braingineer

braingineer

SSH agent forwarding is great. It allows you to ssh from one server to
another all the while using the ssh-agent running on your local
workstation. The benefit is you don't need to generate ssh key pairs
on the servers you are connecting to in order to hop around.

When you ssh to a remote machine the remote machine talks to your
local ssh-agent through the socket referenced by the SSH_AUTH_SOCK
environment variable.

So you the remote server you can do something like: 

braingineer

tmux shortcuts & cheatsheet

start new:

tmux

start new with session name:

tmux new -s myname

braingineer

import requests
import argparse

def download_file_from_google_drive(id, destination):
    URL = "https://docs.google.com/uc?export=download"

    session = requests.Session()

    response = session.get(URL, params = { 'id' : id }, stream = True)
    token = get_confirm_token(response)

braingineer

braingineer

def pad1d(tensor, pad):
    # tensor should be in shape (batch, time, feat)
    # pad should be in shape (left, right)
    tensor = tensor.permute(0, 2, 1).contiguous() # get features on first dim since we are padding time
    original_size = tensor.size() # (batch, feat, time)
    final_new_size = (original_size[0], original_size[1], original_size[2] + pad[0] + pad[1])
    temp_new_size = original_size[:2] + (1,) + original_size[2:]
    assert len(temp_new_size) == 4
    tensor = tensor.view(*temp_new_size)
    pad = pad + (0, 0)

braingineer

def describe_recursively(datum, level=0, level_increment=2):
    indent = lambda i: "|"+"."*i+"|"
    info_str = "L{} - TYPE: {:^6} - LEN: {:^5}"
    if isinstance(datum, list):
        print(indent(level) + info_str.format(level, 'list', len(datum)))
        for i, child in enumerate(datum):
            describe_recursively(child, level+level_increment, level_increment)
            if i>5 and len(datum)>10:
                print(indent(level+level_increment)+"  .....   ")
                describe_recursively(datum[-1], level+level_increment, level_increment)

braingineer

braingineer

def matprint(mat, fmt="g"):
    col_maxes = [max([len(("{:"+fmt+"}").format(x)) for x in col]) for col in mat.T]
    for x in mat:
        for i, y in enumerate(x):
            print(("{:"+str(col_maxes[i])+fmt+"}").format(y), end="  ")
        print("")
  
# Try it!      
import numpy as np

braingineer

import eidos
from toolz import partition
import spacy

nlp = spacy.load('en')

def parse(nlp, input_, n_threads, batch_size):
    nlp.matcher = None
    out = []
    for doc in nlp.pipe(input_, batch_size=batch_size, n_threads=n_threads):