Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

CNFGNN训练曲线 #6

Open
HaibingJin opened this issue Sep 19, 2023 · 3 comments
Open

CNFGNN训练曲线 #6

HaibingJin opened this issue Sep 19, 2023 · 3 comments

Comments

@HaibingJin
Copy link

HaibingJin commented Sep 19, 2023

您好,我在跑实验结果的时候,使用您给的参数:

CNFGNN (64K + 1M)

python main.py --dataset METR-LA --seed 42 --model_name SplitFedAvgNodePredictor --base_model_name GRUSeq2SeqWithGraphNet --batch_size 128 --server_batch_size 48 --hidden_size 64 --suffix mp --use_curriculum_learning --mp_worker_num 16 --sync_every_n_epoch 1 --server_epoch 1 --gcn_on_server --gpus 0, --max_epochs 500 --early_stop_patience 20 --gru_num_layers 1
得到了如下图所示的结果:
image
image
很奇怪为什么train loss会这样,请问您在做实验的时候有出现这个问题吗?这个应该如何解释

@lky-violet
Copy link

我也一样,而且我发现他的数据比原始数据多一个attr维度,请问你知道这个维度代表什么吗?DCRNN 中提供的原始数据貌似没有这个维度

@HaibingJin
Copy link
Author

attr是数据的属性维度,METR-LA数据集除了交通流没有其他维度,所以在实际实现的时候这个attr属性值都是0,也可以删掉它。

@llz12
Copy link

llz12 commented Sep 12, 2024

请问这份代码要运行多久,我跑了十几个小时没有结果

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants