GAT模型val_loss_epoch一直在上升

[fanfanxiaozu]

训练参数如下： d_feat: 20 hidden_size: 64 num_layers: 2 dropout: 0.7 n_epochs: 200 lr: 1e-4 early_stop: 200 metric: loss loss: mse base_model: LSTM model_path: "/root/autodl-tmp/qlib/examples/benchmarks/LSTM/csi300_lstm_ts.pkl" GPU: 0

训练结果，200个epochloss下降到0.94左右，请问这个loss下降是否正常？降到多少才算合理？：

[fanfanxiaozu]

感觉有点奇怪，val_loss_epoch一直呈现上升趋势

[Fivele-Li]

Hi @lwwang1995, I noticed you made a contribution related to this GAT model. I was wondering if you could provide some insight or assistance. Thank you in advance for your help!

[xiangqingcs]

/root/autodl-tmp/qlib/examples/benchmarks/LSTM/csi300_lstm_ts.pkl

How is file csi300_lstm_ts.pkl generated?

[keyboard1st]

求问你怎么画出loss曲线的，谢谢大佬