AIM Weekly for 01-July-2024
Tim Spann @PaaSDev Milvus - Towhee - Attu - Feder - GPTCache - VectorDB Bench
https://www.youtube.com/@FLaNK-Stack
https://medium.com/@tspann/subscribe
https://ossinsight.io/analyze/tspannhw
Please join my meetup group NJ/NYC/Philly/Virtual.
This is Issue #144
Join me:
July 25, 2024 5:30 to 8:30 PM in NYC @ Cloudera 101 5th Ave · New York, NY Cloudera office - 8th Floor https://www.meetup.com/unstructured-data-meetup-new-york/events/301720478/?utm_source=partner&utm_medium=referral&utm_campaign=2024_newsletter_tspann-ai-newsletters_external
Zilliz Cloud https://docs.zilliz.com/docs/release-notes-290
Unity Catalog https://github.com/unitycatalog/unitycatalog/
Necklace AI? https://basedhardware.com/
July 25 - Meetup @ Cloudera NYC August 13 - Meetup @ Hudson Yards NYC
Hardware coming... Transformer enhancement... Sohu
I wonder if this will supercharge Vector Databases?
Milvus Lite is for only one vector per collection. As of current version in 2.4.
There's a lot of cool stuff with Milvus and new models, techniques, libraries and use cases.
Edge AI with Milvus Lite https://medium.com/@tspann/edgeai-edge-vector-database-6a9b5238bffb
Quantization!?!?!? https://medium.com/@tspann/how-good-is-quantization-in-milvus-6d224b5160b0
Vector Embeddings https://zilliz.com/learn/everything-you-should-know-about-vector-embeddings?utm_source=tim
Milvus Lite with LangChain and LLaMaIndex https://medium.com/@zilliz_learn/how-to-connect-to-milvus-lite-using-langchain-and-llamaindex-69ed139c7e4b
Choosing the Right Embedding Model for Your Data https://zilliz.com/blog/choosing-the-right-embedding-model-for-your-data
How Delivery Hero Implemented Safety System for AI https://zilliz.com/blog/how-delivery-hero-implemented-safety-system-for-ai-generated-images?utm_source https://www.slideshare.net/slideshow/i-see-eyes-in-my-soup-how-delivery-hero-implemented-the-safety-system-for-ai-generated-images/267924072
Local Agentic RAG with Langraph and LLAMA3 https://zilliz.com/blog/local-agentic-rag-with-langraph-and-llama3?utm_source=partner&utm_medium=referral&utm_campaign=2024_newsletter_tspann-ai-newsletters_external
Mastering LLM techniques https://developer.nvidia.com/blog/mastering-llm-techniques-inference-optimization/#in-flight_batching
Milvus Performance Benchmark for Vector Databases https://zilliz.com/resources/whitepaper/milvus-performance-benchmark
Vector Search and RAG Balancing Accuracy https://zilliz.com/blog/vector-search-and-rag-balancing-accuracy-and-context?utm_source=li
Promethean Wager AI Vector Databases https://severalnines.com/podcast/promethean-wager-ai-vector-databases-and-data-sovereignty
AI https://news.ycombinator.com/item?id=40789353
Attention Explained https://ai-explained.yoko.dev/1-attention-explained
Polyfill Chain Attack https://sansec.io/research/polyfill-supply-chain-attack
What we learned from Pinterests Text to SQL Solution https://blog.getwren.ai/what-we-learned-from-pinterests-text-to-sql-solution-840fa5840635
The Ultimate Guide to Run Any LLM Locally https://programming.earthonline.us/an-ultimate-guide-to-run-any-llm-locally-eb1a43052053
Drop of a Hat Model https://universe.roboflow.com/test-y7opj/drop-of-a-a-hat/model/2 https://dropofahat.zone/
Structured Output From LLMs https://www.boundaryml.com/blog/structured-output-from-llms
The Death of NYC Congestion Pricing https://www.apricitas.io/p/the-death-of-nyc-congestion-pricing
Finding GPT4S Mistakes with GPT-4 https://openai.com/index/finding-gpt4s-mistakes-with-gpt-4/
Finetuning https://mlops.systems/posts/2024-06-25-evaluation-finetuning-manual-dataset.html
Enterprise RAG at Scale https://medium.com/@dialoglk/asimov-leveraging-rag-models-for-enhanced-efficiency-in-the-telecommunications-engineering-domain-f220fc405571
Try Milvus 2.4 on Zilliz https://www.linkedin.com/pulse/try-milvus-24-features-zilliz-cloud-learn-vector-embeddings-check-b9fyc/
All the Free AI Education https://zilliz.com/learn
Multimodal Embeddings with Fifty One and Milvus https://zilliz.com/blog/exploring-multimodal-embeddings-with-fiftyone-and-milvus
Elevating User Experience with Image Based Fashion Recommendations https://zilliz.com/blog/elevating-user-experience-with-image-based-fashion-recommendations?utm_source=linkedin&utm_medium=social%20&utm_campaign=2024-06-26_social_linkedin-newsletter_zilliz
Training State of the Art General Text Embedding https://www.slideshare.net/slideshow/training-stateoftheart-general-text-embedding/267310506
Fine Tune Florence 2 https://huggingface.co/blog/finetune-florence2
AI Data Infrastructure https://www.felicis.com/insight/ai-data-infrastructure
RAG with Small Language Models https://medium.com/data-science-at-microsoft/evaluating-rag-capabilities-of-small-language-models-e7531b3a5061
Synthetic Data Generation https://blogs.nvidia.com/blog/nemotron-4-synthetic-data-generation-llm-training/
Deep Dive into RAG https://towardsdatascience.com/17-advanced-rag-techniques-to-turn-your-rag-app-prototype-into-a-production-ready-solution-5a048e36cdc8
Live Fun Friday with Unstructed Data Preview https://www.youtube.com/watch?v=_jQB62uPsvc
Running the NVIDIA Milvus Lite Demo https://www.youtube.com/watch?v=7kdYbaw2LSQ
RAG in Production https://www.youtube.com/watch?v=_MpqlnN-TtE
Unstructured Meetup https://www.youtube.com/watch?v=ntiA36Skdrw
Princeton AI Meetup 18-June-2024 Report
AI Camp NYC - 20-June-2024 - Tim Speaks -With Slides https://www.youtube.com/watch?v=2YQiJzwA6BE
AI Camp NYC - 20-June-2024 - Tim Speaks - Raw video feed https://www.youtube.com/watch?v=wYEtg4UuvPM
Unstructured Data Processing with RPI 5 AI Kit https://www.youtube.com/watch?v=tZFJ1DDkD1Q
Using JSON Fields with Milvus https://www.youtube.com/watch?v=HP5L3Hr6Mt8
DSS ML Talk https://www.youtube.com/watch?v=t17Ga4l5gvo
Webinar https://zilliz.com/event/asimov-enterprise-rag-at-dialog-axiata?
