GDScript training dataset for fine-tuning code models on Godot engine development.
Total samples: 476
Train split: 428
Validation split: 48
Format: JSONL (instruction, input, output)
Language: GDScript (Godot 4.x)
Languages: German instructions, GDScript code
godotengine/godot-demo-projects
GDQuest/godot-open-rpg
GDQuest/godot-3d-dodge-the-creeps
bitbrain/beehave (behavior trees)
limboai/limboai (AI for Godot)… See the full description on the dataset page:
https://huggingface.co/datasets/matzejo/godot-lora-dataset.
