ak_transcribe 0.0.1

Transcribe Media Files

ak_transcribe

An awesome README template for your projects!

Table of Contents

• 1. About the Project
  • 1.1. Screenshots
  • 1.2. Features
• 2. Getting Started
  • 2.1. Prerequisites
  • 2.2. Dependencies
  • 2.3. Installation
    • 2.3.1. Production
    • 2.3.2. Development
• 3. Usage
  • 3.1. Ollama - Local AI
  • 3.2. Encryption
• 4. Development
• 5. Roadmap
• 6. FAQ
• 7. License
• 8. Contact
• 9. Acknowledgements

1. About the Project

1.1. Screenshots

1.2. Features

• Feature 1
• Feature 2
• Feature 3

2. Getting Started

2.1. Prerequisites

Python 3.11 or above

2.2. Dependencies

Create the virutual environment and install dependencies

  pip install flit  

2.3. Installation

2.3.1. Production

Install with flit

  flit install --deps production

2.3.2. Development

Download the git and install via flit

  git clone https://github.com/rpakishore/Template-Python.git
  cd Template-Python
  pip install flit
  flit install

3. Usage

from ak_transcribe import debug
debug(True) #For debug messages, Can be skipped.

3.1. Ollama - Local AI

Control the locally hosted Ollama module

from ak_transcribe.ollama import Ollama

ollama = Ollama(url='prx-01-ai:11434')

ollama.local                                                #list local models
ollama.copy(source='llama2', destination='llama2-backup')   #Copy an existing local model
ollama.delete(model_name='llama2')                          #Delete model and it's data
ollama.show(model_name='llama2')                            #Show info of local model
ollama.heartbeat()                                          #Check if Ollama is running
ollama.public_models                                        #List public models
ollama.pull(model_name='llama2')                            #pull a model from registry

#generate response
response, context = ollama.generate(
    model_name='mistral',
    prompt='Why is the sky blue in color?')

#create a model
ollama.create(
    model_name='mario', 
    modelfile='FROM llama2\nSYSTEM You are mario from Super Mario Bros.')

3.2. Encryption

Minor encryption/decryption algoritm w/ Fernet. Also supports direct obscuring/unobscuring function

from ak_transcribe.cryptography import obscure, unobscure, encrypt_text, decrypt_text

obscure(text="Hello World")     # Obscure the text without encryption
unobscure(obscured='eNrzSM3JyVcIzy_KSQEAGAsEHQ==')  # Unobscure the text

encrypt_text(text='Hello World!', password='admin123')  # Encrypt w/ Fernet
decrypt_text(encrypted_text='eaX34dw=', password='admin123')  # Decrypt w/ Fernet

4. Development

1. Open the project directory in vscode
2. Update the app name under pyproject.toml
3. Change the folder name from src\ak_transcribe to src\<app_name>, and propate the changes to the subfolders.
4. Review the dependencies under pyproject.toml and remove as needed.
5. Remove unneeded dependencies from src\<app_name>\

5. Roadmap

• Set up a skeletal framework
• Todo 2

6. FAQ

• Question 1

  • Answer 1

• Question 2

  • Answer 2

7. License

See LICENSE for more information.

8. Contact

Arun Kishore - @rpakishore

Project Link: https://github.com/rpakishore/Template-Python

9. Acknowledgements

• Awesome README Template
• Shields.io