VisionServeNorthwind Labsprod
Overview
  • Dashboard
  • Playground
Inference
  • Jobs
  • Batches
  • Results
Models
  • Registry
  • Deployments
  • Benchmarks
Observability
  • Metrics
  • Usage & Quotas
  • System Health
Workspace
  • API Keys
  • Webhooks
  • Storage
  • Team
  • Audit Log
  • API Reference
  • Settings
MC

Mara Chen

administrator

  1. VisionServe
  2. Models

Model registry

4 registered models · 2 healthy

detector-v1

v1.4.2 · ONNX Runtime CPU

healthy

YOLOv8-based general object detector fine-tuned on street and warehouse imagery. Primary production detector.

detectprod50.0 MB
Requests
1.3M
Latency
41.8 ms
Errors
0.31%

Updated 2 months ago

ocr-v1

v1.2.0 · ONNX Runtime CPU

healthy

Two-stage document text extraction pipeline: DBNet region detection followed by CRNN recognition. Supports 12 languages.

ocrprod23.0 MB
Requests
643K
Latency
118.4 ms
Errors
0.58%

Updated 3 months ago

classifier-v1

v0.9.3 · ONNX Runtime CPU

inactive

EfficientNet-B3 general image classifier, 1000-class ImageNet label space plus 84 fine-tuned domain classes.

classifystaging45.8 MB
Requests
88.4K
Latency
22.6 ms
Errors
2.14%

Updated 3 months ago

segmenter-v0

v0.2.1 · ONNX Runtime CPU

warming

Instance segmentation head (YOLOv8-seg). Beta — exposed only in the development environment.

segmentdev53.1 MB
Requests
2.1K
Latency
96.3 ms
Errors
4.60%

Updated 2 months ago