Issues

• save_pretrained_gguf fails for quantized bases and non-Llama models: proposal to export via llama.cpp (follow-up to #3)

  ARahim3/mlx-tune#21 · Filter by author ghchinoy opened on Sep 27, 2026
  1

• NameError in CPTTrainer._train_with_decoupled_lr when include_embeddings=True (missing local imports from mlx_tune._perf)

  ARahim3/mlx-tune#19 · Filter by author CPiersigilli opened on Aug 21, 2026
  1

• "ValueError: [Primitive::vjp] Not implemented for CustomKernel" on mlx-vlm>=0.6.0

  ARahim3/mlx-tune#18 · Filter by author pizza0rodeo opened on Jun 29, 2026
  1

• Feature Request: Add SenseVoice/Paraformer for STT fine-tuning

  ARahim3/mlx-tune#16 · Filter by author LauraGPT opened on May 31, 2026
  2

• model.save_pretrained_merged("merged", tokenizer) Not functioning properly

  ARahim3/mlx-tune#15 · Filter by author SKillSeven opened on May 25, 2026
  1

• The training examples is wrong for OCR.

  ARahim3/mlx-tune#14 · Filter by author anonymousz97 opened on Apr 16, 2026
  5

• Your project is featured in Awesome MLX

  ARahim3/mlx-tune#11 · Filter by author raullenchai opened on Mar 26, 2026
  1

• Add support for embedding model tuning.

  ARahim3/mlx-tune#10 · Filter by author palindsay opened on Mar 23, 2026
  2

• config.json and adapter_config.json broken

  ARahim3/mlx-tune#9 · Filter by author Turro75 opened on Mar 21, 2026
  1

• qwen 3.5 vision finetuning

  ARahim3/mlx-tune#7 · Filter by author dcrey7 opened on Mar 13, 2026
  1

• saving model as GGUF file is not working

  ARahim3/mlx-tune#3 · Filter by author aykutcayir34 opened on Jan 11, 2026
  5