RTMLib Models: Computer Vision Model Weights from OpenMMLab

A collection of RTMLib model weights, likely for computer vision tasks such as pose estimation, stored locally by the author. The dataset was uploaded on 2026-06-08 following the shutdown of OpenMMLab servers. The author, DavidPagnon, notes that some models are missing and requests community help to locate them.

Use Cases

• Load pre-trained RTMPose models for human pose estimation tasks.
• Benchmark or reproduce computer vision experiments based on the OpenMMLab ecosystem.
• Serve as a backup source for model weights following the shutdown of the original hosting servers.

Strengths

• Contains model weights from the established OpenMMLab computer vision framework.
• Provides a community-sourced backup for models that became unavailable from the original source.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Row count is unknown, which may limit suitability assessment.
• The collection is incomplete, with specific models like 'RTMPose HALPE26' listed as missing.

Provenance

Source

OpenMMLab servers, with local backup by DavidPagnon.

Collection Method

Locally stored model checkpoints uploaded to Hugging Face.

Freshness

Last updated 2026-06-08 10:05:40; freshness should be verified.