roneneldan/TinyStories
Viewer • Updated • 2.14M • 83.3k • 1.15k
This repository contains the trained weights for Nano-GPT-OSS, a language model pre-trained completely from scratch.
The core engineering files, scripts, and logs are hosted on GitHub:
Training/ & Preprocessing/: Scripts used to prepare the dataset and pre-train the model.Architecture/ & main.py: The custom neural network definitions.Inference/: Code required to run text generation using these weights.I'll work on these details soon......