|
|
楼主 |
发表于 2024-11-3 12:38:44
|
显示全部楼层
---------------------------------------------------------------------------
KeyError Traceback (most recent call last)
Cell In[118], line 4
1 print('Training Start!')
2 print('=' * 100)
----> 4 train(model,
5 device,
6 train_dataloader,
7 valid_dataloader,
8 CFG.epochs,
9 loss_fn,
10 optimizer,
11 metric)
13 del model,train_dataloader, valid_dataloader
14 gc.collect()
Cell In[82], line 17, in train(model, device, train_dataloader, valid_dataloader, epochs, loss_fn, optimizer, metric)
14 train_step = 0
15 pbar = tqdm(train_dataloader)#tqdm参数是一个iterable
---> 17 for batch in pbar: # you can also write like "for batch in tqdm(train_dataloader"
18 optimizer.zero_grad() # initialize
19 train_step += 1
File /opt/conda/lib/python3.10/site-packages/tqdm/notebook.py:250, in tqdm_notebook.__iter__(self)
248 try:
249 it = super().__iter__()
--> 250 for obj in it:
251 # return super(tqdm...) will not catch exception
252 yield obj
253 # NB: except ... [ as ...] breaks IPython async KeyboardInterrupt
File /opt/conda/lib/python3.10/site-packages/tqdm/std.py:1181, in tqdm.__iter__(self)
1178 time = self._time
1180 try:
-> 1181 for obj in iterable:
1182 yield obj
1183 # Update and possibly print the progressbar.
1184 # Note: does not call self.update(1) for speed optimisation.
File /opt/conda/lib/python3.10/site-packages/torch/utils/data/dataloader.py:630, in _BaseDataLoaderIter.__next__(self)
627 if self._sampler_iter is None:
628 # TODO(https://github.com/pytorch/pytorch/issues/76750)
629 self._reset() # type: ignore[call-arg]
--> 630 data = self._next_data()
631 self._num_yielded += 1
632 if self._dataset_kind == _DatasetKind.Iterable and \
633 self._IterableDataset_len_called is not None and \
634 self._num_yielded > self._IterableDataset_len_called:
File /opt/conda/lib/python3.10/site-packages/torch/utils/data/dataloader.py:1344, in _MultiProcessingDataLoaderIter._next_data(self)
1342 else:
1343 del self._task_info[idx]
-> 1344 return self._process_data(data)
File /opt/conda/lib/python3.10/site-packages/torch/utils/data/dataloader.py:1370, in _MultiProcessingDataLoaderIter._process_data(self, data)
1368 self._try_put_index()
1369 if isinstance(data, ExceptionWrapper):
-> 1370 data.reraise()
1371 return data
File /opt/conda/lib/python3.10/site-packages/torch/_utils.py:706, in ExceptionWrapper.reraise(self)
702 except TypeError:
703 # If the exception takes multiple arguments, don't try to
704 # instantiate since we don't know how to
705 raise RuntimeError(msg) from None
--> 706 raise exception
KeyError: Caught KeyError in DataLoader worker process 0.
Original Traceback (most recent call last):
File "/opt/conda/lib/python3.10/site-packages/pandas/core/indexes/base.py", line 3805, in get_loc
return self._engine.get_loc(casted_key)
File "index.pyx", line 167, in pandas._libs.index.IndexEngine.get_loc
File "index.pyx", line 196, in pandas._libs.index.IndexEngine.get_loc
File "pandas/_libs/hashtable_class_helper.pxi", line 2606, in pandas._libs.hashtable.Int64HashTable.get_item
File "pandas/_libs/hashtable_class_helper.pxi", line 2630, in pandas._libs.hashtable.Int64HashTable.get_item
KeyError: 2183
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/_utils/worker.py", line 309, in _worker_loop
data = fetcher.fetch(index) # type: ignore[possibly-undefined]
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 52, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/opt/conda/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 52, in <listcomp>
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/tmp/ipykernel_31/2920046666.py", line 34, in __getitem__
labels = self.df.loc[idx]['label']
