我一直在尝试将Fastai与自定义割炬模型一起使用。我的代码如下：
X_train = np.load(dirpath + 'X_train.npy')
X_valid = np.load(dirpath + 'X_valid.npy')
Y_train = np.load(dirpath + 'Y_train.npy')
Y_valid = np.load(dirpath + 'Y_valid.npy')
X_train's shape is : (240, 122, 96),
and Y_train's shape is : (240,1)
然后我将它们转换为火炬张量
# Converting data to torch tensors
def to_torch_data(x,np_type,tch_type):
return torch.from_numpy(x.astype(np_type)).to(tch_type)
X_train = to_torch_data(X_train,float,torch.float32)
X_valid = to_torch_data(X_valid,float,torch.float32)
Y_train = to_torch_data(Y_train,float,torch.float32)
Y_valid = to_torch_data(Y_valid,float,torch.float32)
为fastai DataBunch包装器创建TensorDataSet，
# Creating torch tensor datasets so that data can be used
# on ImageDataBunch function for fastai
train_ds = tdatautils.TensorDataset(X_train,Y_train)
valid_ds = tdatautils.TensorDataset(X_valid,Y_valid)
# Creating DataBunch object to be used as data in fastai methods.
batch_size = 24
my_data_bunch = DataBunch.create(train_ds,valid_ds,bs=batch_size)
这是我的自定义火炬模型：
# Creating corresponding torch model
import torch.nn.functional as F
class Net(nn.Module):
def __init__(self,droprate=0,activationF=None):
super(Net, self).__init__()
self.lstm_0 = nn.LSTM(96, 720)
self.activation_0 = nn.ELU()
self.dropout_0 = nn.Dropout(p=droprate)
self.lstm_1 = nn.LSTM(720,480)
self.activation_1 = nn.ELU()
self.batch_norm_1 = nn.BatchNorm1d(122)
self.fc_2 = nn.Linear(480,128)
self.dropout_2 = nn.Dropout(p=droprate)
self.last = nn.Linear(128,1)
self.last_act = nn.ReLU()
def forward(self, x):
out,hid1 = self.lstm_0(x)
out = self.dropout_0(self.activation_0(out))
out,hid2 = self.lstm_1(out)
out = out[:,-1,:]
out = self.batch_norm_1(self.activation_1(out))
out = self.dropout_2(self.fc_2(out))
out = self.last_act(self.last(out))
return out
#create instance of model
net = Net(droprate=train_droprate,activationF=train_activation) #.cuda()
print(net)
所有这些之后，我运行了Learn for lr_find方法。我得到这个错误：
Empty Traceback (most recent call last)
C:\Anaconda3\envs\fastai\lib\site-packages\torch\utils\data\dataloader.py in _try_get_batch(self, timeout)
510 try:
--> 511 data = self.data_queue.get(timeout=timeout)
512 return (True, data)
C:\Anaconda3\envs\fastai\lib\queue.py in get(self, block, timeout)
171 if remaining <= 0.0:
--> 172 raise Empty
173 self.not_empty.wait(remaining)
Empty:
During handling of the above exception, another exception occurred:
RuntimeError Traceback (most recent call last)
<ipython-input-35-e4b7603c0a82> in <module>
----> 1 my_learner.lr_find()
~\Desktop\fastai\fastai\fastai\train.py in lr_find(learn, start_lr, end_lr, num_it, stop_div, wd)
30 cb = LRFinder(learn, start_lr, end_lr, num_it, stop_div)
31 epochs = int(np.ceil(num_it/len(learn.data.train_dl)))
---> 32 learn.fit(epochs, start_lr, callbacks=[cb], wd=wd)
33
34 def to_fp16(learn:Learner, loss_scale:float=None, max_noskip:int=1000, dynamic:bool=True, clip:float=None,
~\Desktop\fastai\fastai\fastai\basic_train.py in fit(self, epochs, lr, wd, callbacks)
197 callbacks = [cb(self) for cb in self.callback_fns + listify(defaults.extra_callback_fns)] + listify(callbacks)
198 if defaults.extra_callbacks is not None: callbacks += defaults.extra_callbacks
--> 199 fit(epochs, self, metrics=self.metrics, callbacks=self.callbacks+callbacks)
200
201 def create_opt(self, lr:Floats, wd:Floats=0.)->None:
~\Desktop\fastai\fastai\fastai\basic_train.py in fit(epochs, learn, callbacks, metrics)
97 cb_handler.set_dl(learn.data.train_dl)
98 cb_handler.on_epoch_begin()
---> 99 for xb,yb in progress_bar(learn.data.train_dl, parent=pbar):
100 xb, yb = cb_handler.on_batch_begin(xb, yb)
101 loss = loss_batch(learn.model, xb, yb, learn.loss_func, learn.opt, cb_handler)
C:\Anaconda3\envs\fastai\lib\site-packages\fastprogress\fastprogress.py in __iter__(self)
70 self.update(0)
71 try:
---> 72 for i,o in enumerate(self._gen):
73 if i >= self.total: break
74 yield o
~\Desktop\fastai\fastai\fastai\basic_data.py in __iter__(self)
73 def __iter__(self):
74 "Process and returns items from `DataLoader`."
---> 75 for b in self.dl: yield self.proc_batch(b)
76
77 @classmethod
C:\Anaconda3\envs\fastai\lib\site-packages\torch\utils\data\dataloader.py in __next__(self)
574 while True:
575 assert (not self.shutdown and self.batches_outstanding > 0)
--> 576 idx, batch = self._get_batch()
577 self.batches_outstanding -= 1
578 if idx != self.rcvd_idx:
C:\Anaconda3\envs\fastai\lib\site-packages\torch\utils\data\dataloader.py in _get_batch(self)
541 elif self.pin_memory:
542 while self.pin_memory_thread.is_alive():
--> 543 success, data = self._try_get_batch()
544 if success:
545 return data
C:\Anaconda3\envs\fastai\lib\site-packages\torch\utils\data\dataloader.py in _try_get_batch(self, timeout)
517 if not all(w.is_alive() for w in self.workers):
518 pids_str = ', '.join(str(w.pid) for w in self.workers if not w.is_alive())
--> 519 raise RuntimeError('DataLoader worker (pid(s) {}) exited unexpectedly'.format(pids_str))
520 if isinstance(e, queue.Empty):
521 return (False, None)
RuntimeError: DataLoader worker (pid(s) 9584, 7236, 5108, 932, 13228, 13992, 4576, 13204) exited unexpectedly
我已经研究过DataLoader，但是找不到有用的东西。
答案 0 :(得分：2)
尽管我不理解您发布的错误消息，但是我在代码中看到了一个问题。
out = out[:,-1,:] # batch_size x 480
out = self.batch_norm_1(self.activation_1(out))
但是您将batch_norm_1声明为：
self.batch_norm_1 = nn.BatchNorm1d(122)
应该是：
self.batch_norm_1 = nn.BatchNorm1d(480)