2 dépôts
Awesome GitHub RepositoriesDifficulty-Based Datasets
Datasets categorized by complexity levels to study the development of multi-step deductive reasoning.
Distinct from Dataset Curation: Specializes Dataset Curation by focusing on the organization of data by difficulty rather than the general process of cleaning and labeling.
Explore 2 awesome GitHub repositories matching artificial intelligence & ml · Difficulty-Based Datasets. Refine with filters or upvote what's useful.
Awesome Difficulty-Based Datasets GitHub Repositories
Trouvez les meilleurs dépôts grâce à l'IA.Nous recherchons les dépôts les plus pertinents grâce à l'IA.
- hkust-nlp/simplerl-reason
hkust-nlp/simpleRL-reason
3,867Voir sur GitHubsimpleRL-reason is a training framework designed to improve mathematical and logical deduction in large language models. It utilizes reinforcement learning and policy optimization to enhance the accuracy and transparency of step-by-step deduction chains. The project implements a pipeline that establishes baseline capabilities through supervised fine-tuning before applying reinforcement learning to maximize deductive accuracy. It features a reward modeling toolkit that calculates scalar feedback by comparing generated reasoning steps against verified mathematical ground truths. The framework
Provides curated training and evaluation datasets organized by complexity to facilitate multi-step deductive problem solving.
Python - google-deepmind/mathematics_dataset
google-deepmind/mathematics_dataset
1,954Voir sur GitHubThis project provides a structured repository of school-level mathematical problems designed to train and evaluate the reasoning capabilities of neural network models. It functions as a standardized benchmark for measuring the proficiency of artificial intelligence systems in arithmetic, algebra, and logical reasoning. The dataset is generated through procedural synthesis, utilizing formal grammars and template-driven logic to create unique question and answer pairs. To support incremental learning, the content is organized into hierarchical difficulty levels, allowing for the structured sequ
Organizes mathematical problem sets into hierarchical difficulty levels to support structured training and performance benchmarking.
Python
