OpenAI's new Spark model codes 15x faster than GPT-5.3-Codex - but there's a catch ...
To train the models, we use the run_train.py CLI script. The script supports various arguments. See the training/README.md file for more information, along with the hyperparameters used in the paper.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results