container-status-de / README.md
mindchain's picture
Add dataset card
f4849e1 verified
metadata
language:
  - de
  - en
license: mit
tags:
  - logistics
  - shipping
  - container
  - classification
  - port
  - terminal
size_categories:
  - n<100

Container Status Classification Dataset

Dataset for training container status classification models in port logistics.

Labels (15 categories)

Label Description Urgency
DAMAGED Container damaged, inspection required high
DELAYED Container delayed, schedule change medium
CUSTOMS_HOLD Customs inspection or hold high
READY_PICKUP Available for collection low
GATE_IN Container entered terminal low
GATE_OUT Container left terminal low
LOADED Loaded on vessel low
DISCHARGED Unloaded from vessel medium
MISSING_SEAL Seal missing or broken high
WEIGHT_MISMATCH Weight discrepancy high
TEMP_ALERT Temperature alert for reefers critical
DOCUMENTS_MISSING Missing paperwork high
BLOCKED Container blocking access medium
RELEASED Customs cleared, free to go low
INSPECTION_REQUIRED Survey/inspection needed high

Data Structure

{
  "text": "Container MSCU1234567 beschädigt",
  "label": "DAMAGED",
  "urgency": "high",
  "action_required": "inspect_and_claim",
  "container_id": "MSCU1234567"
}

Usage

from datasets import load_dataset

dataset = load_dataset("mindchain/container-status-de")

Intended Models

  • google/t5gemma-2-270m (edge deployment)
  • google/t5gemma-2-1b (serverless)
  • Fine-tuned for text-to-text classification