Skip to content
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

Merge QOL Improvements to Master #2

Closed
wants to merge 137 commits into from
Closed

Merge QOL Improvements to Master #2

wants to merge 137 commits into from

Conversation

blakegood31
Copy link

No description provided.

leathalman and others added 30 commits May 23, 2023 10:57
@leathalman
Updated numpy int to remove deprecation error
b268e67
@leathalman
Enabled score printing, display board, and fixed dimensionality misma…
2c0983b 
…tch when recording loss of neural network
@leathalman
Fixed tensor compatibility issue with numpy and wrong winner in conso…
406cd22 
…le output
@blakegood31
Added code to add board states to text file
a696d7c
@blakegood31
Added code to add game boards to text file
d33df0b
@blakegood31
Added code to add game boards to a text file
8ef8d52
@blakegood31
Finished creating text file with all boards
172658b
@leathalman
Added score to each board state in console and game_history.txt. Incr…
f3a9b4d 
…eased readability of console output.
@blakegood31
Updated Arena.py bar to show proper progress
ff91b16
@blakegood31
Added plot for win rates
364a28d
@leathalman
Added p and v loss graphs for CNN training
e41dd91
@blakegood31
Merge branch 'master' into loss_graphs
61485de
@blakegood31
Merge pull request #1 from leathalman/loss_graphs
dea4ee1 
Add p and v loss graphs for CNN training
@blakegood31
Improved graphing of statistics
f8b9c30
@leathalman
Fixed display enum, Game_History.txt now updates properly on NO_DISPL…
e5a0acd 
…AY toggle, and episodes now print correctly during training on DISPLAY_BOARD toggle
@leathalman
Cleaned up display function
532b7fb
@leathalman
Graphs now update after every iteration
8c9151a
@blakegood31
Updated file system to save new game/training history each time
14cc03f
@blakegood31
Updated to save new Game_History and Training_Results each time
dac78ef
@leathalman
Game_Histories and Training_Results directories will now automaticall…
5a4ff80 
…y be created
@leathalman
Fixed previously introduced bug with save location of csv files
ebb5f1f
@blakegood31
Fixed issue with loading last model/checkpoints
a8a9324
@blakegood31
Merge branch 'master' of https://github.com/leathalman/AZ-Go
2774496
@leathalman
Plots now represent updateThreshold accurately and added debug inform…
9c57c95 
…ation to GoMCTS
@leathalman
Added failsafe for infinite recursion bug
345d90f
@blakegood31
New implementation of PettingZoo. Slightly improved over old implemen…
14b4139 
…tation
@blakegood31
Fixed the infinite recursion error from yesterday, but still dealing …
bbad199 
…with more. Check the updates page on notion for more info.
@leathalman
Updated GoCoach and added board printing
c9bf58c
@leathalman
Fixed bug where valids were not updated to action_mask from PettingZoo
e478dbf
@leathalman
First stable PettingZoo release and code cleanup
39ca041
leathalman and others added 27 commits September 17, 2023 19:47
@leathalman
Fixed bug in engine code and updated temp in config.yaml
8cfc19e
@leathalman
Added dynamic scaling for maximum number of moves allowed in a game
f268416
@leathalman
Fixed incorrect file deletion
149f55e
@leathalman
First stable refactor
5a57016
@blakegood31
Refactored learn() and executeEpisode() in coach.py
e212773
@blakegood31
Added playout cap randomization to self play games
e87b5fa
@leathalman
Added bug fixes for incorrect x-axis on graphs, broken yaml configs, …
5021f42 
…status bar float division error, and fixed naming for arena log/training graphs
@leathalman
Fixed pathing for engine code
9ba2b5e
@leathalman
Fixed engine import path for model.tar
1181082
@leathalman
Updated pathing for distributed_worker
0a35bb3
@leathalman
Updated parameters and number of games on first iteration of distribu…
8d18af3 
…ted training
@blakegood31
Arena play now returns after threshold is met by one of the players
6d40f43
@blakegood31
Added cosine annealing functionality
8351d76
@leathalman
Added support for multiple concurrent worker machines
602acb5
@leathalman
Merge branch 'modified_arena'
8188f71
@leathalman
Restored config.yaml
98804b2
@leathalman
Fixed error where distributed/examples directory was not created by s…
5b0d10c 
…tart_worker script
@blakegood31
Implemented fix for pickler, fixed arena statistics, and unmasked mov…
e779441 
…e probabilities for MCTS
@leathalman
Merge remote-tracking branch 'origin/pickler_fix'
24d30b1
@blakegood31
First attempt at proper implementation of queue length
04284a9
@blakegood31
Added/updated code to use sensibility layer
5c3a83c
@leathalman
Remasked pi values for training, config set to Model Q levels
c6e9a41
@leathalman
Updated pathing for loading in models
378ebce
@blakegood31
Fixed load_model() in coach to actually load model
34d782b
@blakegood31
Modified how acceptance threshold is calculated for new models
0648a00
@blakegood31
Added updated engine code
36418cc
@blakegood31
Updated loss function and added loadsgf command to engine
ea5a7a0
@blakegood31 blakegood31 closed this Apr 5, 2024
@blakegood31
Copy link
Author

Sorry, Accidental Pull Request

@blakegood31 blakegood31 deleted the qol_improvements branch April 5, 2024 21:21
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@blakegood31 @leathalman