GRU（选通重复单元）显示发电机功能数据不足的错误，并且没有为“提供数据”；每个键在；在R（RStudio）

这里的初学者-没有正式的ML背景或任何相近的背景，还有一点自学成才的R。我用RStudio来做这件事。
我已经建立了一个基本的G.R.U.，我希望在此基础上建立更多的流程。我的第一个步骤是按照上的说明，使用生成器函数来训练数据并将其输入到

fit\u generator（）

。第二种方法不使用生成器，而是使用fit.keras.engine.training.Model（） 我在此尝试和出错点的设置是-
带发电机功能：

library(keras)

model <- keras_model_sequential() %>% 
  layer_gru(units = 5, input_shape = list(NULL, dim(data)[[-1]])) %>% 
  layer_dense(units = 1)

model %>% compile(
  optimizer = optimizer_rmsprop(),
  loss = "mae"
)

history <- model %>% fit_generator(
  train_gen,
  steps_per_epoch = 100,
  epochs = 1,
  validation_data = val_gen,
  validation_steps = val_steps
)

library(keras)

model <- keras_model_sequential() %>% 
  layer_gru(units = 32, input_shape = list(NULL, dim(data)[[-1]])) %>% 
  layer_dense(units = 1)

model %>% compile(
  optimizer = optimizer_rmsprop(),
  loss = "mae"
)

history <- model %>% fit(
  x=train_data[,1:2],
  y=train_data[,3:4],
  epochs = 2,
  verbose=1,
  shuffle=F,
  validation_data = val_data[,1:4]
)

（省略了“详细追溯”）

没有发电机（我决定重点关注）：

什么是gru_8_输入？甚至谷歌也不知道它是什么！ 请帮忙

加上，
发电机功能来自

谢谢

library(tibble)
library(readr)

data_dir <- "C:/Users/elech/Desktop/traindats"
USDCADname <- file.path(data_dir, c("USDCAD.csv"))
USDAUDname <- file.path(data_dir, c("USDAUD.csv"))
USDCHFname <- file.path(data_dir, c("USDCHF.csv"))
USDNZDname <- file.path(data_dir, c("USDNZD.csv"))
USDJPYname <- file.path(data_dir, c("USDJPY.csv"))

USDCAD<- read_csv(USDCADname, col_names = F)
USDAUD<- read_csv(USDAUDname, col_names = F)
USDCHF<- read_csv(USDCHFname, col_names = F)
USDNZD<- read_csv(USDNZDname, col_names = F)
USDJPY<- read_csv(USDJPYname, col_names = F)

Close_USDAUD<-USDAUD$X6
Close_USDCAD<-USDCAD$X6
Close_USDCHF<-USDCHF$X6
Close_USDNZD<-USDNZD$X6
Close_USDJPY<-USDJPY$X6

train_data<-data[1:2000,]
val_data<-data[2001:4000,]
test_data<-data[4001:nrow(data),]
data<-tibble(USDAUD=Close_USDAUD[176:5175], USDCAD=Close_USDCAD[173:5172], USDCHF=Close_USDCHF[176:5175], USDJPY=Close_USDJPY[175:5174], USDNZD=Close_USDNZD[170:5169])

Error occurred in generator: argument 'length.out' must be of length 1
WARNING:tensorflow:Your dataset iterator ran out of data; interrupting training. Make sure that your iterator can generate at least `steps_per_epoch * epochs` batches (in this case, 2000 batches). You may need touse the repeat() function when building your dataset.
Error in py_call_impl(callable, dots$args, dots$keywords) : 
  ValueError: Empty training data.

Error in py_call_impl(callable, dots$args, dots$keywords) : 
  ValueError: No data provided for "gru_8_input". Need data for each key in: ['gru_8_input']