You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This repository contains code that implements GPT-2 with tensorflow / pytorch / jax.
3
+
This repository provides a Python implementation of GPT-2 from scratch, along with the ability to load pre-trained weights provided by OpenAI. This allows you to generate text using the power of GPT-2 without relying on external libraries.
4
+
5
+
### Features
6
+
7
+
* GPT-2 implemented from scratch: Understand the inner workings of GPT-2 by exploring its implementation using basic Python libraries like NumPy.
8
+
* Load OpenAI pre-trained weights: Utilize the pre-trained weights provided by OpenAI to generate high-quality text without extensive training.
9
+
* Command-line interface: Interact with the model easily using a simple command-line interface to input your starting text and generate continuations.
5
10
6
11
A detailed explanation of the code is posted on Medium.
7
12
8
-
[Create your own GPT and generate text with OpenAI’s pre-trained parameters](https://medium.com/@satojkovic/create-your-own-gpt-and-generate-text-with-openais-pre-trained-parameters-8d1632d6c92d)
13
+
[Create your own GPT and generate text with OpenAI’s pre-trained parameters](https://medium.com/@satojkovic/create-your-own-gpt-and-generate-text-with-openais-pre-trained-parameters-8d1632d6c92d)
0 commit comments