However, we observe that the speed and accuracy of YOLOs are negatively affected by the NMS. Recently, end-to-end Transformer-based detectors (DETRs) have provided an alternative to eliminating NMS. Nevertheless, the high computational cost limits their practicality and hinders them from fully exploiting the advantage of excluding NMS. In this paper, we propose the Real-Time DEtection TRansformer (RT-DETR), the first real-time end-to-end object detector to our best knowledge that addresses the above dilemma. We build RT-DETR in two steps, drawing on the advanced DETR: first we focus on maintaining accuracy while improving speed, followed by maintaining speed while improving accuracy.…
The RT-DETRv2 model was proposed in RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer by Wenyu Lv, Yian Zhao, Qinyao Chang, Kui Huang, Guanzhong Wang, Yi Liu.
Parameters77M
Context—
Weights307.3 MB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads8.5k
Runs On
What it takes to serve rtdetr_v2_r101vd (77M parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 0.2 GB | 0.2 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 0.1 GB | 0.1 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.0 GB | 0.0 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.
Model Card
By Peking University, published under apache-2.0, revision 2c5dbbd2d4d8.
The RT-DETRv2 model was proposed in RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer by Wenyu Lv, Yian Zhao, Qinyao Chang, Kui Huang, Guanzhong Wang, Yi Liu. RT-DETRv2 refines RT-DETR by introducing selective multi-scale feature extraction, a discrete sampling operator for broader deployment compatibility, and improved training strategies like dynamic data augmentation and scale-adaptive hyperparameters. These changes enhance flexibility and practicality while maintaining real-time performance. This model was contributed by @jadechoghari with the help of @cyrilvallez and @qubvel-hf This is RT-DETRv2 consistently outperforms its predecessor across all…
Read Peking University's full model card
RT-DETRv2
Overview
The RT-DETRv2 model was proposed in RT-DETRv2: Improved Baseline with Bag-of-Freebies for Real-Time Detection Transformer by Wenyu Lv, Yian Zhao, Qinyao Chang, Kui Huang, Guanzhong Wang, Yi Liu. RT-DETRv2 refines RT-DETR by introducing selective multi-scale feature extraction, a discrete sampling operator for broader deployment compatibility, and improved training strategies like dynamic data augmentation and scale-adaptive hyperparameters. These changes enhance flexibility and practicality while maintaining real-time performance.
This model was contributed by @jadechoghari with the help of @cyrilvallez and @qubvel-hf
This is
Performance
RT-DETRv2 consistently outperforms its predecessor across all model sizes while maintaining the same real-time speeds.
How to use
import torch
import requests
from PIL import Image
from transformers import RTDetrV2ForObjectDetection, RTDetrImageProcessor
url = 'http://images.cocodataset.org/val2017/000000039769.jpg'
image = Image.open(requests.get(url, stream=True).raw)
image_processor = RTDetrImageProcessor.from_pretrained("PekingU/rtdetr_v2_r101vd")
model = RTDetrV2ForObjectDetection.from_pretrained("PekingU/rtdetr_v2_r101vd")
inputs = image_processor(images=image, return_tensors="pt")
with torch.no_grad():
outputs = model(**inputs)
results = image_processor.post_process_object_detection(outputs, target_sizes=torch.tensor([(image.height, image.width)]), threshold=0.5)
for result in results:
for score, label_id, box in zip(result["scores"], result["labels"], result["boxes"]):
score, label = score.item(), label_id.item()
box = [round(i, 2) for i in box.tolist()]
print(f"{model.config.id2label[label]}: {score:.2f} {box}")
cat: 0.97 [341.14, 25.11, 639.98, 372.89]
cat: 0.96 [12.78, 56.35, 317.67, 471.34]
remote: 0.95 [39.96, 73.12, 175.65, 117.44]
sofa: 0.86 [-0.11, 2.97, 639.89, 473.62]
sofa: 0.82 [-0.12, 1.78, 639.87, 473.52]
remote: 0.79 [333.65, 76.38, 370.69, 187.48]
Training
RT-DETRv2 is trained on COCO (Lin et al. [2014]) train2017 and validated on COCO val2017 dataset. We report the standard AP metrics (averaged over uniformly sampled IoU thresholds ranging from 0.50 − 0.95 with a step size of 0.05), and APval50 commonly used in real scenarios.
