sapiens2-seg-pose-tensorrt
Results and visual examples
Sapiens2 combined segmentation/pose TensorRT artifacts. Segmentation masks and pose outputs have distinct contracts; use the per-build README and runtime metadata. No exact-build photo comparison was recovered.
Available build records: rtx5090-sm120-trt10.14.1.48-dual-input-max-batch-1-fp32rescue/README.md, rtx5090-sm120-trt10.14.1.48-dual-input-max-batch-1-fp32rescue/sapiens2_0.4b_seg_pose_dual_fp16.metadata.json, rtx5090-sm120-trt10.14.1.48-dual-input-max-batch-1-fp8-default-normqk-exp/sapiens2_0.4b_seg_pose_dual.README.md, rtx5090-sm120-trt10.14.1.48-dual-input-max-batch-1-fp8-default-normqk-exp/sapiens2_0.4b_seg_pose_dual_fp8.metadata.json, rtx5090-sm120-trt10.14.1.48-dual-input-max-batch-1-fp8-default-normqk-exp/sapiens2_0.8b_seg_pose_dual.README.md, rtx5090-sm120-trt10.14.1.48-dual-input-max-batch-1-fp8-default-normqk-exp/sapiens2_0.8b_seg_pose_dual_fp8.metadata.json.
No checkpoint-matched photo gallery was located in the existing evidence reviewed for this update. The files and runtime records below are the available evidence; no visual quality claim is inferred from their presence.
Evidence provenance
This documentation update reuses saved results; it does not rerun inference or change weights. The starting repository revision is eb8213a5355d. Captions distinguish model predictions, training diagnostics, and aggregate metrics. Qualitative examples are not a representative accuracy estimate.