#Open source

TOPLOC: A Local Sensitivity Hashing Scheme for Trusted Verification Inference

TOPLOC detects unauthorized modifications to models, prompts, or computational precision through local sensitivity hashing mechanism, verification speed is 100 times faster than original inference, and submission storage for every 32 new tokens only requires 258 bytes (compared to 262 KB for Llama-3.1-8B-Instruct). This solution maintains robustness across various hardware configurations, GPU types, tensor parallel dimensions, and attention kernel implementations, reducing memory overhead by 1000 times and is suitable for large-scale deployment.

The PRIME-RL algorithm achieves 10 times faster training for the INTELLECT-MATH model through synthetic reasoning data.

Dify has released the Plugins plugin system to modularize component extensions to AI application capabilities.

Dify has launched the Plugins plugin system into Beta, extending AI application functionality through plug-and-play components. Developers can obtain ready-made plugins or build their own on the Dify Marketplace, supporting capabilities such as image processing, audio analysis, text generation, data analysis, logical reasoning, and external integration. The system adopts a decoupled architecture, including five types of components: models, tools, strategies, extensions, and packages, achieving bidirectional communication with Dify's core functionality through custom endpoints.

MetaGene-1: A 7B parameter metagenomic base model for epidemic monitoring.

MetaGene-1 is a 7B parameter metagenomic base model pre-trained on 1.5T DNA/RNA sequencing data for epidemic monitoring and pathogen detection. It achieves SOTA on pathogen detection and metagenomic embedding benchmarks, with a 512-token context length suitable for short sequence anomaly detection, but limited for complex task applications. The model was developed in collaboration by USC, Prime Intellect, and the Nucleic Acid Observatory, and is open-sourced.

Dify v0.15.0 Released: Introducing Parent-Child Retrieval to Enhance Knowledge Acquisition

Dify v0.15.0 Launches Parent-Child Retrieval Technology, Achieving Precise Retrieval by Matching Fine-Grained Child Fragments, Combined with Coarse-Grained Parent Fragments to Supplement Context, Improving RAG System's Response Accuracy and Information Richness. This Feature Supports Two Parent Fragment Splitting Modes (Paragraph/Full Text), Allows Users to Edit Parent-Child Fragments Independently and Choose Whether to Regenerate Child Fragments, While Optimizing Fragment Preview Interface to Enhance Low-Code Platform Usability.

Dify v0.14.0 Released: Enhancing AI Workflow Elasticity Through Error Handling

Dify v0.14.0 introduces powerful error handling features, allowing developers to define custom error handling for LLM nodes, HTTP nodes, tool nodes, and code nodes. It prevents process interruptions through strategies like default values and workflow redirection, enhancing the reliability of AI applications.

INTELLECT-1 Released: The First Global Jointly Trained 10B Parameter Model

INTELLECT-1 is released, the first 10B parameter language model trained through global collaboration, based on the Llama-3 architecture, completing 1T token training on 112 H100 GPUs across 5 countries and 3 continents, with the PRIME framework achieving a 400x reduction in communication bandwidth, 96% computational utilization, supporting sequence lengths of 8192 and a vocabulary of 128256, with the model code and data open-sourced.

Dify v0.10.0 released: Supporting file upload functionality.

Dify v0.10.0 released, adding file upload functionality, supporting document, audio, and video files, enabling applications such as Q&A and summary generation, and allowing the creation of AI-driven podcasts through file upload and LLM nodes.

INTELLECT–1 globally distributed training the first 10B parameter model

INTELLECT-1 launches the world's first 10B parameter model distributed training, inviting the public to contribute computing power to advance the open-source AGI process. The training is based on the OpenDiLoCo open-source implementation, extending the 1B parameter scale of DeepMind DiLoCo's method by 10 times, significantly reducing communication frequency (up to 500 times), and lowering bandwidth requirements by 2000 times through quantized pseudo-gradients (int8) and 500-step outer optimizer synchronization.

Dify partnered with Takin.ai to provide a more convenient GenAI education platform

Dify announced a collaboration with GenAI education startup Takin.ai, simplifying GenAI education through its integrated platform. It supports teachers and students in zero-setting learning various GenAI concepts and tools, and integrates a customized version of Dify into the Takin.ai system as the core feature for building AI agents and agent workflows.

