News Topic Classifier

Runs inside your own AWS account: your text is processed on a SageMaker endpoint you control, inside your VPC, and is never sent to a third-party API.

This SageMaker model package provides a REST API to classify news text into four topics: World, Sports, Business, or Sci/Tech, with a confidence score.

Use it to organize content feeds, tag articles at ingestion time, or build topic-aware recommendation and monitoring pipelines.

The API accepts input as JSON or CSV and supports real-time endpoints and batch transform.

We welcome your feedback at [email protected]

Model and training data The model is a ModernBERT-base (answerdotai/ModernBERT-base) text classifier fine-tuned with PyTorch and Hugging Face Transformers on the public ag_news dataset. It runs in network isolation on SageMaker, so no data leaves your account.

Known limitations - English only. Other languages are not supported. - The model was fine-tuned on inputs up to 512 tokens. Longer text is accepted, but accuracy is only characterized to that length; split long documents into passages before sending. - The label set is fixed to the four AG News topics: World, Sports, Business, and Sci/Tech. Text outside those topics is still forced into one of the four. - Trained on news articles; accuracy on other genres will be lower.

Measured performance - Accuracy 94.4 percent, macro-F1 0.944, measured on 20,000 held-out examples from AG News. The split is seeded (70/20/10, seed 42) and was not used in training. - Real-time endpoint on ml.m5.large: 268 ms median per request (282 ms p95). - Throughput on the same instance: about 23 texts per second at a batch size of 32.