question:
Want to use earlystopping in pytorch, after searching I found that 'EarlyStopping' from 'pytorchtools' can be used.
The tutorial says to use pip install pytorchtools to install, so the installed version is 0.0.2,
Then call from pytorchtools import EarlyStopping ,
But this will report an error ImportError: cannot import name 'EarlyStopping' from 'pytorchtools' .
reason:
After checking, I found that the 'pytorchtools' installed in this way is empty, and there is no 'EarlyStopping' in it.
Solution:
Copy the following code (or the code in the address ) into it, or directly create a new pytorchtools.py file in the project, then copy the code in and call it; )
import numpy as np
import torch
class EarlyStopping:
"""Early stops the training if validation loss doesn't improve after a given patience."""
def __init__(self, patience=7, verbose=False, delta=0):
"""
Args:
patience (int): How long to wait after last time validation loss improved.
上次验证集损失值改善后等待几个epoch
Default: 7
verbose (bool): If True, prints a message for each validation loss improvement.
如果是True,为每个验证集损失值改善打印一条信息
Default: False
delta (float): Minimum change in the monitored quantity to qualify as an improvement.
监测数量的最小变化,以符合改进的要求
Default: 0
"""
self.patience = patience
self.verbose = verbose
self.counter = 0
self.best_score = None
self.early_stop = False
self.val_loss_min = np.Inf
self.delta = delta
def __call__(self, val_loss, model):
score = -val_loss
if self.best_score is None:
self.best_score = score
self.save_checkpoint(val_loss, model)
elif score < self.best_score + self.delta:
self.counter += 1
# print(f'EarlyStopping counter: {self.counter} out of {self.patience}')
if self.counter >= self.patience:
self.early_stop = True
else:
self.best_score = score
self.save_checkpoint(val_loss, model)
self.counter = 0
def save_checkpoint(self, val_loss, model):
'''
Saves model when validation loss decrease.
验证损失减少时保存模型。
'''
if self.verbose:
print(f'Validation loss decreased ({self.val_loss_min:.6f} --> {val_loss:.6f}). Saving model ...')
torch.save(model.state_dict(), 'checkpoint.pth') # 这里会存储迄今最优模型的参数
# torch.save(model, 'finish_model.pkl') # 这里会存储迄今最优的模型
self.val_loss_min = val_loss