Image Segmentation
MLX
English
native-mlx
apple-silicon
medical-imaging
mri
brain-tumor
segmentation
research
Instructions to use MedMLX/segresnet-brats-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use MedMLX/segresnet-brats-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download MedMLX/segresnet-brats-mlx --local-dir segresnet-brats-mlx
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download native-run.json from MedMLX/segresnet-brats-mlx: direct link, hf CLI and curl.
- Browser
- Download file 7.32 kB
-
https://huggingface.co/MedMLX/segresnet-brats-mlx/resolve/main/native-run.json
- Command line
-
hf download hf://MedMLX/segresnet-brats-mlx/native-run.json
-
curl -L -o native-run.json https://huggingface.co/MedMLX/segresnet-brats-mlx/resolve/main/native-run.json
7.32 kB
| { | |
| "schema_version": "medmlx.run.v1", | |
| "model": "segresnet-brats", | |
| "configuration": "MONAI brats_mri_segmentation (SegResNet, 4-channel MRI)", | |
| "date": "2026-10-08", | |
| "package": { | |
| "name": "segresnet-mlx", | |
| "path": "<repositories>/segresnet-mlx", | |
| "git_revision": "40cbce54523632fbc8cdfc564a9cce48d8df3424", | |
| "branch": "main", | |
| "dirty": false, | |
| "remote": "https://github.com/MedMLX/segresnet-mlx.git" | |
| }, | |
| "environment": { | |
| "machine": "MacBook Pro (2021), Apple M1 Max, 10-core CPU, 24-core GPU, 64 GB", | |
| "macos": "27.0", | |
| "python": "3.12.12", | |
| "mlx_version": "0.32.3", | |
| "device": "gpu" | |
| }, | |
| "command": { | |
| "entry_point": "segresnet_mlx.medmlx:run", | |
| "kwargs": { | |
| "bundle": "brats_mri_segmentation", | |
| "weights_path": "<paper-repo>/work/runs/weights/segresnet-brats.safetensors", | |
| "output_dir": "<paper-repo>/work/runs/segresnet-brats/out", | |
| "t1c_path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/T1gd.nii.gz", | |
| "t1_path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/T1w.nii.gz", | |
| "t2_path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/T2w.nii.gz", | |
| "flair_path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/FLAIR.nii.gz", | |
| "overwrite": true | |
| }, | |
| "invocation": "cd <repositories>/segresnet-mlx && .venv/bin/python <paper-repo>/work/runs/_harness/run_one.py segresnet_mlx.medmlx <kwargs.json> <run_result.json>", | |
| "process": "fresh Python process per model" | |
| }, | |
| "checkpoint": { | |
| "path": "<paper-repo>/work/runs/weights/segresnet-brats.safetensors", | |
| "sha256": "9044dcf3e2a82ef18f14d7abf507debbea52e14560c304b990a4c0e412bc9c30", | |
| "source": "huggingface.co/MONAI/brats_mri_segmentation@370f7f9d models/model.pt (local HF cache)", | |
| "source_sha256": "860ccb3f1c21c99d0410ad8a1ac4ef6b8fab60cec0a503b0ba42675741a750ae", | |
| "registry_sha256_match": true, | |
| "conversion": "segresnet_mlx.convert_checkpoint(spec=BUNDLES['brats_mri_segmentation'].spec)" | |
| }, | |
| "inputs": [ | |
| { | |
| "role": "t1c_path", | |
| "path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/T1gd.nii.gz", | |
| "sha256": "96abcdbb541c4af37f03bef466eee5d11a3f91b9a1ea10b1a4a22cb38b7b3978", | |
| "dataset": "MSD Task01 BrainTumour (training)", | |
| "case_id": "BRATS_001", | |
| "derivation": "channel split of imagesTr/BRATS_001.nii.gz (MSD order FLAIR,T1w,T1gd,T2w)" | |
| }, | |
| { | |
| "role": "t1_path", | |
| "path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/T1w.nii.gz", | |
| "sha256": "740b0adfa8de83fcbdfa5cd5f14689f6d947120bfbf057fe82954c07477bb173", | |
| "dataset": "MSD Task01 BrainTumour (training)", | |
| "case_id": "BRATS_001", | |
| "derivation": "channel split of imagesTr/BRATS_001.nii.gz (MSD order FLAIR,T1w,T1gd,T2w)" | |
| }, | |
| { | |
| "role": "t2_path", | |
