Hacker News: montanalow

Plea deal with accused 9/11 plotters revoked

montanalow — Sat, 03 Aug 2024 14:00:57 +0000

Article URL: https://www.bbc.com/news/articles/c7202593k0xo

Comments URL: https://news.ycombinator.com/item?id=41146657

Points: 2

# Comments: 0

New comment by montanalow in "What's Hacker News' problem with open source AI?"

montanalow — Thu, 25 Jul 2024 01:24:14 +0000

Thanks for the feedback. I've updated that paragraph to clarify the product placement is for Llama. I don't really think Meta needs our help to much...

New comment by montanalow in "Korvus: Single-Query RAG with Postgres"

montanalow — Fri, 12 Jul 2024 14:01:59 +0000

If you want to see all the SQL functions and tables Korvus depends on, check out the pgml extension.

https://postgresml.org/docs/open-source/pgml/

New comment by montanalow in "Korvus: Single-Query RAG with Postgres"

montanalow — Thu, 11 Jul 2024 23:01:29 +0000

Yep, one of our other projects, pgcat is exactly to help make the horizontal scaling as easy as possible.

https://github.com/postgresml/pgcat

New comment by montanalow in "We migrated from AWS to GCP with minimal downtime"

montanalow — Fri, 07 Jun 2024 17:28:54 +0000

Most managers and organizations miss the fact that engineers are often motivated by solving puzzles in ways like this, because it's fun. If you want to accomplish big challenges, quickly, make it fun for the engineer. Working long hours on projects like this doesn't burn engineers out, it sharpens their skills, broadens their knowledge and grows the organizations capabilities long term.

This sort of cultural difference of exploration and letting work be fun is one of the big things that accounts for the differences in velocity between big co's and little co's. Does your work give you energy to the point where not only you love doing it, you want to tell everyone else about it?

Serverless LLMs are dead; Long live Serverless LLMs

montanalow — Thu, 30 May 2024 16:57:45 +0000

Article URL: https://postgresml.org/blog/serverless-llms-are-dead-long-live-serverless-llms

Comments URL: https://news.ycombinator.com/item?id=40525947

Points: 1

# Comments: 0

New comment by montanalow in "Show HN: Hacker Search – A semantic search engine for Hacker News"

montanalow — Thu, 02 May 2024 17:30:28 +0000

You can do all of that in a single SQL query, with pgml.embed() and then pgml.train() a custom reranker with xgboost, to pgml.predict() the conversion score of a search result based on click-through-rate, or other objective.

If you'd like free hosting, feel free to reach out. I'm one of the founders at postgresml.org.

LLMs are commoditized; data is the differentiator

montanalow — Fri, 26 Apr 2024 16:01:43 +0000

Article URL: https://postgresml.org/blog/llms-are-commoditized-data-is-the-differentiator

Comments URL: https://news.ycombinator.com/item?id=40171032

Points: 1

# Comments: 0

New comment by montanalow in "Swap OpenAI with any open-source model"

montanalow — Wed, 06 Dec 2023 23:15:04 +0000

This is an SDK built to interact with PostgresML, which provides ML & AI _inside_ a Postgres database. Clients in this case don't perform inference, rather the server does. You could run the open source server locally, or connect to one running in the cloud.

New comment by montanalow in "My favorite database shirts"

montanalow — Mon, 27 Nov 2023 17:49:36 +0000

If anyone would like a free PostgresML T-shirt, we just did our first run. Feel free to email me with your shipping info and size. It'd also be nice to get to know you a bit if your email address isn't obvious.

Show HN: PostgresML – Run open-source ML/LLM models in a Postgres extension

montanalow — Thu, 16 Nov 2023 19:30:54 +0000

We've been hard at work improving PostgresML, and thought it was time for an update now that our cloud offering is generally available.

In case you missed our first posts, I built the open-source ML platform at Instacart way back in 2017. I learned a ton, but primarily that it's better to bring your ML workload to the database rather than bringing the data to the codebase. That's why we made PostgresML.

