kimbochen

Development Workflow

• Login to Datahub and setup ssh key (00:00):
  • Data Hub
  • Generate a new SSH Key: ssh-keygen -t ed25519 -C "your_email@example.com"
  • Start the SSH Agent: eval "$(ssh-agent -s)"
  • Add the private key to the agent: ssh-add ~/.ssh/id_ed25519
  • Copy the public key: cat ~/.ssh/id_ed25519.pub
  • Add the public key to GitHub (GitHub settings page)
  • Check: ssh -T git@github.com

kimbochen

kimbochen

from collections import defaultdict, deque


def topo_sort(n_nodes, edges):
    # Organize info from edges:
    # 1. In degree of each node
    # 2. Out nodes of each nodes
    in_degree = dict.fromkeys(range(n_nodes), 0)
    out_nodes = defaultdict(set)

kimbochen

#!/usr/local/bin/bash -e

rust_env()
{
    docker run --rm \
        -v $PWD:/app -w /app rust:latest \
        $1
}

rust_env "rustc $1.rs"

kimbochen

version: "3.7"

services:
  app:
    image: rust
    container_name: hello-rust
    working_dir: /app
    volumes:
      - ./:/app
    stdin_open: true

kimbochen

Autograd

This is a didactic version of Autograd based on Doug Maclaurin's Ph.D. Thesis. Only single-variable scalar functions are supported.

Constructing a Node

• Input:

x = Node(value=3.9)

kimbochen

import sys
from collections import namedtuple
from enum import Enum


grid = [
    '...#...O',
    '.###..X.',
    '...#....',
    '.X......',

kimbochen

from pdb import set_trace as st

import torch
import numpy as np
import pandas as pd
from torch import nn, optim
from torch.utils.data import Dataset, DataLoader

DEVICE = 'cuda:1'
torch.manual_seed(3985)