Applications
RT-DETRv2 is ideal for real-time object detection in diverse applications such as autonomous driving, surveillance systems, robotics, and retail analytics. Its enhanced flexibility and deployment-friendly design make it suitable for both edge devices and large-scale systems + ensures high accuracy and speed in dynamic, real-world environments.
Configuration
- Architecture
- RtDetrV2ForObjectDetection
- Stored precision
- float32
- Model type
- rt_detr_v2
Identity and Version
- Repository
- PekingU/rtdetr_v2_r101vd
- Publisher
- Peking University
- Task
- Object detection
- Modality
- Image
- Library
- transformers
- Parameters
- 77M parameters
- Languages
- en
- Revision
- 2c5dbbd2d4d8c8814827a3b42737ba1afce3cf2a
- First published
- 2025-01-31
- Last updated
- 2025-02-06
Files and Weights
5 files, 307.3 MB in total. The weights are 1 file totalling 307.3 MB in safetensors.
Weights1 file · 307.3 MB
Configuration2 files · 5.8 KB
Documentation1 file · 3.5 KB
Repository1 file · 1.5 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 307.3 MB | fa2d79aa5006 |
| config.json | Configuration | 5.3 KB | — |
| preprocessor_config.json | Configuration | 444 B | — |
| README.md | Documentation | 3.5 KB | — |
| .gitattributes | Repository | 1.5 KB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 307.3 MB
Download from Peking University
Released by Peking University through its official repository on Hugging Face. Read the license.
Built From
- Described by arXiv:2407.17140
- Trained on (disclosed) coco
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 307.3 MB |
| 16-bit | 0.2 GB |
| 8-bit | 0.1 GB |
| 4-bit | 0.0 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About rtdetr_v2_r101vd
How much GPU memory does rtdetr_v2_r101vd need?
About 0.2 GB at 16-bit and 0 GB at 4-bit: the weights (77M parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run rtdetr_v2_r101vd on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use rtdetr_v2_r101vd commercially?
Yes. rtdetr_v2_r101vd is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
Similar Models
Roof detection model for remote sensing imagery, fine-tuned using RT-DETR. The following example shows roof detections produced by the model: Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model.
This model is a fine-tuned version of YOLOv8x specialized in detecting two specific classes: Face and Person. It has been trained on a large-scale proprietary dataset consisting of approximately 150,000 images. The high capacity of the YOLOv8x architecture combined with a diverse proprietary dataset ensures high accuracy and robustness in various scenarios. You can load the model using the Hugging Face transformers library by enabling custom code execution. If you prefer the standard Ultralytics API, you can download the weights from the Hub and load them directly. This method automatically handles model downloading for ultralytics YOLO model. This model is based on the Ultralytics YOLOv8…
The D-FINE model was proposed in D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement by Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, Feng Wu This model was contributed by VladOS95-cyber with the help of @qubvel-hf This is the HF transformers implementation for D-FINE coco -> model trained on COCO obj365 -> model trained on Object365 obj2coco -> model trained on Object365 and then finetuned on COCO D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global…
The D-FINE model was proposed in D-FINE: Redefine Regression Task in DETRs as Fine-grained Distribution Refinement by Yansong Peng, Hebei Li, Peixi Wu, Yueyi Zhang, Xiaoyan Sun, Feng Wu This model was contributed by VladOS95-cyber with the help of @qubvel-hf This is the HF transformers implementation for D-FINE coco -> model trained on COCO obj365 -> model trained on Object365 obj2coco -> model trained on Object365 and then finetuned on COCO D-FINE, a powerful real-time object detector that achieves outstanding localization precision by redefining the bounding box regression task in DETR models. D-FINE comprises two key components: Fine-grained Distribution Refinement (FDR) and Global…
DEtection TRansformer (DETR) model trained end-to-end on COCO 2017 object detection (118k annotated images). It was introduced in the paper End-to-End Object Detection with Transformers by Carion et al. and first released in this repository. Disclaimer: The team releasing DETR did not write a model card for this model so this model card has been written by the Hugging Face team. The DETR model is an encoder-decoder transformer with a convolutional backbone. Two heads are added on top of the decoder outputs in order to perform object detection: a linear layer for the class labels and a MLP (multi-layer perceptron) for the bounding boxes. The model uses so-called object queries to detect…