Rfdetr Large 2026 Mlx Fp32

--- license: apache-2.0 library_name: mlx pipeline_tag: object-detection tags: - rf-detr - object-detection - mlx - mlx-swift - apple-silicon --- # RF-DETR Large (2026) — MLX, fp32 MLX-format conversion of Roboflow's **RF-DETR Large (2026)** object-detection model, for use with [mlx-swift-rf-detr](https://github.com/mnmly/mlx-swift-rf-detr) on Apple Silicon. - **Upstream:** [roboflow/rf-detr](https://github.com/roboflow/rf-detr) (Apache-2.0), checkpoint `rf-detr-large-2026.pth` - **Task:** object detection — COCO, 90 classes - **Resolution:** 704 · **decoder layers:** 4 · **queries:** 300 · **backbone:** `dinov2_windowed_small` (hidden 256) - **Precision:** fp32 - **Files:** `config.json` + `preprocessor_config.json` + `model.safetensors` ## Conversion Converted from the upstream PyTorch checkpoint with [`Scripts/convert_detection.py large-2026`](https://github.com/mnmly/mlx-swift-rf-detr/blob/main/Scripts/convert_detection.py). Config values are read from the upstream `RFDETRLargeConfig`; weights keep their `model.` prefix and NCHW conv layout (the Swift loader transposes conv weights to NHWC at load). ## Usage (Swift) ```swift import MLXRFDETR let dir = URL(fileURLWithPath: "rfdetr-large-2026-mlx-fp32") // this repo's downloaded files let predictor = try MLXRFDETR.fromPretrained(dir, scoreThreshold: 0.3, nmsThreshold: 0.5) let result = try predictor.predict(url: URL(fileURLWithPath: "image.jpg")) ``` ## License Apache 2.0 — see [`LICENSE`](LICENSE). Model architecture and weights © Roboflow, Inc. This repository redistributes a format-converted copy of the upstream Apache-2.0 weights; no weights were retrained or modified beyond the PyTorch→MLX serialization.

Source context: 5 downloads · 0 likes · Pipeline object-detection · Library mlx · Repo mnmly/rfdetr-large-2026-mlx-fp32