LibreGTR
Collection
GTR (Gated Token Recurrence) for LibreYOLO: detection, instance segmentation, pose, oriented boxes, depth and semantic segmentation. MIT. • 22 items • Updated
LibreGTRl-seg
GTR-L COCO instance-segmentation weights converted for LibreYOLO.
GTR support is being prepared for LibreYOLO v1.6.0. Earlier PyPI releases may not include this model family.
Usage
With a LibreYOLO version that includes GTR segmentation:
from libreyolo import LibreYOLO
model = LibreYOLO("LibreGTRl-seg.pt")
results = model.predict("image.jpg")
masks = results.masks
The default input is 640 by 640 pixels. This checkpoint supports instance segmentation (boxes, classes and per-instance masks).
Source
Official GTR implementation,
source revision 782e737efe2e6437ac537fbdcee089673d3376c1.
Published checkpoint,
weight repository revision 9fc62c8c2b2c976835d0f1c1ffc544dbc0f9e29f.
Copyright (c) 2026 Intellindust-AI-Lab. The source code is MIT licensed and the
publisher's weight repository explicitly declares MIT.
Modifications
Selected the EMA state dict and added LibreYOLO schema v1.0 metadata.
Learned parameters and state-dict keys are unchanged. Training/optimizer state
was removed. Conversion uses weights/convert_gtr_weights.py in the
LibreYOLO source repository.
Validation
Checkpoint schema, exact tensor preservation and strict loading were checked for this artifact. On CPU at 640 pixels the LibreYOLO graph reproduces the pinned upstream graph exactly (logits, boxes and mask logits), using the same portable attention substitute on both sides. CUDA parity, independent COCO accuracy and training validation are deferred to the LibreYOLO v1.6.0 release checks. No independently reproduced accuracy or latency numbers are claimed here.