OpenDiLoCo: An open-source global distributed low-communication training framework

OpenDiLoCo successfully replicated and extended DeepMind's DiLoCo experiment to a 10B parameter model scale, achieving cross-continent, three-country distributed training through Hivemind and PyTorch FSDP, maintaining 90-95% computational resource utilization. The framework is based on the open-source implementation of Hivemind, supporting dynamic resource joining/leaving and fault-tolerant training, significantly reducing communication requirements by 500x while maintaining performance on 1.5B parameter models.

Introduction to DifySandbox

DifySandbox is a code execution sandbox service provided by the Dify platform, used for processing JSON, XML data, and executing user-defined code in the background, while ensuring security by isolating the file system, database, and network. It supports two languages, Python and Node.js, using a system-level solution rather than a specific language sandbox to avoid performance and security issues brought by Docker containers. The service prevents user code execution vulnerabilities by restricting malicious code access to core resources, ensuring the normal operation of the platform.

Dify Integrates LangSmith and Langfuse to Enhance LLM Application Observability

Dify version 0.6.12 integrates LangSmith and Langfuse, allowing users to access detailed data of LLM applications through simple configuration, facilitating cost, latency, and quality evaluation. LangSmith provides rich evaluation features, and Langfuse supports open-source hosted architecture, both helping Dify users monitor and optimize LLM application performance.

Prime Intellect Compute platform officially launched: Global GPU resource aggregation and scheduling market

Prime Intellect Compute platform officially launched, aggregating global GPU resources to democratize instant computing. The platform offers a unified resource pool, supporting instant access from single GPU to 512+ GPU clusters. The first batch integrates 12 cloud service providers, with H100 GPUs priced at $1.5-4 per hour.

Dify v0.6.11 released: Integrating Firecrawl to create a top-tier web data knowledge base solution.

Dify v0.6.11 released, integrating Firecrawl as a web data source, generating structured markdown data from websites to enhance knowledge bases, supporting synchronization of web data and embedding vector databases to build RAG applications, providing 500 free page API credit quota, while optimizing workflow collaboration features.

Dify releases new architecture Dify Beehive, enhancing flexibility and scalability.

Dify refactors its architecture into a modular design, launching the Model Runtime service module to support models with 175B parameters, achieving a unified interface and simplified YAML configuration, dynamically adding models, improving development efficiency and compatibility.

How does Dify.AI's Annotation Reply feature improve chatbot quality and reduce costs?

Dify.AI's Annotation Reply feature allows manual editing of historical conversations or importing batches of Q&A to prioritize matching similar user questions with annotated replies, enabling chatbots to provide precise answers in specific scenarios. This feature helps build an LLMOps data feedback loop, reducing LLM call costs and latency in conversational scenarios, and accumulating valuable data assets for future model fine-tuning.

Dify.AI integrates the Jina AI embedding model to enhance the performance of RAG applications.

Dify.AI integrates the Jina Embeddings model into its platform, which supports a context length of 8,192 tokens, surpassing most open-source models that only support 512 tokens.

Dify.AI v0.3.31 released: RAG technology improved by 20% over OpenAI Assistants API.

Dify.AI v0.3.31 update enhances RAG technology through hybrid search, semantic reordering models, and multi-path retrieval, improving question-answering efficiency by 20%, significantly outperforming the OpenAI Assistants API. Ragas evaluation shows an increase of 18.44% in Ragas score, 20% improvement in context accuracy, and 35.71% increase in faithfulness.

Dify.AI v0.3.29 Released: Supports GPT-4 Turbo and Vision Models

Dify.AI v0.3.29 version adds support for OpenAI GPT-4 Turbo (128K context) and GPT-4 Vision models, and introduces external data API tools and sensitive content review features.

Dify v0.3.13 released: Easily integrate open-source large models like Llama2 and ChatGLM

Dify v0.3.13 released, integrating open-source large models like Llama2 and ChatGLM, supporting models from Hugging Face and Replicate platforms, allowing users to easily access and quickly build AI applications via API keys. The platform supports rapid switching between models, enabling users to compare performance by clicking the model selector, lowering the barrier for model selection and tuning.

Dify.AI: A visual platform for creating and operating native AI applications with LLMOps