InnovArul

curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list
sudo apt-get update
sudo apt-get install -y nvidia-docker2
sudo pkill -SIGHUP dockerd

InnovArul

Answering https://discuss.pytorch.org/t/a-set-of-data-sum-as-the-dividend-how-to-find-grad/182303:

We can find $grad_x = \frac{dz}{dx}$ in the following way:

$$grad_x = \frac{dz}{dx} = \left[\frac{dz}{dx_0},\frac{dz}{dx_1}, \frac{dz}{dx_2}\right]$$

$$ = \frac{dz}{dx} = \left[\frac{dy}{dx_0}\frac{dz}{dy},\frac{dy}{dx_1}\frac{dz}{dy},\frac{dy}{dx_2}\frac{dz}{dy}\right]$$

Now let's concentrate on finding one element $\frac{dz}{dx_0} = \frac{dy}{dx_0}\frac{dz}{dy}$

InnovArul

from turtle import forward
import torch
import torch.nn as nn, torch.nn.functional as F

class CNN(nn.Module):
    def __init__(self) -> None:
        super().__init__()

        self.conv1 = nn.Conv2d(in_channels=1, out_channels=8, kernel_size=6, stride=1, padding=2)
        self.RL1 = nn.ReLU()

InnovArul

import torch, torch.nn as nn
import torch.optim as optim

def print_grads(modules, string):
    print(string)
    for mod in modules:
        for p in mod.parameters():
            print(p.grad)
        print('**')
    print("-----")

InnovArul

command:
--------
sudo useradd -m -d /home/<user> -s /bin/bash -c "<rollnumber>" -U <user>

password:
---------
sudo passwd <user>

Add user to sudo
-----------------

InnovArul

import torch, torch.nn as nn                                                         
import torch.optim as optim, torch.nn.functional as F                                
                                                                                     
class CustomLinearNoWeightDecay(nn.Module):                                                       
    def __init__(self, mask):                                                        
        super().__init__()                                                           
        self.register_buffer("mask", mask)                                           
        out_channels, in_channels = mask.shape                                       
        self.weight = nn.Parameter(torch.randn(out_channels, in_channels))       
        

InnovArul

import torch, torch.nn as nn

class LowlevelModule(nn.Module):
    def __init__(self, custom_val):
        super().__init__()
        self.custom_val = custom_val

    def print_custom_val(self):
        print(self.custom_val.item())

InnovArul

import torchvision, copy
import torch, torch.nn as nn

def reset_all_weights(model: nn.Module) -> None:
    """
    refs:
        - https://discuss.pytorch.org/t/how-to-re-set-alll-parameters-in-a-network/20819/6
        - https://stackoverflow.com/questions/63627997/reset-parameters-of-a-neural-network-in-pytorch
        - https://pytorch.org/docs/stable/generated/torch.nn.Module.html
    """

InnovArul

TORCH_CUDA_ARCH_LIST="5.2 6.0 6.1 7.0 7.5 8.0 8.6+PTX" python setup.py build
python setup.py install

D:\cmder_mini\Cmder.exe "%ActivDir%"

InnovArul

mv Day.. day_  # move Day folder to day_ folder

# move MTS files to root dir 
mv ./01/2019/* .

# to concate MTS files into MP4
ffmpeg -i "concat:$(echo *.MTS | tr ' ' '|')" -strict -2 concat_out.mp4