Skip to content

Instantly share code, notes, and snippets.

View max-kuk's full-sized avatar
🎓
Focusing

Max Kukuškin max-kuk

🎓
Focusing
View GitHub Profile
import torch
from collections import OrderedDict
from typing import List
checkpoints_weights_paths: List[str] = ... # sorted in descending order by score
model: torch.nn.Module = ...
def average_weights(state_dicts: List[dict]):
everage_dict = OrderedDict()
@max-kuk
max-kuk / segment-satellite-images.ipynb
Last active December 18, 2019 10:36
Compare predicted and original masks.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@max-kuk
max-kuk / notebook2.ipynb
Last active December 18, 2019 10:38
Segment satellite images.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@max-kuk
max-kuk / gradientboostingv3.ipynb
Created January 17, 2020 09:19
GradientBoostingV3.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
# Copyright 2020 by Maksim Kukushkin, The Information Systems Institute, University of Leipzig.
# Weather crawler to collect weather data from Wunderground.com
# This file is part of the tools used
# for master thesis "Flight price prediction and Dynamic Pricing in European low-cost airlines"
import csv
import json
import time
from calendar import monthrange
@max-kuk
max-kuk / global_convolution_tf.py
Last active October 20, 2022 14:03
The Global Convolution Network (GCN) Block is essentially a kx1 followed by 1xk convolution summed with a parallely computed 1xk followed by kx1 convolution. This results in a large kxk kernel with dense connections.
class GCN(tf.Module):
def __init__(self, filters, k=7):
super(GCN, self).__init__()
self.padding_l1 = tf.keras.layers.ZeroPadding2D(padding=((k - 1) // 2, 0))
self.conv_l1 = tf.keras.layers.Conv2D(filters, kernel_size=(k, 1))
self.padding_l2 = tf.keras.layers.ZeroPadding2D(padding=(0, (k - 1) // 2))
self.conv_l2 = tf.keras.layers.Conv2D(filters, kernel_size=(1, k))
self.padding_r1 = tf.keras.layers.ZeroPadding2D(padding=((k - 1) // 2, 0))
self.conv_r1 = tf.keras.layers.Conv2D(filters, kernel_size=(1, k))