AI Model Training

Gdpval OpenAI

OpenAI is an open source AI model economic value assessment framework that measures the economic value of AI through 1,320 real tasks in 44 occupations.

OmniManip OmniManip

Object center constraint-driven universal robot operation method

Aether Aether

An open source world model research framework that unifies 4D reconstruction, video prediction, and visual planning.

Gemma 3 Google DeepMind

Gemma 3 is a model training platform for AI developers and researchers, supporting model fine-tuning, training management and deployment.

Gradio Gradio

Use a few lines of Python to build an interactive web interface for machine learning models and share it as an open source library

AGI-Eval AGI-Eval

A Chinese evaluation community for large models, providing evaluation sets, lists and ability comparisons

AnythingLLM AnythingLLM

All-in-one RAG desktop and server applications that can be deployed privately, integrating documents, models and agents into the same workspace

HELM HELM

The large model overall evaluation system launched by Stanford University emphasizes transparent evaluation of multiple scenarios and indicators.

C-Eval C-Eval

A multi-level and multi-disciplinary assessment package for Chinese large models, containing 13948 questions, covering 52 subjects

Jan Jan

Open source, local large model desktop client that can be run offline, compatible with cloud models and comes with OpenAI style API

Cohere AIStartMap

Cohere is an enterprise-level AI platform with "data sovereignty" as its core, providing the North full-stack workbench, Compass intelligent search and the full range of Command/Embed/Rerank models.

Kimi Linear Moonshot AI

Dark Side of the Moon Open Source Hybrid Linear Attention Long Context Model