luminousresearch/L0-Luau-4B-Instruct

Training Data

This model was trained on a dataset derived from TorpedoSoftware/Roblox-Luau-Reasoning-v1.0, which is released under the MIT License.

The original authors are not affiliated with or responsible for this model.

Base Model

Base model: Qwen/Qwen3-4B-Instruct-2507

Fine-tuning Method

• —Adapter: DoRA
• —Method: SFT
• —Precision: trained with 4-bit base weights + BF16 compute, then merged to safetensors

Training Details

• —Training time: ~12 hours
• —Hardware: 1x NVIDIA RTX 5060 Ti