rtdetr
Real-time object detection you can ship: PyTorch training, OpenVINO inference, 100% Apache-2.0 — code and pretrained weights alike.
Install
pip install rtdetr # inference
pip install "rtdetr[train]" # + training
Detect
from rtdetr import RTDETR
model = RTDETR("rtdetr-r18") # COCO weights, downloaded on first use
r = model("bus.jpg", conf=0.5)[0]
r.boxes.xyxy, r.boxes.conf, r.boxes.cls # plain numpy
r.names[int(r.boxes.cls[0])] # "person"
r.plot(); r.save(); r.show()
Point it at an image, a folder, a glob, a video, an RTSP stream, a webcam index, or a numpy array. Long sources stream:
for r in model.predict(0, stream=True, show=True): # webcam, q or Esc quits
print(r.boxes.xyxyn)
model.track("clip.mp4") # adds r.boxes.id
Label
rtdetr label source=images/ names=can,bottle # browser tool, auto-label + fix
Train
model = RTDETR("rtdetr-r18")
model.train(data="data.yaml", epochs=100, imgsz=640, batch=8, device=0)
model.val(data="data.yaml").box.map50
model.export(format="openvino", half=True) # IR + labels.txt
# data.yaml
path: /data/cans
train: images/train
val: images/val
names: {0: can, 1: bottle}
Command line
rtdetr predict model=rtdetr-r18 source=bus.jpg conf=0.5
rtdetr train model=rtdetr-r18 data=data.yaml epochs=100
rtdetr val model=best.pt data=data.yaml
rtdetr export model=best.pt format=openvino half=true
rtdetr label source=images/ names=can,bottle
Models
| name | params | COCO AP | notes |
|---|---|---|---|
rtdetr-r18 |
20M | 46.4 | fastest |
rtdetr-r34 |
31M | 48.9 | |
rtdetr-r50 |
43M | 53.1 | most accurate |
Docs
- Using the model — sources, results, tracking, saving
- Labelling — the browser tool, auto-labelling
- Training — datasets, validation, export, CLI
- Weights — the mirror, building it, offline use
- Performance — measured speeds and how to improve them
- Design — architecture and provenance
- Development — tests, releases
한국어
from rtdetr import RTDETR
model = RTDETR("rtdetr-r18") # COCO 사전학습 가중치 자동 다운로드
model("bus.jpg", conf=0.5)[0].save() # 결과 이미지 저장
model.train(data="data.yaml", epochs=100) # 내 데이터로 학습
model.export(format="openvino", half=True) # 배포용 IR + labels.txt
가중치 캐시는 ~/.rtdetr/, 사내 미러는 RTDETR_ASSETS_URL 환경변수로 지정합니다.
License
1 maintainer
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
rtdetr-0.4.0.tar.gz
(66.2 kB
view details)
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
rtdetr-0.4.0-py3-none-any.whl
(60.4 kB
view details)
File details
Details for the file rtdetr-0.4.0.tar.gz.
File metadata
- Download URL: rtdetr-0.4.0.tar.gz
- Upload date:
- Size: 66.2 kB
- Tags: Source
- Uploaded using Trusted Publishing? Yes
- Uploaded via:
twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
90b2365c4ea45232600420d4c0ff0460186e65d5c2ec850b0b0d98fc8075828c
|
|
| MD5 |
cac5257f19c65957a70b3cd007b4db4b
|
|
| BLAKE2b-256 |
158cf6f6c8b281d97d6df03be1e6124e356fdec106d9adb8790e7763b02e560e
|
Provenance
The following attestation bundles were made for rtdetr-0.4.0.tar.gz:
Publisher:
publish.yml on leeyunjai82/rtdetr
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
rtdetr-0.4.0.tar.gz -
Subject digest:
90b2365c4ea45232600420d4c0ff0460186e65d5c2ec850b0b0d98fc8075828c - Sigstore transparency entry: 2754264882
- Sigstore integration time:
-
Permalink:
leeyunjai82/rtdetr@3787637d3959033d58eb5c77204a4c0dfc954444 -
Branch / Tag:
refs/heads/main - Owner: https://github.com/leeyunjai82
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@3787637d3959033d58eb5c77204a4c0dfc954444 -
Trigger Event:
workflow_dispatch
-
Statement type:
File details
Details for the file rtdetr-0.4.0-py3-none-any.whl.
File metadata
- Download URL: rtdetr-0.4.0-py3-none-any.whl
- Upload date:
- Size: 60.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? Yes
- Uploaded via:
twine/7.0.0 CPython/3.13.14
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
3b12c7d6de9bf7457ba51570e7181fec2809092499422ececbabdbc1cc3dc696
|
|
| MD5 |
f2a7a1834cb6043fbfefd73271c06710
|
|
| BLAKE2b-256 |
0a163938abd1beeab2b947ece9bb5505e6e9f17b3ef12f2b2212967f884e77f9
|
Provenance
The following attestation bundles were made for rtdetr-0.4.0-py3-none-any.whl:
Publisher:
publish.yml on leeyunjai82/rtdetr
-
Statement:
-
Statement type:
https://in-toto.io/Statement/v1 -
Predicate type:
https://docs.pypi.org/attestations/publish/v1 -
Subject name:
rtdetr-0.4.0-py3-none-any.whl -
Subject digest:
3b12c7d6de9bf7457ba51570e7181fec2809092499422ececbabdbc1cc3dc696 - Sigstore transparency entry: 2754264986
- Sigstore integration time:
-
Permalink:
leeyunjai82/rtdetr@3787637d3959033d58eb5c77204a4c0dfc954444 -
Branch / Tag:
refs/heads/main - Owner: https://github.com/leeyunjai82
-
Access:
public
-
Token Issuer:
https://token.actions.githubusercontent.com -
Runner Environment:
github-hosted -
Publication workflow:
publish.yml@3787637d3959033d58eb5c77204a4c0dfc954444 -
Trigger Event:
workflow_dispatch
-
Statement type:
