Skip to content

Instantly share code, notes, and snippets.

@jonbinney
Created November 24, 2025 21:46
Show Gist options
  • Select an option

  • Save jonbinney/3f63eaebffd76eec875db0e8a4a6235f to your computer and use it in GitHub Desktop.

Select an option

Save jonbinney/3f63eaebffd76eec875db0e8a4a6235f to your computer and use it in GitHub Desktop.
Example sweep file for deep_quoridor
program: deep_quoridor/src/train_alphazero.py
method: grid
name: sweep
metric:
goal: maximize
name: win_perc
parameters:
alphazero:
parameters:
batch_size:
value: 40
mcts_n:
values: [10, 100]
replay_buffer_size:
value: 20000
save_replay_buffer:
value: "always"
validation_ratio:
value: 0.2
nn_type:
value: "resnet"
nn_resnet_num_blocks:
value: 4
nn_resnet_num_channels:
value: 8
nn_mask_training_predictions:
value: true
command:
- ${env}
- python
- ${program}
- "--sweep=deep_quoridor/sweep_test.yaml"
- "-w=log_from_workers=false"
- "--board-size=3"
- "--max-walls=0"
- "--max-steps=40"
- "--seed=10"
- "-g=2"
- "-e=1"
- "--num-workers=1"
- "-pg=8"
- "--benchmarks=greedy"
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment