--- title: "The greatest progress that the human race has made lies in learning how to make correct inferences." newsletter: "MLOps Community" date: 2024-05-23 source: https://aaif.live/newsletters/mlopscommunity/2024-05-23-the-greatest-progress-that-the-human-race-has-made-lies-in-l --- # The greatest progress that the human race has made lies in learning how to make correct inferences. *Plus, getting down with the data doyen, MLOps and e-shops, benchmarks up to speed, and hidden gems.* *MLOps Community — Agentic AI Foundation, 2024-05-23* ## Beavis and Butthead https://www.myinstants.com/en/instant/beavis-butthead-laugh/?utm_source=copy&utm_medium=share Tim Allen https://www.myinstants.com/en/instant/home-improvement-huh-67362/ Tom Jones https://www.myinstants.com/en/instant/tom-huh-11800/ ## Retrieval Augmented Generation // Syed Asad // MLOps Podcast #233 It's kind of ironic that we talked a lot about the inference layer in this episode when you won’t need inference skills to understand what Asad means, as he's very honest with his answers. We opened the chat with him sharing a production issue he was dealing with at that moment. That’s the level of openness he goes for. He discussed the challenges of using tools for handling unstructured data and streaming data, the necessity of premium connectors, insights on integrating tools in production environments, and addressing complexities with vector embeddings and CSV files. A nice little bonus was learning about his company’s innovative ML research team that tests and compares frameworks and tools, documenting their findings in a GitHub repository for client-specific solutions. So, use your inference skills to read between the lines. _____________________________________________________ Click below to watch the episode! Video [https://home.mlops.community/home/videos/retrieval-augmented-generation] || Spotify [https://open.spotify.com/episode/6Q7qRUgMqQgzT86oJuU60o?si=_4hFRjIYQ5G4JV5JVthPpA] || Apple [https://podcasts.apple.com/us/podcast/retrieval-augmented-generation/id1505372978?i=1000655962869] [https://podcasts.apple.com/us/podcast/retrieval-augmented-generation/id1505372978?i=1000655962869](https://podcasts.apple.com/us/podcast/retrieval-augmented-generation/id1505372978?i=1000655962869) [https://home.mlops.community/home/videos/retrieval-augmented-generation](https://home.mlops.community/home/videos/retrieval-augmented-generation) [https://open.spotify.com/episode/6Q7qRUgMqQgzT86oJuU60o?si=_4hFRjIYQ5G4JV5JVthPpA](https://open.spotify.com/episode/6Q7qRUgMqQgzT86oJuU60o?si=_4hFRjIYQ5G4JV5JVthPpA) ## Open Standards Make MLOps Easier and Silos Harder // Cody Peterson // MLOps Podcast #234 This is one for the data devotees! There was some chat about the challenges and potential of LLMs and the whole AI hype, but it wasn’t the main thing. We talked a lot about the Ibis project, which compiles data frame code to multiple backends, looking to bridge the gap between SQL and Python data frames for efficient data processing. He told me about plans for developments to Ibis including enhancing stability, developing an ML package for data preprocessing, and considering community requests for new backends and complex features. We also talked about how crucial open standards are for making data systems work smoothly together, the challenges in setting industry standards, and how Voltron Data is focusing on open-source contributions. Data dynamos, definitely dive-in! Video [https://home.mlops.community/home/videos/open-standards-make-mlops-easier-and-silos-harder] || Spotify [https://open.spotify.com/episode/7a9fUqZkZzMQ8coQxwuPS6?si=k2oFY-dDQoay1VxxBhVWEg]|| Apple [https://podcasts.apple.com/us/podcast/open-standards-make-mlops-easier-and-silos-harder/id1505372978?i=1000656298533] [https://podcasts.apple.com/us/podcast/open-standards-make-mlops-easier-and-silos-harder/id1505372978?i=1000656298533](https://podcasts.apple.com/us/podcast/open-standards-make-mlops-easier-and-silos-harder/id1505372978?i=1000656298533) [https://home.mlops.community/home/videos/open-standards-make-mlops-easier-and-silos-harder](https://home.mlops.community/home/videos/open-standards-make-mlops-easier-and-silos-harder) [https://open.spotify.com/episode/7a9fUqZkZzMQ8coQxwuPS6?si=k2oFY-dDQoay1VxxBhVWEg](https://open.spotify.com/episode/7a9fUqZkZzMQ8coQxwuPS6?si=k2oFY-dDQoay1VxxBhVWEg) ## Job