File "/opt/conda/lib/python3.10/site-packages/pandas/core/indexing.py", line 1191, in __getitem__
return self._getitem_axis(maybe_callable, axis=axis)
File "/opt/conda/lib/python3.10/site-packages/pandas/core/indexing.py", line 1431, in _getitem_axis
return self._get_label(key, axis=axis)
File "/opt/conda/lib/python3.10/site-packages/pandas/core/indexing.py", line 1381, in _get_label
return self.obj.xs(label, axis=axis)
File "/opt/conda/lib/python3.10/site-packages/pandas/core/generic.py", line 4301, in xs
loc = index.get_loc(key)
File "/opt/conda/lib/python3.10/site-packages/pandas/core/indexes/base.py", line 3812, in get_loc
raise KeyError(key) from err
KeyError: 2183
以上为报错信息，相关代码如下def prepare_input(cfg, text):
inputs = cfg.tokenizer.encode_plus(
text,
return_tensors=None,
add_special_tokens=True,#自动在每个文本前后添加特殊标记(如CLS和SEP)
return_attention_mask=True # 计算注意力（attention）时忽略那些paddle值
)
if len(inputs['input_ids']) > CFG.max_len:#句子长度大于max_length时截断
inputs['input_ids'] = inputs['input_ids'][:CFG.max_len]
inputs['attention_mask'] = inputs['attention_mask'][:CFG.max_len]
inputs['token_type_ids'] = inputs['token_type_ids'][:CFG.max_len]
for k, v in inputs.items():
inputs[k] = torch.tensor(v, dtype=torch.long)
return inputs
class LLMDataset(Dataset):
def __init__(self, cfg, df,is_grad):
self.cfg = cfg
self.df = df
self.texts = df["cleaned"].values
self.is_grad = is_grad
def __len__(self):
return len(self.texts)
def __getitem__(self,idx):
inputs = prepare_input(self.cfg, self.texts[idx])#inputs是一个字典
if self.is_grad:#训练集
labels = self.df.loc[idx]['label']
# [batch,1,max_len(84)] -> [batch,max_len]#使用squeeze降维
return {'input_ids':inputs['input_ids'].squeeze(),
'attention_mask':inputs['attention_mask'].squeeze(),
'token_type_ids':inputs['token_type_ids'].squeeze(),
# Our loss_fn wants it to be a "float" type
'labels':torch.tensor(labels,dtype=torch.float).unsqueeze(dim=0)}
else:#测试集
# [batch,1,max_len(84)] -> [batch,max_len]
return {'input_ids':inputs['input_ids'].squeeze(),
'attention_mask':inputs['attention_mask'].squeeze(),
'token_type_ids':inputs['token_type_ids'].squeeze()}
class CollateCls:
def __init__(self, cfg):
self.tokenizer = cfg.tokenizer
self.cfg = cfg
def __call__(self, batch):
output = dict()
output["input_ids"] = [sample["input_ids"] for sample in batch]
output["attention_mask"] = [sample["attention_mask"] for sample in batch]
output["token_type_ids"] = [sample["token_type_ids"] for sample in batch]
# calculate max token length of this batch
batch_max = max([len(ids) for ids in output["input_ids"]])
# add padding
if self.tokenizer.padding_side == "right":
output["input_ids"] = [
list(s) + (batch_max - len(s)) * [self.tokenizer.pad_token_id]
for s in output["input_ids"]
]
output["attention_mask"] = [
list(s) + (batch_max - len(s)) * [0] for s in output["attention_mask"]
]
output["token_type_ids"] = [list(s) + (batch_max - len(s)) * [0] for s in output["token_type_ids"]]
else:
output["input_ids"] = [
(batch_max - len(s)) * [self.tokenizer.pad_token_id] + list(s)
for s in output["input_ids"]
]
output["attention_mask"] = [
(batch_max - len(s)) * [0] + list(s) for s in output["attention_mask"]
]
output["token_type_ids"] = [(batch_max - len(s)) * [0] + list(s) for s in output["token_type_ids"]]
# convert to tensors
output["input_ids"] = torch.tensor(output["input_ids"], dtype=torch.long)
output["attention_mask"] = torch.tensor(output["attention_mask"], dtype=torch.long)
output["token_type_ids"] = torch.tensor(output["token_type_ids"], dtype=torch.long)
return output
#生成训练集
train_dataset = LLMDataset(CFG,pd.concat([X_train,y_train],axis=1),True)
valid_dataset = LLMDataset(CFG,pd.concat([X_valid,y_valid],axis=1),False)
from torch.utils.data import DataLoader
