🎛️ Finetuner Studio

Low-code fine-tuning on Apple Silicon · powered by mlx-tune · 12 training paradigms, zero boilerplate

🟡 GUI-only mode — mlx-tune requires an Apple Silicon Mac (arm64/macOS).

1 · Choose a task and a base model

Training task
Every mlx-tune trainer is available here.

SFT — Supervised Fine-Tuning

Instruction tuning of chat/completion LLMs on text or conversation data.

  • Backend: mlx_tune.SFTTrainer + SFTConfig
  • Dataset schema: text
Model source
Model
Pick a search result or type any repo id.
256 32768

2 · LoRA adapters

1 256
1 256
0 0.5
Target modules (text models)
Finetuner Studio · load a model → drop a dataset → press train. The generated Python script is yours to keep.