Downloads · 30 days
85
54% of all-time downloads
bambezius/yolo11n-soccernet-ball
yolo11n-soccernet-ball is a object detection model from bambezius. Use it when you need objects located in an image. It is set up for ultralytics.
Private copy of a checkpoint fine-tuned for ball detection using the https://github.com/anthonyceponis/glitch project and SoccerNet-v3 annotations.
Downloads · 30 days
85
54% of all-time downloads
All-time downloads
158
Public
Repo size
449 MB
Likes
0
Public
Click a slice to open those files.
.gz362 MB · 97%
From the Hugging Face model README
Private copy of a checkpoint fine-tuned for ball detection using the https://github.com/anthonyceponis/glitch project and SoccerNet-v3 annotations.
best.pt is the selected run's best checkpoint by validation mAP50-95.
It is not selected by pixel error. checkpoint-info.json records its SHA-256.
args.yaml and results.csv, if available, provide settings and full run history;
the final CSV row is not necessarily the best checkpoint's epoch.
This is a model transfer, not a grant of additional rights to the upstream model or training dataset. Refer to Ultralytics and SoccerNet's applicable terms. No dataset images, credentials, or other checkpoints are uploaded by this script.
Download with hf download bambezius/yolo11n-soccernet-ball best.pt --local-dir models/yolo11-soccernet.
The original model and class metadata are retained in the checkpoint.
For inference, load with ultralytics.YOLO("models/yolo11-soccernet/best.pt").
For exact training recovery, use a full numbered epoch checkpoint; best.pt may
have its optimizer state stripped after a normally completed training run.