LongxingTan

@YakubFahimLuckyarno sorry i missed to mention another directory change. in /example/data/read_data.py line 102 ``` def get_examples(self, data_dir, sample=1, start_date=None, plot=False, model_dir='../weights'): ```

> > @YakubFahimLuckyarno > > sorry i missed to mention another directory change. > > in /example/data/read_data.py line 102 > > ``` > > def get_examples(self, data_dir, sample=1, start_date=None, plot=False,...

@Hzither Hi, can you copy more your training code? from the screen shot, it seems fine. Epoch 1/100

是不是跑的时间比较长, 其实还在跑？我用了kaggle一份上传的数据，跑了三个epoch是没有问题的。不过上传的人改了一些代码导致后来会报错 [https://www.kaggle.com/code/tanlongxing/kdd-cup-2022-wind-power-forecast](https://www.kaggle.com/code/tanlongxing/kdd-cup-2022-wind-power-forecast)

当时是把这个压缩文件传到网上提交出分了。可以根据[预测代码](https://github.com/LongxingTan/KDDCup2022-WPF/blob/master/src/inference/predict.py)，自己从本地数据里给预测