typoverflow/UtilsRL issues and pull requests

#42 - [Feature]: forced argument type during arg parse

Issue - State: open - Opened by typoverflow about 1 year ago
Labels: enhancement

#41 - delete: remove redudant features

Pull Request - State: closed - Opened by typoverflow about 1 year ago

#40 - refactor: loggers

Pull Request - State: closed - Opened by typoverflow about 1 year ago

#39 - feat: DMControl Wrappers

Pull Request - State: closed - Opened by typoverflow over 1 year ago

#38 - update dmc wrapper

Pull Request - State: closed - Opened by LyndonKong over 1 year ago

#37 - feat: Implement RWKV-v4

Pull Request - State: closed - Opened by typoverflow over 1 year ago

#36 - feat: log histogram for tb logger

Pull Request - State: closed - Opened by typoverflow over 1 year ago

#35 - feat: Transformer Architecture for RL

Pull Request - State: closed - Opened by typoverflow over 1 year ago

#34 - fix: DT input token was not layer normed

Pull Request - State: closed - Opened by typoverflow over 1 year ago

#33 - Add dmc control env wrapper

Pull Request - State: closed - Opened by LyndonKong over 1 year ago

#32 - dropout after the attention

Issue - State: closed - Opened by Howuhh over 1 year ago - 2 comments
Labels: bug, question

#31 - Fix PrioritizedSimpleReplay batch_update bug

Pull Request - State: closed - Opened by LyndonKong over 1 year ago

#30 - update rwkv

Pull Request - State: closed - Opened by LyndonKong over 1 year ago

#29 - Fix transition_replay_buffer unsqueeze

Pull Request - State: closed - Opened by LyndonKong over 1 year ago

#28 - [Bug]: using TanhNormal may result in too complex gradient computing

Issue - State: closed - Opened by typoverflow over 1 year ago - 1 comment
Labels: bug

#27 - [feature] Enable profile decorator to be context manager

Pull Request - State: closed - Opened by LyndonKong over 1 year ago

#26 - refactor NameSpace

Pull Request - State: closed - Opened by typoverflow almost 2 years ago

#25 - [Bug]: NameSpace should support dict and argparse.Namespace features

Issue - State: closed - Opened by typoverflow almost 2 years ago - 1 comment
Labels: bug

#24 - [Feature]: WandB Logger

Pull Request - State: closed - Opened by typoverflow almost 2 years ago - 1 comment

#23 - Update: Implementations of trajectory replays

Pull Request - State: closed - Opened by mansicer almost 2 years ago

#22 - refactor: RL buffer

Pull Request - State: closed - Opened by typoverflow almost 2 years ago - 1 comment

#21 - [Feature]: bridging module for backends and frontends

Issue - State: open - Opened by typoverflow almost 2 years ago
Labels: enhancement

#20 - [Update]: Fix argparse logic for float precision; fix interface error for ensemble linear

Pull Request - State: closed - Opened by typoverflow almost 2 years ago - 1 comment

#19 - [Feature]: Wanb logger

Issue - State: closed - Opened by typoverflow almost 2 years ago - 2 comments
Labels: enhancement

#18 - [Feature]: Add FP16 and unify it with custom precision logic

Issue - State: closed - Opened by typoverflow almost 2 years ago - 1 comment
Labels: enhancement

#17 - Fix: return proper value for `random_batch` method of `ReplayPool`

Pull Request - State: closed - Opened by mansicer about 2 years ago

#16 - [Feature]: SumTree implemented in c++

Pull Request - State: closed - Opened by typoverflow over 2 years ago - 1 comment

#15 - [Feature]: high-performance buffer or data structure implementations

Issue - State: closed - Opened by typoverflow over 2 years ago - 2 comments
Labels: enhancement

#14 - [Feature]: ensemble mlp network and transformer blocks

Pull Request - State: closed - Opened by typoverflow over 2 years ago - 2 comments

#13 - [Feature]: ensemble model

Issue - State: closed - Opened by typoverflow over 2 years ago - 2 comments
Labels: enhancement

#12 - [Update]: use register_buffer for no grad parameters

Pull Request - State: closed - Opened by typoverflow over 2 years ago

#11 - [Bug]: register_buffer should be used rather than registering a no_grad parameter

Issue - State: closed - Opened by typoverflow over 2 years ago - 3 comments
Labels: bug

#10 - [Feature]: Support for custom float precision

Pull Request - State: closed - Opened by typoverflow over 2 years ago

#9 - [Feature]: gym env wrappers

Issue - State: closed - Opened by typoverflow over 2 years ago
Labels: enhancement

#8 - [Feature]: Light-weight conversion between np.ndarray and torch.tensor

Issue - State: closed - Opened by typoverflow over 2 years ago - 1 comment
Labels: enhancement

#7 - [Feature]: Implement prioritized experience replay buffer

Issue - State: closed - Opened by typoverflow over 2 years ago - 2 comments
Labels: enhancement

#6 - [Feature]: Full attention block support

Issue - State: open - Opened by typoverflow over 2 years ago
Labels: enhancement

#5 - [Feature]: Full image input support

Issue - State: open - Opened by typoverflow over 2 years ago
Labels: enhancement

#4 - [Feature]: Full support for RNN backend

Issue - State: open - Opened by typoverflow over 2 years ago
Labels: enhancement

#3 - [Feature]: Manual iteration of UtilsRL.monitor.Monitor

Issue - State: closed - Opened by typoverflow over 2 years ago - 1 comment
Labels: enhancement

#2 - [Feature]: Choice of float number precision

Issue - State: closed - Opened by typoverflow over 2 years ago - 1 comment
Labels: enhancement

#1 - [BUG]: NameSpace adds arguments to classdicts

Issue - State: closed - Opened by typoverflow over 2 years ago

GitHub / typoverflow/UtilsRL issues and pull requests