Neo4j Live: Enhancing text2cypher with In-Context Learning & Fine-Tuning

This podcast episode delves into the innovative intersection of text-to-Cypher technology and artificial intelligence, featuring hosts Alex and Wilsen. They emphasize the importance of fundamental coding knowledge while exploring Wilsen's journey from data science to the realm of graph databases and natural language processing (NLP). The conversation highlights the transformative power of in-context learning, especially through few-shot learning methods, in enhancing the text-to-Cypher experience by providing accurate query translations. Wilsen demonstrates practical implementations using Langchain and Neo4j, touching on fine-tuning strategies like LoRA to bolster query handling capabilities. Ultimately, the episode equips listeners with actionable insights into leveraging LLMs for graph queries while underscoring the necessity of underlying comprehension in coding.

Outlines

Sign in to continue reading, translating and more.

Continue

Neo4j

Introduction to Text-to-Cypher and the Role of AI in Code Generation

Wilsen's Journey into Graph Databases and NLP

Enhancing Text-to-Cypher with In-Context Learning: Zero-Shot, One-Shot, and Few-Shot Learning

Hands-on Demonstration: Few-Shot Learning with Langchain and Neo4j

Fine-tuning Text-to-Cypher Models with LoRA: Techniques and Implementation

Fine-tuning Implementation and Evaluation: Using Unslot and the Hugging Face Hub

Neo4j Live: Enhancing text2cypher with In-Context Learning & Fine-Tuning

Neo4j

00:00Introduction to Text-to-Cypher and the Role of AI in Code Generation

Introduction to Text-to-Cypher and the Role of AI in Code Generation

03:16Wilsen's Journey into Graph Databases and NLP

Wilsen's Journey into Graph Databases and NLP

08:16Enhancing Text-to-Cypher with In-Context Learning: Zero-Shot, One-Shot, and Few-Shot Learning

Enhancing Text-to-Cypher with In-Context Learning: Zero-Shot, One-Shot, and Few-Shot Learning

18:54Hands-on Demonstration: Few-Shot Learning with Langchain and Neo4j

Hands-on Demonstration: Few-Shot Learning with Langchain and Neo4j

44:48Fine-tuning Text-to-Cypher Models with LoRA: Techniques and Implementation

Fine-tuning Text-to-Cypher Models with LoRA: Techniques and Implementation

1:00:04Fine-tuning Implementation and Evaluation: Using Unslot and the Hugging Face Hub

Fine-tuning Implementation and Evaluation: Using Unslot and the Hugging Face Hub