2024-05-07 Meeting #12

TODO

1. Pabeigt uzdevumus no ieprējā note

2. Ievērojami uzlabot atskaites, pievienot CSV failus un grafikus ar rezultātiem

3. CSV vailos noteikti jāglabā last_test_loss, best_test_loss, last_test_acc, best_test_acc un train mainīgie

   1. Piemēri ar korektām atskaitēm:

      1. https://evalds.notion.site/2022-08-04-Report-5e3e0ed31f59430aa96acd88582cc67f?pvs=4

      2. https://evalds.notion.site/alarm_bicyclegan_2-grid-search-3effe116573e43a2bcad30513284c3dd?pvs=4

4. Implementēt korekti multi-dataloder piemēru

xxxxxxxxxx
30
1
import numpy as np
2

3

4
class DataLoader():
5
    def __init__(self):
6
        self.data = [1, 2, 3, 4, 5]
7
        self.index = 0
8

9
    def __iter__(self):
10
        self.index = 0
11
        np.random.shuffle(self.data)
12
        return self
13

14
    def __next__(self):
15
        if self.index < len(self.data):
16
            result = self.data[self.index]
17
            self.index += 1
18
            return result
19
        else:
20
            raise StopIteration()
21

22
loader_train = DataLoader()
23

24
iter2 = iter(DataLoader())
25
iter3 = iter(DataLoader())
26

27
for each1 in loader_train:
28
    each2 = next(iter2)
29
    each3 = next(iter3)
30
    print(each1, each2, each3)