https://github.com/stoyan-stoyanov/transformers-calculator
Transformer Calculator - Estimate training time for transformer models.
https://github.com/stoyan-stoyanov/transformers-calculator
llm llm-training transformers
Last synced: 4 months ago
JSON representation
Transformer Calculator - Estimate training time for transformer models.
- Host: GitHub
- URL: https://github.com/stoyan-stoyanov/transformers-calculator
- Owner: stoyan-stoyanov
- License: mit
- Created: 2023-11-30T04:18:04.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2023-11-30T07:12:18.000Z (almost 2 years ago)
- Last Synced: 2025-04-09T06:35:15.520Z (7 months ago)
- Topics: llm, llm-training, transformers
- Language: JavaScript
- Homepage: https://transformers-calculator.vercel.app
- Size: 430 KB
- Stars: 6
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# 👾 Transformers Calculator
Live demo available here: https://transformers-calculator.vercel.app
## Introduction
Transformer Calculator is a web application that can help you estimate the training time for transformer models based on various input parameters. It's built using React and React Bootstrap, and it is inspired from this [blog post](https://medium.com/@dzmitrybahdanau/the-flops-calculus-of-language-model-training-3b19c1f025e4).
If you are developing LLM apps check out [LLMFlows](https://github.com/stoyan-stoyanov/llmflows).
## Features
- Calculate the estimated training time for transformer models.
- Responsive design with React Bootstrap.
- Special message for the GPU rich.
- Input validation and formatted results.
## Getting Started
To run Transformer Calculator locally, follow these steps:
### Prerequisites
- Node.js
- npm
### Running locally
1. Clone the repository:
```bash
git clone git@github.com:stoyan-stoyanov/transformers-calculator.git
cd transformers-calculator
npm install
npm run start
```
2. Open http://localhost:3000 in your browser.
## License
This project is licensed under the MIT Licence - see the LICENSE.md file for details.