Software Used

• Data Generation: https://chatgpt.com

• Video Generation: https://www.canva.com

• YouTube Scheduling: https://www.make.com/en/register?pc=cmyong

ChatGPT Prompts

Generate data

Generate 10 sets of intriguing and thought-provoking facts based on the style of the examples given. Arrange these facts in a table with six columns: Question, Fact 1, Fact 2, Fact 3, Fact 4, and Fact 5. Make sure each fact is unique and does not repeat. The focus will be on dating facts. Each set should start with a question and be followed by five different facts. Use simple language that anyone can understand, suitable for a reading level below grade 7. Here are three examples:

5 clear signs someone likes you a lot
Fact 1: 1. They subtly touch you...
Fact 2: 2. Their eyebrows raise suddenly at you...
Fact 3: 3. They lean towards you often, sit next to you, and remove barriers between you...
Fact 4: 4. They ask about you and want to know more about you...
Fact 5: 5. They smile a lot around you and try to make you smile...

5 signs he like you (School Edition)
Fact 1: 1. He fixes his hair a lot...
Fact 2: 2. Waits for you to leave class first so he can walk behind you...
Fact 3: 3. He stares when you aren't looking...
Fact 4: 4. Looks at you when you walk into class or when he walks in...
Fact 5: 5. He teases you when getting told off by the teacher.

5 Dangerous words used by girls when she's angry (Boys take note)
Fact 1: 1. I'm fine...
Fact 2: 2. Your wish...
Fact 3: 3. Please shut up...
Fact 4: 4. I don't care...
Fact 5: 5. Leave me alone.

Regenerate data

Generate 10 more using the same specifications, make sure you don't repeat previous ones.

Make

Sign up your free account here

Initial Population Workflow

Download here

Sample Json

Download here

YouTube Scheduling Workflow

Download here

Resources

1. YouTube Shorts monetization policies

2. Connecting to Google services using custom OAuth client

3. YouTube Shorts Safe Zone

   Subscribe to my newsletter where I share things that inspire me!

   
   

1. Go to https://github.com/KwaiVGI/LivePortrait

2. Click on Code and then download Zip

1. Then create a folder called e.g. LivePortrait in place like Desktop or Documents. Then, copy all the content of the downloaded files into this folder

Step 2 - Download pretrained weights

1. Go to this link based on the Google Drive link below in https://github.com/KwaiVGI/LivePortrait

1. Download both folder into your local computer

3. Go to the LivePortrait folder you created in Step 1 and copy both of the folders into pretrained_weights folder

Step 3 - Create conda environment

We will create a Python conda environment to run the code. For those of you who does not now what conda environment is, you can just imagine that it’s a place for you to install all the dependencies/software required to run the LivePortrait software.

1. Install Miniconda from https://docs.anaconda.com/miniconda/#miniconda-latest-installer-links

2. Create conda environment, the environment will be created at the default directory of conda (e.g. /opt/miniconda3/envs/LivePortrait for macOS or C:\Users\<YourUsername>\Miniconda3\envs\LivePortrait for Windows)

# create env using conda
conda create -n LivePortrait python=3.9

1. Activate conda environment

conda activate LivePortrait

1. Navigate to the LivePortrait directory and install requirements in your conda environment

# for Linux and Windows users
pip install -r requirements.txt

# for macOS with Apple Silicon users
# Edit the requirements_macOS.txt to change the version of onnxruntime-silicon==1.16.3 to onnxruntime-silicon==1.16.0
cd /Users/chenming/Documents/LivePortrait
pip install -r requirements_macOS.txt

Step 4 - Install ffmpeg

macOS

1. Install homebrew at https://github.com/Homebrew/brew/releases/tag/4.3.12

1. Install ffmpeg using Terminal, just open the Terminal app

Open Terminal

# Install ffmpeg using brew
brew install ffmpeg

# Verify that ffmpeg is installed
ffmpeg

Windows

1. Install by following this step here: https://phoenixnap.com/kb/ffmpeg-windows

Step 5 - Open web interface

# For Linux and Windows users
cd C:\Users\chenming\Documents\LivePortrait
python app.py

# for macOS with Intel
cd /Users/chenming/Documents/LivePortrait
python app.py

# For macOS with Apple Silicon users
cd /Users/chenming/Documents/LivePortrait
PYTORCH_ENABLE_MPS_FALLBACK=1 python app.py

You should see something like this 👇

Open your web url at:

http://127.0.0.1:8890