Yiqi Lu

NVIDIA Research developed VILA, a visual language model with a holistic pretraining, instruction tuning, and deployment pipeline that helps our NVIDIA clients succeed in their multi-modal products. VILA achieves SOTA performance both on image QA benchmarks and video QA benchmarks, having strong multi-image reasoning capabilities and in-context learning capabilities. It is also optimized for speed. VILA offers an efficient design recipe to augment LLMs toward vision tasks, from training to inference catering. Leveraging the full strength of unfreezing the LLM, interleaved image-text data curation, and careful text data re-blending, VILA has surpassed state-of-the-art methods for vision tasks while preserving text-only capabilities. VILA has demonstrated strong reasoning capability for multi-image analysis, in-context learning, and zero/few-shot tasks. We hope VILA can help NVIDIA build better multi-modal foundation models with diverse applications in NVIDIA Metropolis, audiovisual, robotics, generative AI, and more. For more information, see the VILA: On Pre-training for Visual Language Models paper and the /Efficient-Large-Model/VILA GitHub repo.

3

1 Comment

AIQCon is coming up fast and we have a new speaker lined up for you! Cindy Xing, Software Engineer Manager at Google, is presenting "Open Model and Its Curation over Kubernetes". Open Generative AI (GenAI) models are transforming the AI landscape. But which one is right for your project? What are the quality metrics for one to evaluate his/her own trained model? For application developers and AI practitioners enhancing their applications with GenAI, it’s critical to choose and evaluate the model that meets both quality and performance requirements. Cindy will examine customer scenarios and discuss the model selection process. She will explore the current landscape of open models and collection mechanisms to measure model quality and share insights from Google’s experience. There are a limited amount of tickets left! Use code "BeOurGuest" to secure your seat for 20% off! https://lnkd.in/gxTzZSXR

5

Google Cloud and Anthropic made some exciting announcements: 📣 Claude 3 Opus, Anthropic’s most intelligent model, is now generally available on Vertex AI ⚙️ Tool use is available for the Claude 3 model family on Vertex AI, enabling Claude to act as an autonomous agent 📈Provisioned throughput is available for the Claude 3 model family on Vertex AI

1

1 Comment

How is AI transforming product management? Join Chris Butler, Staff Product Operations Manager at GitHub to explore this question with other senior leaders on May 22nd at 12pm PT ✨ RSVP at https://lu.ma/bnh9s3y1 to save your seat! ✨ enrich conversations are an opportunity to ask questions and meet other AMAZING people in similar roles. The small, intimate format makes it easy to raise your hand and ask about your particular situation, unlike a 100 person webinar. #peerlearning #leaders #productmanagement #AI #leadership

4

Using generative AI, NVIDIA operations built an AI Planner agent, developed on NVIDIA Inference Microservices (NIM). The agent leverages LLM, NeMo Retriever and CuOpt NIM to reduce re-planning time from hours to just seconds. https://nvda.ws/4bIfS2T

2

Implement a Q&A experience using a RAG technique with Elasticsearch as a vector database by following the steps in this blog: https://gag.gl/V2hwuA #ElasticSearchLabs #RAG #GenAI

5

Check out Confluent’s second issue of Chronicles! Inside, you’ll find a clever, but relatable approach to learn about the fundamentals of Apache Flink, stream processing, and Confluent. Our heroes, developer and architect team, Ada and Jax are struggling with decentralized data processing challenges and need a manageable, cost-effective stream processing solution for an important upcoming launch. They embark on their next adventure to learn why Apache Kafka and Apache Flink are better together. Get your copy now!

3

2 Comments

Llama 3 has arrived and the newest version of Meta AI is dropping, The world leading AI is seamless to use and a game changer for our family of apps. #NextGen #AI #ML #LlamaDrama #Llama3 #Meta #MetaAI

13

Using generative AI, NVIDIA operations built an AI Planner agent, developed on NVIDIA Inference Microservices (NIM). The agent leverages LLM, NeMo Retriever and CuOpt NIM to reduce re-planning time from hours to just seconds. https://nvda.ws/4bIfS2T

7

Speed up semantic search performance with token pruning for our ELSER model: https://lnkd.in/gePkdDpn #ElasticSearchLabs #SemanticSearch

19

Just announced at #MongoDBlocal NYC: Atlas Stream Processing is GA and ready to support your production workloads! With this developers can: - Effortlessly handle complex and rapidly changing data structures - Use the familiar MongoDB Query API for processing streaming data - Seamlessly integrate with MongoDB Atlas - Benefit from a fully managed service that eliminates operational overhead Dive deeper into the announcement today. https://lnkd.in/g79qXPHy #artificialintelligence #ai #machinelearning #technology #datascience #python

4

New toy in town: Qwen2 is the new model series of large language models from the Qwen team

2

🔓 Unlock the power of your data with NVIDIA NeMo Curator. Discover how to #customize and optimize your model by integrating high-quality, domain-specific datasets. https://nvda.ws/3WO1QYZ

6

While speaking at the GenAI Summit (GenAIx- Helping you to 10x your company with AI) last week, Rodrigo Liang, CEO and co-founder of SambaNova Systems, spoke with ABC7 News Bay Area, discussing how San Francisco has become the center of the AI boom and what the future of AI looks like.    Rodrigo says the Bay Area's AI innovation is incredible, and it is helping redefine and reshape San Francisco. Check out the full story below.    #AI #SanFrancisco

5

1 Comment

🏠🤖 We're pushing the boundaries of AI with our new OpenEQA benchmark! How well can AI understand and navigate physical spaces? Let's explore the challenges and what's next for vision+language models. Check out the details here. #AI #Meta 🌟

6

Now we have two first class data catalogs integrated with MS Fabric now available in Azure. Of course DataBricks is the heavy heavy favorite in the enterprise lakehouses as well as AI (as of the conference that just finished 🤓)

6

See how Expedia Group implements Alluxio to federate cross-region data lakes in AWS. Alluxio unifies geo-distributed data silos without replication, enabling consistent and high performance with ~50% reduced costs. Learn More: https://buff.ly/3Qtx97l #egresscost #s3 #aws #hybridcloud #multicloud #datasilos #data #distributedcomputing #caching #storage #compute

1

Check out this in-depth recap of Riyu Wang’s presentation at FlinkForward 2023, where the Senior Development Engineer from Alibaba Cloud Intelligence discussed Using Apache Paimon + StarRocks for High-speed Batch and Streaming Lakehouse Analysis. The article explores the development history, main scenarios, technical principles, performance tests, and future planning of Apache Paimon + StarRocks. Dive into the full article to explore these aspects in detail: https://lnkd.in/gW4Jb2MF #DataAnalytics #DataEngineering 

9