Verba Releases

The first stable release of Resume Matcher, featuring tailored resume creation, multiple ATS-compatible templates, and support for multiple AI providers.

Jan 6, 2026

This release adds a new rotation detection parameter to assist with document orientation handling.

Aug 29, 2025

This update focuses on fixing color space issues in PDF images, reducing retry logic for failed requests, and adding metadata support to parsed documents.

Aug 22, 2025

Introduces the early preview of confidence scores for field extraction, moving the parameter name from 'experimental' to 'confidence'.

Jul 21, 2025

This release adds support for additional data types including CSV and Excel files, along with new Ollama environment variables for model configuration.

Jul 14, 2025

A major version bump that addresses critical bugs regarding Windows file encoding and bounding box cropping, while introducing a ParseConfig class for better configuration management.

Jul 9, 2025

This is an archival release of the 'Old Resume Matcher' containing extensive UI enhancements, Streamlit upgrades, and dependency updates.

May 26, 2025

Open-Sora V1.3.0 focuses on architectural improvements, specifically enhancing the ST-DiT model and upgrading video compression to deliver higher quality video generation.

Feb 21, 2025

This release introduces Novita AI generator support and fixes spaCy language issues while also improving OpenAI model name retrieval.

Jan 27, 2025

This update introduces a new 'High Fidelity' model, significantly expands language support to nine languages, and overhauls the user interface with new tools like a Lens Viewer and clipboard integration.

Dec 25, 2024

This release introduces new AI providers like Upstage and Groq, along with enhanced deployment options and Docker support. It features significant improvements in document processing capabilities, including audio support via AssemblyAI and better Markdown chunking. Additionally, the update resolves various bugs, including collection name conflicts and race conditions, while upgrading dependencies.

Dec 10, 2024

A major overhaul introducing asynchronous ingestion, a new DocumentExplorer with vector visualization, and migration to Weaviate v4 Client.

Sep 6, 2024

This release adds .docx support, GitLabReader integration, and improves embedding models including MixedBread and AllMPNet embedders.

Jul 2, 2024

Version 1.2.0 introduces a video compression network, rectified-flow training, enhanced multi-stage training with more data, and better model conditioning and evaluation metrics.

Jun 18, 2024

This minor update focuses on fixing various bugs, including issues with notification icons on Linux, non-ASCII character recognition, Windows custom width upscaling, and logs reporting 100.00%.

Jun 1, 2024

This release improves the ST-DiT architecture with rope positional encoding and QK normalization, supports training with any resolution and duration, and adds support for image/video conditioning and video editing.

May 7, 2024

A significant update that revamps the backend, allowing scaling up to 16X, introducing lossy/lossless compression for image formats, adding scale options to the main interface, and removing post-processing steps.

Apr 25, 2024

The initial release of Open-Sora, a project dedicated to efficiently producing high-quality video, featuring source code for data processing and training pipelines, optimized by ColossalAI and flash attention, and model weights for specific resolutions.

Mar 17, 2024

Version 5.6 introduces full compatibility for Demucs versions 1 through 4 and MDX23C models, accompanied by new MDX23C models via the Download Center and expanded GPU acceleration for macOS.

Sep 26, 2023

This release focuses on adding brand new MDX-Net models and ensuring full compatibility with Demucs versions 1 through 4, alongside continued stability improvements and model downloads via the Settings menu.

Dec 19, 2022