awesome-archive / gpt-2-pytorch

Simple Text-Generator with OpenAI gpt-2 Pytorch Implementation

GPT2-Pytorch with Text-Generator

Better Language Models and Their Implications

Our model, called GPT-2 (a successor to GPT), was trained simply to predict the next word in 40GB of Internet text. Due to our concerns about malicious applications of the technology, we are not releasing the trained model. As an experiment in responsible disclosure, we are instead releasing a much smaller model for researchers to experiment with, as well as a technical paper. from openAI Blog

This repository is simple implementation GPT-2 about text-generator in Pytorch with compress code

• The original repertoire is openai/gpt-2. Also You can Read Paper about gpt-2, "Language Models are Unsupervised Multitask Learners". To Understand more detail concept, I recommend papers about Transformer Model.

• Good implementation GPT-2 in Pytorch which I referred to, huggingface/pytorch-pretrained-BERT, You can see more detail implementation in huggingface repository.

• Transformer(Self-Attention) Paper : Attention Is All You Need(2017)

• First OpenAi-GPT Paper : Improving Language Understanding by Generative Pre-Training(2018)

• See OpenAI Blog about GPT-2 and Paper

Quick Start

1. download GPT2 pre-trained model in Pytorch which huggingface/pytorch-pretrained-BERT already made! (Thanks for sharing! it's help my problem transferring tensorflow(ckpt) file to Pytorch Model!)

$ git clone https://github.com/graykode/gpt-2-Pytorch && cd gpt-2-Pytorch
# download huggingface's pytorch model 
$ curl --output gpt2-pytorch_model.bin https://s3.amazonaws.com/models.huggingface.co/bert/gpt2-pytorch_model.bin
# setup requirements
$ pip install -r requirements.txt

2. Now, You can run like this.

• Text from Book 1984, George Orwell

$ python main.py --text "It was a bright cold day in April, and the clocks were striking thirteen. Winston Smith, his chin nuzzled into his breast in an effort to escape the vile wind, slipped quickly through the glass doors of Victory Mansions, though not quickly enough to prevent a swirl of gritty dust from entering along with him."

3. Also You can Quick Starting in Google Colab

Option

• --text : sentence to begin with.
• --quiet : not print all of the extraneous stuff like the "================"
• --nsamples : number of sample sampled in batch when multinomial function use
• --unconditional : If true, unconditional generation.
• --batch_size : number of batch size
• --length : sentence length (< number of context)
• --temperature: the thermodynamic temperature in distribution (default 0.7)
• --top_k : Returns the top k largest elements of the given input tensor along a given dimension. (default 40)

See more detail option about temperature and top_k in here

Dependencies

• Pytorch 0.41+
• regex 2017.4.5

Author

• Tae Hwan Jung(Jeff Jung) @graykode
• Author Email : [email protected]

License

• OpenAi/GPT2 follow MIT license, huggingface/pytorch-pretrained-BERT is Apache license.
• I follow MIT license with original GPT2 repository

Acknowledgement

Jeff Wu(@WuTheFWasThat), Thomas Wolf(@thomwolf) for allowing referring code.