e96031413

// 每60分钟自动运行代码刷新，解除90分钟断开限制.

// 使用方法：colab页面按下 F12或者 Ctrl+Shift+I (mac按 Option+Command+I) 在console（控制台） 输入以下代码并回车.

// 复制以下代码粘贴在浏览器console！！不要关闭浏览器以免失效
function ConnectButton(){
    console.log("Connect pushed"); 
    document.querySelector("#connect").click() 
}
setInterval(ConnectButton,60000);

e96031413

from tsnecuda import TSNE
from tsne.resnet import ResNet18

# 使用 PyTorch內建的 ResNet18
import os
import torch
import torchvision.models as models
import torch.optim
from torchvision import transforms
model = models.resnet18()

e96031413

# https://stackoverflow.com/questions/63761717/load-image-dataset
# https://stackoverflow.com/questions/60655280/how-to-split-an-image-dataset-in-x-train-y-train-x-test-y-test-by-tensorflow

import tensorflow as tf
import pandas as pd

train_df = pd.read_csv('train.csv')
train_df['class'] = train_df['class'].apply(str)

train_datagen = tf.keras.preprocessing.image.ImageDataGenerator(horizontal_flip=True, vertical_flip=True,)

e96031413

# 針對每個元件計算其Overkill rate和Leakage rate
# 目前只適用於batch size = 1的情形
# 用來產生overkill和leakage數值的dataframe
import pandas as pd
import torch

test_df_mapping2_label = test_df.copy()                                          #複製一份要mapping到2個label的testing資料
test_df_mapping2_label.loc[test_df_mapping2_label['class'] == 0, 'class'] = 0    #將大於1的label轉成1
test_df_mapping2_label.loc[test_df_mapping2_label['class'] == 1, 'class'] = 1
test_df_mapping2_label.loc[test_df_mapping2_label['class'] == 2, 'class'] = 1

e96031413

# REF https://stackoverflow.com/questions/55497789/find-a-word-in-multiple-powerpoint-files-python/55763992#55763992
from pptx import Presentation
from pptx.enum.shapes import MSO_SHAPE_TYPE

import os

path = "./"

files = [x for x in os.listdir(path) if x.endswith(".pptx")] 

e96031413

import numpy as np
from sklearn import metrics
from sklearn.metrics import roc_auc_score
y = np.array([1, 1, 2, 2])
#scores可以是模型預測結果(Label)
#scores也可以是模型預測的confidence(softmax probability)
scores = np.array([1, 1, 2, 2])
scores = np.array([0.1, 0.4, 0.35, 0.8])
area_under_curve = roc_auc_score(y, scores)

e96031413

 def eval(loader, gt_labels_t, output_file="output.txt"):
    G.eval()  # 特徵提取器
    F1.eval() # 分類器
    size = 0
    correct = 0
    
    y_pred=[]
    y_true=[]
    pred_prob = None
    pred_result = None

e96031413

import torch.nn.functional as F
import torch
import numpy as np
from torch.autograd import Function
import torch.nn as nn
from pdb import set_trace as breakpoint
import sys
import math
from torch.nn.parameter import Parameter
from torch.nn import init

e96031413

# LabelSmoothing.py

# https://www.aiuai.cn/aifarm1333.html 示例 3
# From: Github - NVIDIA/DeepLearningExamples/PyTorch/Classification
# smoothing.py
import torch
import torch.nn as nn

# 一般版本LabelSmoothing
class LabelSmoothing(nn.Module):

e96031413

# https://stackoverflow.com/a/45925049/13369757
import matplotlib.pyplot as plt 
import numpy as np

fig, host = plt.subplots(figsize=(12,10)) # (width, height) in inches
    
par1 = host.twinx()
par2 = host.twinx()

host.set_xlabel("Threshold")