Introducing Prompt Compression and Query Optimization, a short course created in collaboration with MongoDB and taught by Richmond Alake, Developer Advocate at MongoDB. 📈 Learn to integrate traditional database features with vector search capabilities to enhance the performance and cost-efficiency of large-scale RAG applications. In the course, you’ll explore these techniques: - Prefiltering and Postfiltering: Filter results based on specific conditions. - Projection: Minimize output size by selecting a subset of fields from a query. - Reranking: Reorder search results to prioritize the most desired outcomes. - Prompt Compression: Reduce the length of prompts for cost-effective large-scale applications. With practical exercises, you’ll also implement vector search for RAG using MongoDB, develop multi-stage aggregation pipelines, and use metadata to refine search results. Learn more and sign up for free: https://hubs.la/Q02G14DM0
DeepLearning.AI
Software Development
Palo Alto, California 1,023,572 followers
Making world-class AI education accessible to everyone
About us
DeepLearning.AI is making a world-class AI education accessible to people around the globe. DeepLearning.AI was founded by Andrew Ng, a global leader in AI.
- Website
-
http://DeepLearning.AI
External link for DeepLearning.AI
- Industry
- Software Development
- Company size
- 11-50 employees
- Headquarters
- Palo Alto, California
- Type
- Privately Held
- Founded
- 2017
- Specialties
- Artificial Intelligence, Deep Learning, and Machine Learning
Products
DeepLearning.AI
Online Course Platforms
Learn the skills to start or advance your AI career | World-class education | Hands-on training | Collaborative community of peers and mentors.
Locations
-
Primary
2445 Faber Pl
Palo Alto, California 94303, US
Employees at DeepLearning.AI
Updates
-
Initialization can have a significant impact on convergence in training deep neural networks. Simple initialization schemes can accelerate training, but they require some care to avoid common pitfalls. Read our guide to learn how to initialize neural network parameters effectively: https://hubs.la/Q02Gn9Fk0
-
Anthropic introduced Artifacts, which allows users to work on generated outputs as independent files on Claude. Find more details in #TheBatch: https://hubs.la/Q02GkffD0
-
Yesterday we launched Prompt Compression and Query Optimization, our latest short course, taught by MongoDB's Richmond Alake! 🔍 Combine vector search capabilities with traditional database operations to build efficient, cost-effective RAG applications. 🛠️ Learn key techniques like pre-filtering, post-filtering, and projection for faster query processing and optimized output. 💡 Reduce prompt lengths with prompt compression for large-scale applications. Enroll now and start optimizing your RAG applications: https://hubs.la/Q02GjnWS0
-
DeepLearning.AI reposted this
AI/GenAI Team Leader and IIT Chair at NielsenIQ | PhD in AI at UAH | Top 10 technologists under 35 years old in Spain (Nova111 winner) | 16x professional and academic awards | 12x patents | 33x publications
🔬 I love continuous learning. #RAG is one of the hot topics in #GenAI, so I decided to have a look into the new course by Andrew Ng (DeepLearning.AI) and Richmond Alake (MongoDB) about Prompt Compression and Query Optimization. 🚀 This course teaches you to combine traditional database capabilities with vector search using MongoDB for RAG. You'll learn these techniques: - Vector search: For semantic matching of user queries. - Filtering using metadata: Pre- and post-filtering to narrow search results. - Projections: Selecting only necessary fields to minimize data returned. - Boosting: Reranking results to improve relevance. - Prompt compression: Using a small LLM to compress context, significantly reducing token count and processing costs. 💡 If you want to enroll and complete this short course, I strongly recommend it, great contents. Follow this link to do it: https://lnkd.in/dJ_pYk3Y
-
-
This week in #TheBatch: 🎨 All about Claude’s Artifacts feature 🌱 AI growth hinders carbon emissions goals 🧠 GaLore, a memory-efficient method for fine-tuning and pretraining Plus: Andrew Ng shares his concerns about the proposed California regulation SB 1047. Read The Batch now: https://hubs.la/Q02G5YpM0
AI's Cloudy Path to Zero Emissions, Amazon's Agent Builders, and more
deeplearning.ai
-
DeepLearning.AI reposted this
As RAG and agent-based LLM applications hit production, keeping operational costs down is key. Our free course on Prompt Compression and Query Optimization, in partnership with DeepLearning.AI, will teach you how to combine traditional and vector database techniques to make RAG more cost-effective and efficient. In the course, you will learn to: ✂️ Reduce prompt length to save inference costs. 🗂️ Filter results based on conditions, applied at index creation or after vector search. 🔝 Reorder search results to improve relevance. 📋 Select a subset of fields to minimize inputs to LLM. Enroll today and start optimizing: https://lnkd.in/gD4_v5bh Andrew Ng, Richmond Alake
Prompt Compression and Query Optimization
deeplearning.ai
-
DeepLearning.AI reposted this
Learn to optimize RAG for cost and performance in our new short course, Prompt Compression and Query Optimization, created with MongoDB and taught by Richmond Alake. This course teaches you to combine traditional database capabilities with vector search using MongoDB for RAG. You'll learn these techniques: - Vector search: For semantic matching of user queries - Filtering using metadata: Pre- and post-filtering to narrow search results - Projections: Selecting only necessary fields to minimize data returned - Boosting: Reranking results to improve relevance - Prompt compression: Using a small LLM to compress context, significantly reducing token count and processing costs These methods address scaling, performance, and security challenges in large-scale RAG applications. You can sign up here: https://lnkd.in/gMVN3hzM
-
Researchers at Sakana AI developed an automated method for merging models. The resulting new models show improved performance without further training. Sakana’s team used the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) to merge the Japanese-language LLM Shisa-Gamma with two math-specific, English-language LLMs: Abel and WizardMath. This process produced a high-performing model that excelled in solving math problems written in Japanese, achieving 55.2 percent accuracy on the Multilingual Grade School Math dataset. Read our summary of the paper in #TheBatch: https://hubs.la/Q02FWzKF0
-
-
Are you wondering how to build a career in AI? Then we’ve got the perfect eBook for you! It covers necessary steps, including: Developing technical skills, finding projects, nailing the big interview, and more. Get it here for free: https://hubs.la/Q02FQJjC0
-