of the Week [https://docs.google.com/document/d/1mTjsriocJfhGwagj44eT5ZBPiPcf7yINeB0xxIiPb8o/edit#heading=h.fexct7ijaqv3](https://docs.google.com/document/d/1mTjsriocJfhGwagj44eT5ZBPiPcf7yINeB0xxIiPb8o/edit#heading=h.fexct7ijaqv3) ## Lessons Learned from Doing MLOps within E-commerce // VMarcus Svensson // IRL #78 Stockholm ## MLOps Community IRL Meetup Looking for a new outfit to dazzle everyone this conference season? Marcus has got you covered. In his IRL talk, he discusses how machine learning technologies like recommender systems, product ranking, and search functionality enhance consumer experiences on e-commerce sites. It’s no cat-walk in the park, though. He explores the technical challenges of maintaining system availability, particularly during high-traffic events such as sales or promotions. He also explains the significance of A/B testing and iterative approaches in optimizing algorithms to improve metrics like conversion rates and average order values. Watch the video, then wait for the invite to next year’s MET Gala. Watch it here [https://home.mlops.community/home/videos/lessons-learned-from-doing-mlops-within-e-commerce] [Watch it here](https://home.mlops.community/home/videos/lessons-learned-from-doing-mlops-within-e-commerce) ## Exploring LLMs Speed Benchmarks I feel the need... the need for detailed benchmarks! Not quite as catchy as the original, but much more useful. This blog gives an independent analysis of the speed performance of three advanced 7 billion parameter language models: Mistral 7Bn, Llama-2 7Bn, and Gemma 7Bn. It breaks down the testing environment, configuration settings, types of prompts, and token ranges used in the evaluations. Each model and library combination had it's own unique strengths, meaning there's no one-size-fits-all solution, but it does mean there's great set of benchmarks to help you pick the most suitable model for your needs. This blog can be your wingman anytime! With thanks to Rajdeep Borgohain and Aishwarya Goel for their contribution. [Exploring LLMs Speed Benchmarks](https://home.mlops.community/home/blogs/exploring-llms-speed-benchmarks-2024-05-17) ## Hidden Gems ## IRL Meetups Amsterdam [https://www.meetup.com/amsterdam-mlops-community/events/300696760/] - May 28 Munich [https://www.meetup.com/munich-mlops-community/events/300673506/] - May 28 California [https://www.meetup.com/mlops-community/events/301129835/?utm_medium=referral&utm_campaign=share-btn_savedevents_share_modal&utm_source=link] - May 29 Stockholm [https://www.meetup.com/stockholm-mlops-community/events/300881538/] - May 30 Frankfurt [https://www.meetup.com/frankfurt-mlops-community/events/300877018/?utm_medium=referral&utm_campaign=share-btn_savedevents_share_modal&utm_source=link] - June 12 Denver [https://www.meetup.com/denver-mlops-community/events/300432149/?utm_medium=referral&utm_campaign=share-btn_savedevents_share_modal&utm_source=link] - June 12 San Francisco [https://www.meetup.com/mlops-community/events/300375509/?utm_medium=referral&utm_campaign=share-btn_savedevents_share_modal&utm_source=link] - June 25 [Amsterdam](https://www.meetup.com/amsterdam-mlops-community/events/300696760/) Thanks for reading. See you in Slack [https://go.mlops.community/slack], YouTube [https://www.youtube.com/channel/UCG6qpjVnBTTT8wLGBygANOQ?view_as=subscriber], and podcast [https://home.mlops.community/public/content/] land. Oh yeah, and we are also on X [https://twitter.com/mlopscommunity]. The MLOps Community newsletter is edited by Jessica Rudd [https://www.linkedin.com/in/jmrudd/]. --- Source: https://aaif.live/newsletters/mlopscommunity/2024-05-23-the-greatest-progress-that-the-human-race-has-made-lies-in-l