New default net nn-33c9d39e5eb6.nnue #3557
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
as the previous net, trained on Leela games, as provided by borg. See also https://lczero.org/blog/2021/06/the-importance-of-open-data/ The particular data set, which is a mix of T60 and T74 data, is now available as a single binpack: https://drive.google.com/file/d/1RFkQES3DpsiJqsOtUshENtzPfFgUmEff/view?usp=sharing The training command was: python train.py ../../training_data_pylon.binpack ../../training_data_pylon.binpack --gpus 1 --threads 2 --num-workers 2 --batch-size 16384 --progress_bar_refresh_rate 300 --smart-fen-skipping --random-fen-skipping 10 --features=HalfKAv2^ --lambda=1.0 --max_epochs=440 --seed $RANDOM --default_root_dir exp/run_2 passed STC: https://tests.stockfishchess.org/tests/view/60c887cb457376eb8bcab054 LLR: 2.94 (-2.94,2.94) <-0.50,2.50> Total: 12792 W: 1483 L: 1311 D: 9998 Ptnml(0-2): 62, 989, 4131, 1143, 71 passed LTC: https://tests.stockfishchess.org/tests/view/60c8e5c4457376eb8bcab0f0 LLR: 2.95 (-2.94,2.94) <0.50,3.50> Total: 11272 W: 601 L: 477 D: 10194 Ptnml(0-2): 9, 421, 4657, 535, 14 also had strong LTC performance against another strong net of the series: https://tests.stockfishchess.org/tests/view/60c8c40d457376eb8bcab0c6 Bench: 5032320
|
Congratulations! Excited to see if there's more elo in that data set. For now I concentrate on pure T74 data. |
|
thanks, yes, I want to explore a bit as well, T60, T74, your mix, and your mix shuffled. Let's see what is the magic ingredient. |
|
merged via 8ec9e10, congrats :-) Now we have to re-test all the other pull requests which passed Elo gaining bounds on the previous net? |
|
yes, I think it would be wise to retest #3556 also because it needs some additional code to deal correctly with the 'adjusted' flag. |
MichaelB7
pushed a commit
to MichaelB7/Stockfish
that referenced
this pull request
Jul 6, 2021
As the previous net, this net is trained on Leela games as provided by borg. See also https://lczero.org/blog/2021/06/the-importance-of-open-data/ The particular data set, which is a mix of T60 and T74 data, is now available as a single binpack: https://drive.google.com/file/d/1RFkQES3DpsiJqsOtUshENtzPfFgUmEff/view?usp=sharing The training command was: python train.py ../../training_data_pylon.binpack ../../training_data_pylon.binpack --gpus 1 --threads 2 --num-workers 2 --batch-size 16384 --progress_bar_refresh_rate 300 --smart-fen-skipping --random-fen-skipping 10 --features=HalfKAv2^ --lambda=1.0 --max_epochs=440 --seed $RANDOM --default_root_dir exp/run_2 passed STC: https://tests.stockfishchess.org/tests/view/60c887cb457376eb8bcab054 LLR: 2.94 (-2.94,2.94) <-0.50,2.50> Total: 12792 W: 1483 L: 1311 D: 9998 Ptnml(0-2): 62, 989, 4131, 1143, 71 passed LTC: https://tests.stockfishchess.org/tests/view/60c8e5c4457376eb8bcab0f0 LLR: 2.95 (-2.94,2.94) <0.50,3.50> Total: 11272 W: 601 L: 477 D: 10194 Ptnml(0-2): 9, 421, 4657, 535, 14 also had strong LTC performance against another strong net of the series: https://tests.stockfishchess.org/tests/view/60c8c40d457376eb8bcab0c6 closes official-stockfish#3557 Bench: 5032320
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
as the previous net, trained on Leela games, as provided by borg.
See also https://lczero.org/blog/2021/06/the-importance-of-open-data/
The particular data set, which is a mix of T60 and T74 data, is now available as a single binpack:
https://drive.google.com/file/d/1RFkQES3DpsiJqsOtUshENtzPfFgUmEff/view?usp=sharing
The training command was:
python train.py ../../training_data_pylon.binpack ../../training_data_pylon.binpack --gpus 1 --threads 2 --num-workers 2 --batch-size 16384 --progress_bar_refresh_rate 300 --smart-fen-skipping --random-fen-skipping 10 --features=HalfKAv2^ --lambda=1.0 --max_epochs=440 --seed $RANDOM --default_root_dir exp/run_2
passed STC:
https://tests.stockfishchess.org/tests/view/60c887cb457376eb8bcab054
LLR: 2.94 (-2.94,2.94) <-0.50,2.50>
Total: 12792 W: 1483 L: 1311 D: 9998
Ptnml(0-2): 62, 989, 4131, 1143, 71
passed LTC:
https://tests.stockfishchess.org/tests/view/60c8e5c4457376eb8bcab0f0
LLR: 2.95 (-2.94,2.94) <0.50,3.50>
Total: 11272 W: 601 L: 477 D: 10194
Ptnml(0-2): 9, 421, 4657, 535, 14
also had strong LTC performance against another strong net of the series:
https://tests.stockfishchess.org/tests/view/60c8c40d457376eb8bcab0c6
Bench: 5032320