language: en
thumbnail: >-
https://www.huggingtweets.com/jeremyphoward-karpathy-ylecun/1620433354561/predictions.png
tags:
- huggingtweets
widget:
- text: My dream is
I was made with huggingtweets.
Create your own bot based on your favorite user with the demo!
How does it work?
The model uses the following pipeline.
To understand how the model was developed, check the W&B report.
Training data
The model was trained on tweets from Andrej Karpathy & Yann LeCun & Jeremy Howard.
| Data | Andrej Karpathy | Yann LeCun | Jeremy Howard |
|---|---|---|---|
| Tweets downloaded | 3217 | 3249 | 3246 |
| Retweets | 426 | 940 | 1437 |
| Short tweets | 98 | 185 | 154 |
| Tweets kept | 2693 | 2124 | 1655 |
Explore the data, which is tracked with W&B artifacts at every step of the pipeline.
Training procedure
The model is based on a pre-trained GPT-2 which is fine-tuned on @jeremyphoward-karpathy-ylecun's tweets.
Hyperparameters and metrics are recorded in the W&B training run for full transparency and reproducibility.
At the end of training, the final model is logged and versioned.
How to use
You can use this model directly with a pipeline for text generation:
from transformers import pipeline
generator = pipeline('text-generation',
model='huggingtweets/jeremyphoward-karpathy-ylecun')
generator("My dream is", num_return_sequences=5)
Limitations and bias
The model suffers from the same limitations and bias as GPT-2.
In addition, the data present in the user's tweets further affects the text generated by the model.
About
Built by Boris Dayma
For more details, visit the project repository.
