sayef

import math
from torch import nn, Tensor

class PositionalEncoding(nn.Module):
    def __init__(self, embedding_dim: int, dropout: float = 0.1, max_len: int = 5000):
        super().__init__()
        self.dropout = nn.Dropout(p=dropout)

        position = torch.arange(max_len).unsqueeze(1)
        div_term = torch.exp(

sayef

• If the bucket is not full, add the record.
• Otherwise, split the bucket.
  • Allocate a new leaf and move half the bucket elements to the new bucket.
  • Insert the new leaf's smallest key and address into the parent.
  • If the parent is full, split it too.
    • Add the middle key to the parent node.
  • Repeat until a parent is found that need not split.
• If the root splits, create a new root that has one key and two pointers. (That is, the value that gets pushed to the new root gets removed from the original node)

sayef

import hashlib as hash

# Specify how many bytes of the file you want to open at a time
BLOCKSIZE = 65536

sha = hash.sha256()
with open('kali.iso', 'rb') as kali_file:
    file_buffer = kali_file.read(BLOCKSIZE)
    while len(file_buffer) > 0:
        sha.update(file_buffer)

sayef

1. create a ssh-proxy ssh -D 9999 -qCN user@server.net

2. make git connect through the ssh-proxy

• [current script only] export GIT_SSH_COMMAND='ssh -o ProxyCommand="connect -S 127.0.0.1:9999 %h %p"'
• OR [git global setting] git config --global core.sshCommand 'ssh -o ProxyCommand="connect -S 127.0.0.1:9999 %h %p"'
• OR [one-time only use] git clone -c=core.sshCommand 'ssh -o ProxyCommand="connect -S 127.0.0.1:9999 %h %p"' git@github.com:user/repo.git

sayef

{
    "cardinal number": [
        "A Russian diver has found the bodies of three of the [E] 118 [/E] sailors who were killed when the nuclear submarine Kursk sank in the Barents Sea .",
        "Navy officials do not expect that all [E] 118 [/E] bodies can be recovered .",
        "In the latest daily CNN USA Today Gallup popular opinion survey , George W Bush , the Governor of Texas , the Republican nominee , holds a 49 % to 42 % advantage over his Democratic opponent Al Gore , the Vice President , but [E] two [/E] other major surveys gave Mr. Bush only a one - point lead .",
        "In the latest daily CNN USA Today Gallup popular opinion survey , George W Bush , the Governor of Texas , the Republican nominee , holds a 49 % to 42 % advantage over his Democratic opponent Al Gore , the Vice President , but two other major surveys gave Mr. Bush only a [E] one [/E] - point lead .",
        "The paper concludes that there is [E] one [/E] candidate who will do that and that is George W Bush .",
        "The bat

sayef

1. Convert list to bytes and get back list again:

import numpy as np
from io import BytesIO

def list2bytes(v):
    np_array = np.array(v)
    np_bytes = BytesIO()
 np.save(np_bytes, np_array, allow_pickle=True)

sayef

LATEST=$(curl -s https://github.com/$1/releases | grep -ioE $1"/archive/refs/tags/LMDB_.*tar.gz" | head -n 1)
wget https://github.com/$LATEST

sayef

# add this file in .vscode folder as compile.sh
# edit ~/.local/bin/ if it's different in your machine

FILE=$(wslpath $1)
EXT=$(rev <<< "$FILE" | cut -d"." -f1 | rev)

if [ "$EXT" == "py" ]; then
    echo "Running autopep8 and autoflake on $FILE $EXT"
    ~/.local/bin/autopep8 -i -a -a $FILE
    ~/.local/bin/autoflake -i --remove-all-unused-imports --remove-unused-variables $FILE

sayef

# specify your svg element and run this in chrome console
var svg = document.getElementsByTagName("svg")[0];
var serializedSVG = new XMLSerializer().serializeToString(svg);
var base64Data = window.btoa(serializedSVG);
var dataURL = "data:image/svg+xml;base64," + base64Data 
var dl = document.createElement("a");
document.body.appendChild(dl); // This line makes it work in Firefox.
dl.setAttribute("href", dataURL);
dl.setAttribute("download", "test.svg");
dl.click();

sayef

"""
Compute score for decoded text in a CTC-trained neural network using TensorFlow:
1. decode text with best path decoding (or some other decoder)
2. feed decoded text into loss function
3. loss is negative logarithm of probability

Example data: two time-steps, 2 labels (0, 1) and the blank label (2). 
Decoding results in [0] (i.e. string containing one entry for label 0).
The probability is the sum over all paths yielding [0], these are: [0, 0], [0, 2], [2, 0]
with probability