This repository aims to be a comprehensive, community-driven catalog of state-of-the-art (SoTA) results across various machine learning problem domains, including NLP, Computer Vision, Speech, and Reinforcement Learning. It serves researchers and practitioners seeking to track the latest advancements and benchmark their own work against leading methodologies.

How It Works

The project aggregates SoTA results by collecting information on research papers, datasets, evaluation metrics, source code availability, and publication year. It categorizes findings across supervised, semi-supervised, unsupervised, transfer, and reinforcement learning paradigms, providing a structured overview of progress in specific ML tasks like language modeling, machine translation, image classification, and speech recognition.

Quick Start & Requirements

This repository is a curated list and does not have a direct installation or execution command. Users can browse the README for information.

Highlighted Details

• Comprehensive coverage across NLP, Computer Vision, Speech, and Reinforcement Learning.
• Detailed breakdown of results by task, including specific datasets, metrics, and source code links (where available).
• Actively seeks community contributions to maintain up-to-date information.

Maintenance & Community

The project is maintained by a single individual seeking collaborators, particularly in NLP, Computer Vision, and Reinforcement Learning. Community contributions are encouraged via GitHub issues or a Google Form. The last update mentioned was February 20th, 2019.

Licensing & Compatibility

The repository itself does not contain code to license. The licensing of the individual research papers and their associated code would need to be checked separately.

Limitations & Caveats

The repository's information is only as current as its last update (February 2019), meaning it may not reflect the absolute latest SoTA results. Source code availability is inconsistent, with several entries marked as "NOT FOUND" or "NOT YET AVAILABLE."