summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorhaoyuren <13851610112@163.com>2026-02-22 12:17:41 -0600
committerhaoyuren <13851610112@163.com>2026-02-22 12:17:41 -0600
commit7345f23d69fe40313907c9eac3094c0f05673166 (patch)
tree31f6c778771212d394c11e98ae65cbd3e2c3e4af
parentc41ea8629ce1351e415cce1551a7b52260a66790 (diff)
Use auto-calibrated collect_batch in Colab notebook
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
-rw-r--r--train_colab.ipynb4
1 files changed, 2 insertions, 2 deletions
diff --git a/train_colab.ipynb b/train_colab.ipynb
index 66b244e..a0e0c61 100644
--- a/train_colab.ipynb
+++ b/train_colab.ipynb
@@ -46,14 +46,14 @@
{
"cell_type": "code",
"metadata": {},
- "source": "# 2-player training (GPU PPO + batched collection)\n!python train.py \\\n --num_players 2 \\\n --episodes 200000 \\\n --collect_batch 128 \\\n --save_path blazing_ppo_2p",
+ "source": "# 2-player training (GPU PPO + batched collection)\n# collect_batch auto-calibrated if not specified\n!python train.py \\\n --num_players 2 \\\n --episodes 200000 \\\n --save_path blazing_ppo_2p",
"execution_count": null,
"outputs": []
},
{
"cell_type": "code",
"metadata": {},
- "source": "# (Optional) 3-player training\n# !python train.py --num_players 3 --episodes 300000 --collect_batch 128 --save_path blazing_ppo_3p\n\n# (Optional) Larger batch for faster throughput\n# !python train.py --num_players 2 --episodes 200000 --collect_batch 256 --save_path blazing_ppo_2p\n\n# (Optional) Skip greedy warmup\n# !python train.py --num_players 2 --episodes 200000 --greedy_warmup 0 --save_path blazing_ppo_2p_no_warmup",
+ "source": "# (Optional) 3-player training\n# !python train.py --num_players 3 --episodes 300000 --save_path blazing_ppo_3p\n\n# (Optional) Manual batch size (skip auto-calibration)\n# !python train.py --num_players 2 --episodes 200000 --collect_batch 256 --save_path blazing_ppo_2p\n\n# (Optional) Skip greedy warmup\n# !python train.py --num_players 2 --episodes 200000 --greedy_warmup 0 --save_path blazing_ppo_2p_no_warmup",
"execution_count": null,
"outputs": []
},