https://www.slideshare.net/slideshow/06-18-2024-princeton-meetup-introduction-to-milvus/269765983
Oct 27 - 29, Raleigh, NC - All Things Open https://2024.allthingsopen.org/speakers/timothy-spann
Nov 5-7, 10-12, 2024: CloudX. Online/Santa Clara. https://www.developerweek.com/cloudx/
Nov 19, 2024: XtremePython. Online. https://xtremepython.dev/2024/
Building an Agentic RAG locally with Milvus, Ollama and LangGraph July 11, 2024 | 9:00 AM PT/12:00PM ET | Stephen Batifol, Zilliz Get hands-on and learn how to:
- Enable agent planning, memory, and tool use for tasks
- Allow LLM web searches and custom function calls
- Implement fallbacks and self-correction for agent errors https://zilliz.com/event/rag-agents-with-langchain-and-milvus?utm_campaign=tim
RAG Evaluation with Ragas July 18 | 9:00 AM PT/12:00PM ET | Christy Bergman, Zilliz
- Evaluate a RAG pipeline using metrics like context F1-score and answer correctness, then learn the differences between:
- Foundation model evaluation vs RAG evaluation
- Human evaluation vs LLM-as-a-judge evaluations
- Overall RAG vs RAG component evaluations https://zilliz.com/event/rag-evaluation-with-ragas?utm_campaign=tim
Hands-On Demo: Building and Scaling Vector Search Apps with Zilliz Cloud July 25, 2023 | 9:00 AM PT/12:00PM ET | Frank Liu, Zilliz Learn how to build and scale vector search applications with live examples. Walk through the following:
- Live Zilliz Cloud setup and configuration
- Building a simple chatbot step-by-step
- Advanced search techniques with examples https://zilliz.com/event/hands-on-zilliz-cloud-demo?utm_campaign=tim
- https://github.com/tspannhw/AIM-MilvusLite
- https://github.com/tspannhw/AIM-NYCStreetCams
- https://github.com/tspannhw/AIM-MotorVehicleCollisions
- https://github.com/milvus-io/milvus?utm_source=partner&utm_medium=referral&utm_campaign=2024_newsletter_tspann-ai-newsletters_external
- https://huggingface.co/mistralai/Codestral-22B-v0.1
- https://huggingface.co/IDEA-Research/grounding-dino-tiny
- https://huggingface.co/datasets/nvidia/HelpSteer2
- https://ftfy.readthedocs.io/en/latest/
- https://github.com/lmstudio-ai
- https://github.com/eclipse-theia/theia/releases
- https://github.com/wavetermdev/waveterm
- https://github.com/constacts/milvus-clj
- https://www.tessell.com/services/tessell-for-milvus
- https://github.com/devflowinc/trieve
- https://github.com/constacts/ragtacts/tree/main
- https://github.com/knuddelsgmbh/jtokkit
- https://github.com/spring-projects/spring-ai
- https://github.com/exadel-inc/CompreFace
- https://github.com/mayneyao/eidos
- https://www.fuzzmap.io/
- https://amphi.ai/
- https://github.com/google-deepmind/magiclens
- https://git-cliff.org/docs/
- https://github.com/CerebriumAI/examples/tree/master/18-realtime-voice-agent
- https://github.com/y-scope/clp?uclick_id=a585c5dc-9268-410e-8eb0-31f1ac8679b0
- https://github.com/AutoMQ/automq
- https://github.com/fiddlecube/fiddlecube-sdk
- https://www.jetson-ai-lab.com/agent_studio.html#__tabbed_1_4
- https://github.com/stephen37/Milvus_demo/tree/main/multimodal_milvus_clip
- https://github.com/mifi/lossless-cut
- https://www.labgopher.com/
- https://huggingface.co/blog/finetune-florence2
- https://github.com/andimarafioti/florence2-finetuning
- https://hatch.pypa.io/latest/
- https://pickcode.io/
© 2020-2024 Tim Spann https://www.youtube.com/@FLaNK-Stack
