Werpy is a robust yet lightweight Python package, engineered to swiftly compute and analyzes the Word Error Rate (WER) between two text sets. It boasts versatility in managing various types of input data, including strings, lists, and numpy arrays.
The library encompasses an extensive array of functionalities, encompassing text normalization to accommodate discrepancies in data collection. It also empowers users to effortlessly allocate distinct weights or penalties to specific error categories such as insertions, deletions, and substitutions.
Moreover, the summary function furnishes a comprehensive breakdown of the computed outcomes, facilitating a swift and detailed analysis of individual errors.
Features
- Text normalization capabilities
- Calculate a single word error rate
- Calculate multiple word error rates
- Calculates weighted word error rates
- Easily show all the insertion, deletion and substitution errors
Other Useful Business Software
Auth for GenAI | Auth0
Easily implement secure login experiences for AI Agents - from interactive chatbots to background workers with Auth0. Auth for GenAI is now available in Developer Preview
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of werpy!