Skip to content

Instantly share code, notes, and snippets.

View InnovArul's full-sized avatar
💭
wandering around in the AI space

Arulkumar InnovArul

💭
wandering around in the AI space
39 followers · 13 following
View GitHub Profile
@InnovArul
InnovArul / new_machine.sh
Created August 10, 2018 08:53
installations on new machine
sudo apt-get install zsh
sudo apt-get install git
sudo add-apt-repository ppa:webupd8team/sublime-text-3
sudo apt-get update
sudo apt-get install sublime-text-installer
dropbox
teamviewer
google-chrome
@InnovArul
InnovArul / pytorch_shared_module_weights.py
Created September 30, 2018 00:10
To validate if the shared module weight is still shared while saving and loading state_dict()
import torch
import torch.nn as nn
class SubModule(nn.Module):
def __init__(self, embedding):
super(SubModule, self).__init__()
self.embedding = embedding
self.fc = nn.Linear(200, 200)
@InnovArul
InnovArul / cuda-extension-pytorch1.0-compile.sh
Created October 9, 2018 09:17
Compilation log for pytorch 1.0 lltm extension in the tutorial
running install
running bdist_egg
running egg_info
creating lltm_cuda.egg-info
writing lltm_cuda.egg-info/PKG-INFO
writing dependency_links to lltm_cuda.egg-info/dependency_links.txt
writing top-level names to lltm_cuda.egg-info/top_level.txt
writing manifest file 'lltm_cuda.egg-info/SOURCES.txt'
reading manifest file 'lltm_cuda.egg-info/SOURCES.txt'
writing manifest file 'lltm_cuda.egg-info/SOURCES.txt'
@InnovArul
InnovArul / mars_split_bug.py
Last active January 8, 2019 11:07
Code to find Mars person reid data gallery-query split bug
# taken and improvised from : https://github.com/jiyanggao/Video-Person-ReID/issues/6
import numpy as np
from scipy.io import loadmat
q = loadmat('./mars/info/query_IDX.mat')['query_IDX'][0]
t = loadmat('./mars/info/tracks_test_info.mat')['track_test_info']
query_inds = q - 1 # to get 0 based predefined array indices for query instances
@InnovArul
InnovArul / filter_learning.py
Created February 16, 2019 08:14
pytorch forum (Model parameters are not being updated?)
from __future__ import print_function
import torch.nn.functional as F
# from torch.autograd import Variable
# import copy
import torch
import torch.nn as nn
import torch.optim as optim
from torch.utils.data import DataLoader
from torchvision import datasets, transforms
@InnovArul
InnovArul / inception-v3-query.py
Created June 1, 2019 06:43
Pytorch inception-v3 query from https://discuss.pytorch.org/t/pytorch-inception-v3/46681
from __future__ import print_function
from __future__ import division
import torch
import torch.nn as nn
import torch.optim as optim
import numpy as np
import torchvision
from torchvision import datasets, models, transforms
from torch.autograd import Variable
import matplotlib.pyplot as plt
@InnovArul
InnovArul / imgaug_segfault.py
Created January 1, 2020 11:22
import imgaug as ia
import imgaug.augmenters as iaa
print("loaded")
prob = 0.5
sometimes = lambda aug: iaa.Sometimes(prob, aug)
output_shape=(473,473)
seq = iaa.Sequential([
# apply the following augmenters to most images
iaa.Fliplr(0.5),
sometimes(iaa.Affine(
@InnovArul
InnovArul / drive_script_gen.ipynb
Created July 25, 2020 11:51 — forked from immuntasir/drive_script_gen.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@InnovArul
InnovArul / tensor_dataset.py
Last active October 11, 2020 15:13
def get_dataloader_from_pth(path, batch_size=4):
contents = torch.load(path)
dataset = torch.utils.data.TensorDataset(contents['x'], contents['y'])
dataloader = torch.utils.data.DataLoader(dataset, batch_size=batch_size,
shuffle=True, num_workers=2)
return dataloader
#----------------------------------------------------------------------
import os.path as osp
@InnovArul
InnovArul / freeze_test.py
Created December 13, 2020 16:20
https://discuss.pytorch.org/t/freezing-a-network-problem/105938
import torch, torch.nn as nn
import torch.nn.functional as F
import os, sys
import copy
import torch.optim as optim
class ConvReLU(nn.Module):
def __init__(self, indim, outdim):
super().__init__()
self.conv = nn.Conv2d(indim, outdim, kernel_size=1)