train_dataloader = DataLoader(train_dataset,batch_size=32,collate_fn=CollateCls(CFG),shuffle=True,num_workers = 4,pin_memory=True)#锁页内存(pin_memory)能够保持与GPU进行高速传输,在训练时加快数据的读取
valid_dataloader = DataLoader(valid_dataset,batch_size=32,collate_fn=CollateCls(CFG),shuffle=False,num_workers = 4,pin_memory=True)
#collate_fn用于自定义数据加载和批处理的方式
import gc,os
from tqdm.auto import tqdm # visualizing tool for progress
# They will be used to pick the best model.pt given to the valid loss
best_model_epoch, valid_loss_values = [],[]
valid_loss_min = [1] # arbitrary loss I set here
def train(model,device,train_dataloader,valid_dataloader,epochs,loss_fn,optimizer,metric):
for epoch in range(epochs):
gc.collect() # memory cleaning垃圾回收机制，减少占用内存
model.train()
train_loss = 0
train_step = 0
pbar = tqdm(train_dataloader)#tqdm参数是一个iterable
for batch in pbar: # you can also write like "for batch in tqdm(train_dataloader"
optimizer.zero_grad() # initialize
train_step += 1
train_input_ids = batch['input_ids'].to(device)
train_attention_mask = batch['attention_mask'].to(device)
train_token_type_ids = batch['token_type_ids'].to(device)
train_labels = batch['labels'].squeeze().to(device).long()#label真实值long()转化成一维张量
# You can refer to the class "TweetsModel" for understand
# what would be logits
logits = model(train_input_ids, train_attention_mask,train_token_type_ids).to(device)
predictions = torch.argmax(logits, dim=1) # get an index from larger one
detached_predictions = predictions.detach().cpu().numpy()
loss = loss_fn(logits, train_labels)
loss.backward()
optimizer.step()
model.zero_grad()
train_loss += loss.detach().cpu().numpy().item()
pbar.set_postfix({'train_loss':train_loss/train_step})#设置进度条显示信息
pbar.close()
with torch.no_grad():
model.eval()
valid_loss = 0
valid_step = 0
total_valid_score = 0
y_pred = [] # for getting f1_score that is a metric of the competition
y_true = []
pbar = tqdm(valid_dataloader)
for batch in pbar:
valid_step += 1
valid_input_ids = batch['input_ids'].to(device)
valid_attention_mask = batch['attention_mask'].to(device)
valid_token_type_ids = batch['token_type_ids'].to(device)
valid_labels = batch['labels'].squeeze().to(device).long()
logits = model(valid_input_ids, valid_attention_mask).to(device)
predictions = torch.argmax(logits, dim=1)
detached_predictions = predictions.detach().cpu().numpy()
loss = loss_fn(logits, valid_labels)
valid_loss += loss.detach().cpu().numpy().item()
y_pred.extend(predictions.cpu().numpy())
y_true.extend(valid_labels.cpu().numpy())
valid_loss /= valid_step
f1 = f1_score(y_true,y_pred)
print(f'Epoch [{epoch+1}/{epochs}] Score: {f1}')
print(f'Epoch [{epoch+1}/{epochs}] Valid_loss: {valid_loss}')
if valid_loss < min(valid_loss_min):
print('model improved!')
else:
print('model not improved')
torch.save(model.state_dict(), f'epoch:{epoch+1}_model.pt')#state_dict 是一个字典对象，包含了模型的所有可学习参数（如权重和偏置）及其当前值
print('save checkpoint!')
valid_loss_min.append(valid_loss)
print(f'valid_loss_min:{min(valid_loss_min)}')
best_model_epoch.append(f'/kaggle/working/epoch:{epoch+1}_model.pt')
valid_loss_values.append(valid_loss)
print('='*100)
select_best_model() # refer to below function
print('Train/Valid Completed!!')
del train_dataloader, valid_dataloader # memory cleaning
gc.collect()
def select_best_model():
best_model = best_model_epoch[np.array(valid_loss_values).argmin()]
os.rename(best_model, best_model.split('.pt')[0] + '_best.pt')#重命名文件
print('Training Start!')
print('=' * 100)
train(model,
device,
train_dataloader,
valid_dataloader,
CFG.epochs,
loss_fn,
optimizer,
metric)
del model,train_dataloader, valid_dataloader
gc.collect()
|
|
( 粤ICP备18085999号-1 | 粤公网安备 44051102000585号)
狗仔卡
发表于 2024-11-3 12:17:11
置顶卡
千斤顶
显身卡