License: MIT Hugging Face Parameters Format

This is a transformer-based GPT model.

Trained from scratch ( 10 tokens per parameter, according to Chinchilla rules, the optimal ratio is 20 tokens per parameter. )

Trained on texts, 500 million tokens (and TinyStoryes tokens)

πŸ’¬ Example Conversations

Temperature: 0.5


You: Once upon a tree and I 

AI: had a big, red apple. It was very pretty and sweet. Many people in the forest were there to eat the apple. One day, a little girl came to the tree. She saw the apple and wanted to eat it.
The little girl said, "I can help you!" The little girl said, "Thank you, apple! You are very kind." The little girl gave the apple to the little girl. She said, "You are very kind. But be careful not to bite you." The little girl was happy and ate the apple. She knew that she could have a tasty treat for the little girl.

...

You: Once upon

 a time, there was a little boy named Tim. He loved to play with his toy car. One day, he saw a big box in the park. He wanted to see what was inside.
Tim went to the box and opened it. Inside, he found a big, red ball. He thought, "I can find this ball!" So, he picked up the ball and ran to the box. Inside, he found a small, shiny ball. He was so happy!
But then, something unexpected happened. The ball started to bounce! It bounced back and forth! Tim was surprised but happy. He played with the
...

use.py

Downloads last month
-
Safetensors
Model size
50M params
Tensor type
F32
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for VDC-team/VDrontGPT50m-Storyteller

Finetuned
(1)
this model

Dataset used to train VDC-team/VDrontGPT50m-Storyteller

Collection including VDC-team/VDrontGPT50m-Storyteller