https://github.com/soumik12345/gpt

A minimal implementation of Generative Pre-Training or GPT
https://github.com/soumik12345/gpt

deep-learning gpt-2 keras nlp openai tensorflow

Last synced: 19 days ago
JSON representation

A minimal implementation of Generative Pre-Training or GPT

• Host: GitHub
• URL: https://github.com/soumik12345/gpt
• Owner: soumik12345
• Created: 2020-10-29T22:56:30.000Z (about 4 years ago)
• Default Branch: master
• Last Pushed: 2020-11-07T18:16:52.000Z (about 4 years ago)
• Last Synced: 2024-12-25T05:51:52.981Z (about 1 month ago)
• Topics: deep-learning, gpt-2, keras, nlp, openai, tensorflow
• Language: Jupyter Notebook
• Homepage: https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf
• Size: 146 KB
• Stars: 3
• Watchers: 2
• Forks: 3
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# GPT (Ongoing)

Tensorflow implementation of Generative Pre-Training on GPT.

![](./assets/gpt.png)

# Experiments

## Language Model

![](./assets/text_entailment.png)

```python

from gpt.experiments.utils import init_wandb

from gpt.experiments.language_model import IMDBReviewLanguageExperiment

experiment = IMDBReviewLanguageExperiment()

init_wandb(

    project_name='gpt', experiment_name='imdb_language_model',

    wandb_api_key='69696969696969696969696969696969696969696'

)

experiment.build_dataset('https://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz')

experiment.compile()

start_text = 'the actor was'

start_tokens = experiment.tokenize(start_text=start_text)

experiment.train(

    epochs=30, start_tokens=start_tokens,

    max_length=100, max_tokens=40, top_k=10,

    infer_every=1, log_on_wandb=True

)

```