - State: open - Opened by kaustabpal about 5 years ago - 6 comments

#44 - sed: can't read python.BUILD: No such file or directory

Issue - State: closed - Opened by kaustabpal about 5 years ago - 3 comments

#43 - Replay buffer?

Issue - State: open - Opened by xuehy about 5 years ago

#42 - Frequency of actor update

Issue - State: closed - Opened by xuehy over 5 years ago - 2 comments

#41 - AttributeError: PyProcessDmLab object has no attribute b'initial'

Issue - State: open - Opened by alex-petrenko about 6 years ago - 3 comments

#39 - Customize action space and reward structure

Issue - State: closed - Opened by CodeGrinder96 about 6 years ago

#38 - incorporating other environments

Issue - State: closed - Opened by esquires about 6 years ago - 1 comment

#37 - question for scan function in vtrace.py

Issue - State: closed - Opened by chagmgang about 6 years ago - 1 comment

#36 - dynamic batching_test not terminating

Issue - State: closed - Opened by ninji3 about 6 years ago - 1 comment

#35 - problem running the py_process_test

Issue - State: open - Opened by lljgreedyfox over 6 years ago - 2 comments

#34 - Small improvements in the vtrace implementation

Pull Request - State: closed - Opened by heiner over 6 years ago - 1 comment

#33 - Small improvements in the vtrace implementation

Pull Request - State: closed - Opened by heiner over 6 years ago

#32 - Small improvements in the vtrace implementation

Pull Request - State: closed - Opened by heiner over 6 years ago - 1 comment

#31 - The number of CPU in multi-task mode

Issue - State: open - Opened by sasforce over 6 years ago

#30 - Problem running the code snippets in py_process.py

Issue - State: closed - Opened by mingyr over 6 years ago - 3 comments

#29 - Updating Dockerfile (pip install)

Pull Request - State: closed - Opened by Draichi over 6 years ago - 5 comments

#28 - Error running the experiment.py (Sonnet requires tensorflow_probability)

Issue - State: closed - Opened by Draichi over 6 years ago - 4 comments

#26 - Potential discrepancy vis-a-vis the original IMPALA paper

Issue - State: closed - Opened by snailrowen1337 over 6 years ago - 1 comment

#25 - How to extend this implementation

Issue - State: closed - Opened by rnunziata almost 7 years ago - 1 comment

#24 - How to freeze the test / inference graph?

Issue - State: open - Opened by purn3ndu almost 7 years ago

#23 - unable to find a detail explanation of INSTR

Issue - State: closed - Opened by rnunziata almost 7 years ago - 4 comments

#22 - running learner on GPU 0, actors on GPU 1, having trouble placing actors on GPU 1

Issue - State: closed - Opened by mengdong almost 7 years ago - 2 comments

#21 - is there any way to have the game window displayed.

Issue - State: closed - Opened by rnunziata almost 7 years ago - 3 comments

#20 - 'PyProcess' object has no attribute 'name'

Issue - State: closed - Opened by rnunziata almost 7 years ago - 1 comment

#19 - dynamic_batching_test error

Issue - State: closed - Opened by rnunziata almost 7 years ago - 1 comment

#18 - Docker build file to work with nvidia docker 2

Pull Request - State: closed - Opened by mengdong almost 7 years ago - 3 comments

#17 - Replay buffer implementation?

Issue - State: closed - Opened by OhadRubin almost 7 years ago - 1 comment

#16 - unable to install with docker file as given

Issue - State: closed - Opened by rnunziata almost 7 years ago - 3 comments

#15 - Confusing use of bootstrap_value

Issue - State: closed - Opened by Serious-Joker almost 7 years ago - 1 comment

#14 - Reproduction of 80K/sec throughput

Issue - State: open - Opened by pengsun almost 7 years ago - 15 comments

#13 - Performance Question

Issue - State: closed - Opened by jasperzhong almost 7 years ago - 1 comment

#12 - Where in the code do the actors receive new parameters from the Learners?

Issue - State: closed - Opened by jsuit about 7 years ago - 1 comment

#11 - questions about tf.FIFOQueue capacity

Issue - State: closed - Opened by pengsun about 7 years ago - 3 comments

#10 - Update to reflect new versions of TF and DMLab

Pull Request - State: closed - Opened by lespeholt about 7 years ago

#9 - Quick question

Issue - State: closed - Opened by OhadRubin about 7 years ago - 3 comments

#8 - --num_actors in the dist tf example

Issue - State: closed - Opened by pengsun about 7 years ago - 1 comment

#7 - error when running distributed version

Issue - State: closed - Opened by pengsun about 7 years ago - 2 comments

#6 - How to fix the probelm about multiprocesing.Process().start

Issue - State: closed - Opened by cumttang about 7 years ago - 6 comments

#5 - Question about batch_size and capacity of queue

Issue - State: closed - Opened by jsuit about 7 years ago - 7 comments

#4 - Having some issue running experiment.py in test mode.

Issue - State: closed - Opened by rnunziata about 7 years ago - 3 comments

#3 - dynamic_batching_test failure

Issue - State: closed - Opened by rnunziata about 7 years ago - 2 comments

#2 - [Question] Multiple or single host?

Issue - State: closed - Opened by idibidiart about 7 years ago - 5 comments

#1 - Modernize Python 2 code to get ready for Python 3

Pull Request - State: closed - Opened by cclauss about 7 years ago - 1 comment