Skip to content

🧠 两小时从 0 训出你的 64M 大模型。Train your own 64M LLM from scratch in 2 hours.

[fix] lora compile

jingyaogonge796b8028a79
diff --git a/trainer/train_lora.py b/trainer/train_lora.py
index 271fecc..4ea259a 100644
--- a/trainer/train_lora.py
+++ b/trainer/train_lora.py
@@ -97,7 +97,6 @@ if __name__ == "__main__":
parser.add_argument('--from_resume', default=0, type=int, choices=[0, 1], help="是否自动检测&续训(0=否,1=是)")
parser.add_argument("--use_wandb", action="store_true", help="是否使用wandb")
parser.add_argument("--wandb_project", type=str, default="MiniMind-LoRA", help="wandb项目名")
- # PS:多卡 DDP + LoRA 场景下建议关闭 use_compile,避免触发 torch.compile 的兼容性问题
parser.add_argument("--use_compile", default=0, type=int, choices=[0, 1], help="是否使用torch.compile加速(0=否,1=是)")
args = parser.parse_args()
@@ -162,8 +161,8 @@ if __name__ == "__main__":
# ========== 8. 编译和分布式包装 ==========
if args.use_compile == 1:
- model = torch.compile(model)
- Logger('torch.compile enabled')
+ args.use_compile = 0
+ Logger('[LoRA] monkey-patch forward 与 torch.compile 不兼容,use_compile 已自动关闭')
if dist.is_initialized():
model._ddp_params_and_buffers_to_ignore = {"freqs_cos", "freqs_sin"}
model = DistributedDataParallel(model, device_ids=[local_rank])