sentence-transformers How to use hasinthakapiyumal/bge-reasoner-embed-ai-patterns with sentence-transformers:
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("hasinthakapiyumal/bge-reasoner-embed-ai-patterns")
sentences = [
"The code implements several data curation and filtering patterns crucial for autonomous driving AI datasets. It includes filtering scenarios based on the fractional presence of specific LiDAR point cloud tokens, selecting scenarios exhibiting particular ego vehicle dynamics (e.g., non-stationary, starting, stopping), and ensuring contextual relevance by checking for route-intersecting roadblocks. These patterns are essential for preparing targeted and high-quality datasets for training and evaluating perception, prediction, and planning models.",
"The code implements the Intelligent Driver Model (IDM) as a reactive car-following policy, dictating longitudinal acceleration based on ego and lead agent states, desired velocity, and safety parameters. This policy is central to an agent-based simulation framework, where `IDMAgent`s dynamically plan and propagate their trajectories by interacting with an `OccupancyMap` that represents other vehicles and traffic light statuses.",
"The code demonstrates a pattern for building highly customizable AI agents (`CustomizeAgent`) that integrate external tools (e.g., `MCPToolkit`, `ArxivToolkit`) to augment their capabilities. A core AI pattern is the robust handling of structured output through various parsing modes (`json`, `xml`, `title`, `custom`), ensuring precise information extraction. Furthermore, it showcases the orchestration of these agents and tools into complex workflows (`WorkFlowGraph`) for multi-step task execution, representing a multi-agent system pattern.",
"This code implements a Retrieval Augmented Generation (RAG) pipeline, primarily focusing on the ingestion and embedding of documents. It employs a `DocumentProcessor` for pre-processing, including chunking strategies (`split_by`, `split_length`) and blocklist filtering, before using an `RAGEmbedder` to generate vector embeddings via an external model (e.g., Ollama). These embeddings are then indexed into a vector store like Elasticsearch, with integrated metrics tracking for pipeline observability."
]
embeddings = model.encode(sentences)
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [4, 4] 
