Spaces:
Configuration error
Configuration error
| ## 👋 About DMind.AI | |
| DMind is an open-source AGI research organization focused on next-generation digital finance, like cryptocurrency, DeFi and RWA. | |
| | [Website](https://dmind.ai) | [X](https://x.com/dmind_ai) | [Discord](https://discord.com/invite/R7xk2cgD) | | |
| ## 💡 Core Focus Areas | |
| 1. **LLM Web3 Capability Enhancement**: We focus on training and fine-tuning large language models to better understand blockchain concepts, smart contracts, and DeFi protocols, delivering more accurate Web3 knowledge services. | |
| 2. **AI-Driven Web3 Analytics**: Developing intelligent analysis tools to help users understand on-chain data, predict market trends, identify potential risks, and provide AI-powered support for investment decisions. | |
| 3. **Model Evaluation & Benchmark**: Establishing AI model evaluation standards for the Web3 domain, conducting comprehensive assessments of market-available LLMs, and publishing authoritative reports to guide industry development. | |
| 4. **Smart Contract Optimization**: Utilizing AI technology to assist in the design, audit, and optimization of smart contracts, improving their security, efficiency, and usability. | |
| ## 🔍 Latest Research & Projects | |
| - [**DMind Benchmark**](https://huggingface.co/datasets/DMindAI/DMind_Benchmark): a domain-specific evaluation suite designed to assess the capabilities of large language models in the Web3 context. | |
| - [**DMind-1**](https://huggingface.co/DMindAI/DMind-1): the first LLM dedicated to Web3. It outperforms general-purpose models of similar size in Web3 tasks significantly. | |
| - [**DMind-2**](https://huggingface.co/DMindAI/DMind-2-107B): the second-generation series of finance-native LLMs (107B & 4B), built for crypto investment research, risk analysis, and real execution through advanced tool-calling. | |
| ## Official Ollama Distribution Authorization | |
| DMindAI authorizes [EonHao](https://github.com/EonHao) to represent | |
| DMindAI in submitting and coordinating first-party requests for the | |
| official, unnamespaced distribution of DMindAI models in the Ollama | |
| Library. | |
| This authorization covers: | |
| - communication with the Ollama team; | |
| - official model-family naming and tag proposals; | |
| - verification of canonical source repositories and revisions; | |
| - model conversion, quantization, packaging, and compatibility | |
| coordination; | |
| - validation of chat templates, tool calling, reasoning, and | |
| multimodal behavior; | |
| - coordination of future DMindAI model releases and updates. | |
| Subject to the licenses of each model and its base model, DMindAI | |
| authorizes Ollama to convert, quantize, host, and distribute DMindAI | |
| model artifacts as official Ollama Library entries. | |
| This authorization does not transfer ownership of the models and does | |
| not grant permission to relicense them. | |
| Official DMindAI channels: | |
| - Website: https://dmind.ai/ | |
| - GitHub: https://github.com/DMindAI | |
| - Hugging Face: https://huggingface.co/DMindAI | |
| Authorized representative: | |
| - GitHub: https://github.com/EonHao | |
| Organization contact: | |
| - Email: team@dmind.ai | |
| Effective date: July 23, 2026 | |
| This authorization remains valid until revoked or replaced by a later | |
| public statement from DMindAI. |