| --- |
| library_name: mlx |
| license: mit |
| pipeline_tag: image-to-image |
| base_model: microsoft/Mage-Flow-Edit-Base |
| tags: |
| - mlx |
| - mlx-swift |
| - diffusion |
| - rectified-flow |
| - mage-flow |
| --- |
| |
| # Mage-Flow-Edit-Base — MLX-Swift |
|
|
| Apple-Silicon MLX-Swift support for |
| [microsoft/Mage-Flow-Edit-Base](https://huggingface.co/microsoft/Mage-Flow-Edit-Base) |
| (edit, defaults: 30 steps / cfg 5.0). One of six checkpoints of the Mage-Flow family, |
| all sharing one architecture — this repo carries the port's support artifact; |
| the component weights come from the upstream microsoft repo (MIT) and are |
| **not re-hosted** (multi-component diffusion pipeline → xocialize namespace per |
| mlx-community conventions). |
|
|
| Code, parity gates, and full porting notes: |
| **[github.com/xocialize/mage-flow-swift](https://github.com/xocialize/mage-flow-swift)**. |
| The DiT/VAE/Gaussian-Shading watermark/scheduler are parity-locked against the |
| PyTorch oracle; T2I and CFG paths are gated per-step and validated with renders |
| at the real defaults (`sample_512.png` is this checkpoint's). |
|
|
| ## Use |
|
|
| ```bash |
| hf download microsoft/Mage-Flow-Edit-Base --local-dir Mage-Flow-Edit-Base |
| hf download microsoft/Mage-Flow-Edit-Turbo --include "text_encoder/*" --include "vae/*" --local-dir Mage-Flow-Edit-Base # shared components (byte-identical family-wide) |
| hf download xocialize/Mage-Flow-Edit-Base-mlx folded_adaln.safetensors --local-dir Mage-Flow-Edit-Base |
| swift run -c release mage-flow-edit --repo Mage-Flow-Edit-Base --ref dog.jpg --prompt "make the background a snowy forest" --out out.png --steps 30 --cfg 5.0 |
| ``` |
|
|
| Notes: the mandatory upstream RAI content filter is wired in (fail-closed); the |
| Gaussian-Shading provenance watermark is reproduced bit-exactly; the DiT runs |
| bf16 with one GEMM row-chunked around mlx-swift's NAX split-K JIT bug |
| ([mlx#3797](https://github.com/ml-explore/mlx/issues/3797)) until a release |
| ships [#3810](https://github.com/ml-explore/mlx/pull/3810). |
|
|
| ## Pre-quantized DiT (no bf16 peak) |
|
|
| | file | recipe | gate | |
| |---|---|---| |
| | `transformer-int8.safetensors` | int8 g32 affine, attn+MLP Linears only | `int8 quant-vs-bf16 0.999917 (family threshold 0.9998) -> PASS` | |
| | `transformer-int4.safetensors` | int4 g64 affine, same scope | `int4 quant-vs-bf16 0.997226 (family threshold 0.9900) -> PASS` | |
|
|
| Modulation/io/norm layers and the trailing block 11 stay bf16 (the final norm |
| amplifies last-block quant error ~1.6× into `proj_out`). Files are |
| self-describing (metadata carries bits / group size / kept blocks) and load |
| directly with **no bf16 peak** — int8 gate: quant-vs-bf16 per-pass, family threshold ≥ 0.9998; int4 gate: quant-vs-bf16 per-pass, family threshold ≥ 0.99. |
| Quantization is weight-only: the mandatory content filter and the bit-exact |
| Gaussian-Shading watermark are unaffected. Measured t2i peak GPU @1024²: |
| bf16 19.6 → int8 17.2 → int4 16.1 GB. |
|
|
| ```bash |
| # quant tier: skip the 7.7 GB bf16 transformer entirely |
| hf download microsoft/Mage-Flow-Edit-Base --include "text_encoder/*" --include "vae/*" --local-dir Mage-Flow-Edit-Base |
| hf download xocialize/Mage-Flow-Edit-Base-mlx folded_adaln.safetensors --local-dir Mage-Flow-Edit-Base |
| hf download xocialize/Mage-Flow-Edit-Base-mlx transformer-int4.safetensors --local-dir Mage-Flow-Edit-Base |
| swift run -c release mage-flow-edit --repo Mage-Flow-Edit-Base --dit-quant Mage-Flow-Edit-Base/transformer-int4.safetensors ... |
| ``` |
|
|
| MLXEngine: `MLXMageFlow` (same repo) wraps the family as `MageFlowT2IPackage` |
| / `MageFlowEditPackage` — quant tiers auto-materialize these files via |
| `WeightSourcing`. |
|
|
