MAZiqing/FEDformer

Issues

• Getting a 'nan' Loss after about 12 training epochs

  #85 opened 4 months ago by marc-martini
  4

• minute ignored when running ETTm2 using provided code & script

  #86 opened 4 months ago by efg001
  0

• RuntimeError: einsum(): operands do not broadcast with remapped shapes [original->remapped]

  #73 opened a year ago by asdfzlcl
  7

• 关于exp_main.py中的一个错误

  #72 opened 4 months ago by superguymj
  4

• RuntimeError

  #71 opened 4 months ago by wbxibcy
  2

• 关于代码效率问题

  #79 opened 4 months ago by superguymj
  1

• 关于模型参数量的问题

  #82 opened 4 months ago by nihaoya12
  1

• 为什么FourierCrossAttention的代码没有用到v呢

  #84 opened 4 months ago by wiluen
  2

• 运行结果中test_results这个文件夹里的pdf文件0 20 40 60 80 都是些什么，为什么在result中的结果很差，但是这个test_results中的图片看起来就很好呢

  #78 opened 4 months ago by kooper9934
  2

• Train FEDformer with wavelets

  #83 opened 4 months ago by philipperbd
  1

• 关于Fedformer模型输出结果的output size问题

  #80 opened 10 months ago by FAFUuser
  3

• Inconsistent results when recreating model instance

  #77 opened 10 months ago by b-nils
  2

• 11

  #81 opened 9 months ago by yaoandye
  0

• 请问对于序列分解所使用的多个移动平均核体现在哪里

  #76 opened 10 months ago by ooooooover
  2

• 为什么会一直降低学习率反复运行，怎么关闭这部分功能

  #74 opened a year ago by kooper9934
  2

• Why does it take so long before the training really starts?

  #75 opened 10 months ago by jannichorst
  2

• 关于运行时报错：nvrtc: error: invalid value for --gpu-architecture (-arch)

  #70 opened 10 months ago by lzb0854
  1

• 关于masked attention的疑问

  #69 opened a year ago by sunxiaoyao-git
  2

• train epochs

  #68 opened a year ago by jayyounghappy
  1

• forecast result

  #67 opened a year ago by anranlovelife
  1

• Output the attention weights of the model

  #66 opened a year ago by MatthiasNeumueller
  1

• 预测结果为负值

  #51 opened 2 years ago by ZFY-git
  8

• 关于使用自己的数据集做未来预测的问题

  #53 opened 2 years ago by Luojlong
  3

• TypeError: Adam.__init__() missing 1 required positional argument: 'params'

  #64 opened a year ago by Med-Rokaimi
  2

• Can't get the ideal result in table2(FEDformer-w for ILI24pl)

  #65 opened a year ago by GiaoBruce
  2

• How to train, test, and forecast my own datasets?

  #45 opened a year ago by TungyuYoung
  5

• About seq_len and label_len

  #52 opened a year ago by nevoliu
  1

• running stuck

  #42 opened a year ago by guetaro
  5

• About pred_len at training and pred_len at inference

  #59 opened a year ago by somaonishi
  3

• FEA-f 中\quat V的问题

  #62 opened a year ago by Xiangyi-Sufe
  2

• 双卡运行bug

  #63 opened a year ago by qaqxk
  1

• 关于Frequency Enhanced Attention

  #60 opened a year ago by linfeng-du
  4

• 关于Attention Map可视化

  #61 opened a year ago by Whimmey
  6

• "min_all_cuda" not implemented for 'ComplexFloat'

  #58 opened 2 years ago by UngQuangHuy
  1

• About projection in AttentionLayer

  #57 opened 2 years ago by somaonishi
  2

• RuntimeError

  #44 opened 2 years ago by yingdd0
  3

• ValueError: __len__() should return >= 0

  #55 opened 2 years ago by DLW30
  1

• Scirpt/run_M.sh 部分有误

  #56 opened 2 years ago by huangsiyuan924
  1

• FEDformer model training results are not ideal

  #43 opened 2 years ago by LivingRoomcode
  1

• Questions about the number of model parameters

  #49 opened 2 years ago by nevoliu
  2

• FourierBlock中scale的选择

  #50 opened 2 years ago by rouniuyizu
  1

• information leakage

  #48 opened 2 years ago by LCZ2815
  2

• 关于论文中图5里的Us(L)和Ud(L)

  #47 opened 2 years ago by vaeaea
  1

• 如何解决该代码中归一化和反归一化问题，应该调整哪些参数，为什么输出的预测值是经过归一化的结果

  #46 opened 2 years ago by Xjt128
  1

• About environment requirements

  #41 opened 2 years ago by Zeying-Gong
  1

• 请问验证过程选择的M个频率分量是和训练过程的一样的吗？还是也是随机选择的呢？

  #40 opened 2 years ago by lzl2040
  1

• 在FEDformer中我咋没找到哪里用了傅里叶啊

  #39 opened 2 years ago by jhzmike
  1

• Loading State_dict Error happend when do predict

  #37 opened 2 years ago by jwee-star
  2

• run_M.sh ETTh1 --feature should be M

  #38 opened 2 years ago by dongjiaxiang
  1

• Training on multiple time series

  #36 opened 2 years ago by uday-rao-aera
  1