| "path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/T2w.nii.gz", | |
| "sha256": "8e2de5df0abd8b25ac1136c6570a6d03f2255107ef1265145b606f07acd1f1af", | |
| "dataset": "MSD Task01 BrainTumour (training)", | |
| "case_id": "BRATS_001", | |
| "derivation": "channel split of imagesTr/BRATS_001.nii.gz (MSD order FLAIR,T1w,T1gd,T2w)" | |
| }, | |
| { | |
| "role": "flair_path", | |
| "path": "<home>/Documents/Codex/2026-10-08/https-drive-google-com-drive-folders/outputs/medmlx/datasets/qualification-2026-10-08/derived/msd-brain-channels/BRATS_001/FLAIR.nii.gz", | |
| "sha256": "7571148ce691dccff5a84109091f43385b7c069a99cd7a967b2002aaeec64cdd", | |
| "dataset": "MSD Task01 BrainTumour (training)", | |
| "case_id": "BRATS_001", | |
| "derivation": "channel split of imagesTr/BRATS_001.nii.gz (MSD order FLAIR,T1w,T1gd,T2w)" | |
| } | |
| ], | |
| "outputs": { | |
| "mask": { | |
| "path": "<paper-repo>/work/runs/segresnet-brats/out/brats_mri_segmentation_mask.nii.gz", | |
| "sha256": "ce7f2ca25eeb2e1dea09e748dfdf1ec666a16296bc523a0b537608575ed5dcd6", | |
| "label_voxel_counts": { | |
| "0": 8791633, | |
| "1": 27348, | |
| "2": 74690, | |
| "4": 34329 | |
| } | |
| }, | |
| "details": { | |
| "labels_present": [ | |
| 0, | |
| 1, | |
| 2, | |
| 4 | |
| ], | |
| "input_shape": [ | |
| 240, | |
| 240, | |
| 155 | |
| ], | |
| "sliding_window": { | |
| "roi_size": [ | |
| 240, | |
| 240, | |
| 160 | |
| ], | |
| "sw_batch_size": 1, | |
| "overlap": 0.5 | |
| }, | |
| "seconds": { | |
| "load_inputs_and_weights": 0.3394, | |
| "preprocess": 0.1134, | |
| "inference": 1.4535, | |
| "postprocess_and_write": 0.2894 | |
| } | |
| } | |
| }, | |
| "checks": [ | |
| { | |
| "name": "mask_reopens", | |
| "result": "pass", | |
| "detail": { | |
| "dtype": "uint8", | |
| "shape": [ | |
| 240, | |
| 240, | |
| 155 | |
| ] | |
| } | |
| }, | |
| { | |
| "name": "finite_integer_labels", | |
| "result": "pass", | |
| "detail": null | |
| }, | |
| { | |
| "name": "shape_matches_input", | |
| "result": "pass", | |
| "detail": { | |
| "input": [ | |
| 240, | |
| 240, | |
| 155 | |
| ], | |
| "mask": [ | |
| 240, | |
| 240, | |
| 155 | |
| ] | |
| } | |
| }, | |
| { | |
| "name": "affine_matches_input", | |
| "result": "pass", | |
| "detail": { | |
| "max_abs_diff": 0.0 | |
| } | |
| }, | |
| { | |
| "name": "labels_in_expected_set", | |
| "result": "pass", | |
| "detail": { | |
| "observed": [ | |
| 0, | |
| 1, | |
| 2, | |
| 4 | |
| ], | |
| "allowed": [ | |
| 0, | |
| 1, | |
| 2, | |
| 4 | |
| ] | |
| } | |
| }, | |
| { | |
| "name": "foreground_nonempty", | |
| "result": "pass", | |
| "detail": { | |
| "foreground_voxels": 136367 | |
| } | |
| }, | |
| { | |
| "name": "required_labels_present", | |
| "result": "pass", | |
| "detail": { | |
| "required": [ | |
| 1, | |
| 2 | |
| ] | |
| } | |
| }, | |
| { | |
| "name": "tumor_overlaps_annotated_region", | |
| "result": "pass", | |
| "detail": { | |
| "whole_tumor_dice_vs_msd_label_informational": 0.8897 | |
| } | |
| } | |
| ], | |
| "duration_s": 2.271, | |
| "timing_boundary": "wall clock around the single run() call: weight loading, preprocessing, MLX inference and export; interpreter start-up and package import excluded; one run, no warmup", | |
| "mlx_peak_gb": 6.954, | |
| "memory_definition": "mx.get_peak_memory() after mx.reset_peak_memory() immediately before run(); decimal GB", | |
| "panel": "paper/figures/runs/segresnet-brats.png", | |
| "status": "success", | |
| "failure_cause": null, | |
| "notes": "BRATS_001 is training-exposed MSD data; overlap with the MSD label is a sanity check, not an accuracy claim.", | |
| "key_check": "labels [1, 2, 4]; whole_tumor_dice_vs_msd_label_informational = 0.8897" | |
| } | |