kanshan/minimind
1
0
Fork 0
mirror of https://github.com/jingyaogong/minimind.git synced 2026-04-25 08:48:16 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

[feat] update lr

Browse Source
This commit is contained in:
jingyaogong 2025-12-31 10:27:09 +08:00
parent 936d105e9b
commit 6242980917
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
trainer/train_full_sft.py
View File

@ -87,7 +87,7 @@ if __name__ == "__main__":
parser.add_argument('--save_weight', default='full_sft', type=str, help="保存权重的前缀名")
parser.add_argument("--epochs", type=int, default=2, help="训练轮数")
parser.add_argument("--batch_size", type=int, default=16, help="batch size")
parser.add_argument("--learning_rate", type=float, default=5e-7, help="初始学习率")
parser.add_argument("--learning_rate", type=float, default=1e-6, help="初始学习率")
parser.add_argument("--device", type=str, default="cuda:0" if torch.cuda.is_available() else "cpu", help="训练设备")
parser.add_argument("--dtype", type=str, default="bfloat16", help="混合精度类型")
parser.add_argument("--num_workers", type=int, default=8, help="数据加载线程数")