Fundamentally, it enables Postgres to act as a GPU-powered AI application database — where you can both save models and index data. That eliminates the need for the myriad of separate services you have to tie together for your ML workflow. Pgml + pgvector create a complete ML platform (vector DB, model store, inference service, open-source LLMs) all within open-source extensions for PostgreSQL. That takes a lot of the complexity out of your infra, and it's ultimately faster for your users.

We wanted to give folks a way to tinker with it easily, so we've added a few interactive demos to our site.

At the top of the homepage, you can test out a few models right on our site. If you scroll down, you can actually try out our Python, JavaScript, Rust and SQL SDKs locally and for free. Just copy the example code and you can play around with open-source models in your favorite database.

Let us know what you think.

Comments URL: https://news.ycombinator.com/item?id=38294209

Points: 2

# Comments: 0

How-to improve search results with machine learning

montanalow — Mon, 04 Sep 2023 16:18:12 +0000

Article URL: https://postgresml.org/blog/how-to-improve-search-results-with-machine-learning

Comments URL: https://news.ycombinator.com/item?id=37381910

Points: 1

# Comments: 0

Pgml-chat: A CLI for deploying low-latency knowledge-based chatbots

montanalow — Wed, 16 Aug 2023 22:31:20 +0000

Article URL: https://postgresml.org/blog/pgml-chat-a-command-line-tool-for-deploying-low-latency-knowledge-based-chatbots-part-I

Comments URL: https://news.ycombinator.com/item?id=37154757

Points: 5

# Comments: 0

LLM based pipelines with PostgresML and DBT

montanalow — Mon, 17 Jul 2023 16:01:58 +0000

Article URL: https://postgresml.org/blog/llm-based-pipelines-with-postgresml-and-dbt

Comments URL: https://news.ycombinator.com/item?id=36759858

Points: 4

# Comments: 0

New comment by montanalow in "PostgresML Adds GPTQ and GGML Quantized LLM Support for HuggingFace Transformers"

montanalow — Tue, 20 Jun 2023 16:14:08 +0000

Quantization allows PostgresML to fit larger models in less RAM. These algorithms perform inference significantly faster on NVIDIA, Apple and Intel hardware. Half-precision floating point and quantized optimizations are now available for your favorite LLMs downloaded from Huggingface.

PostgresML Adds GPTQ and GGML Quantized LLM Support for HuggingFace Transformers

montanalow — Tue, 20 Jun 2023 16:14:07 +0000

Article URL: https://postgresml.org/blog/announcing-gptq-and-ggml-quantized-llm-support-for-huggingface-transformers

Comments URL: https://news.ycombinator.com/item?id=36406261

Points: 4

# Comments: 1

MindsDB vs. PostgresML

montanalow — Thu, 08 Jun 2023 16:08:40 +0000

Article URL: https://postgresml.org/blog/mindsdb-vs-postgresml

Comments URL: https://news.ycombinator.com/item?id=36244224

Points: 4

# Comments: 0

New comment by montanalow in "Python SDK for PostgresML with scalable LLM embedding memory and text generation"

montanalow — Fri, 02 Jun 2023 16:21:49 +0000

We've been working on a Python SDK[1] for PostgresML to make it easier for application developers to get the performance and scalability benefits of integrated memory for LLMs, by combining embedding generation, vector recall and LLM tasks from HuggingFace in a single database query.

This work builds on our previous efforts that give a 10x performance improvement from generating the LLM embedding[2] from input text along with tuning vector recall[3] in a single process to avoid excessive network transit.

We'd love your feedback on our roadmap[4] for this extension, if you have other use cases for an ML application database. So far, we've implemented our best practices for scalable vector storage to provide an example reference implementation for interacting with an ML application database based on Postgres.

[1]: https://github.com/postgresml/postgresml/tree/master/pgml-sd... [2]: https://postgresml.org/blog/generating-llm-embeddings-with-o... [3]: https://postgresml.org/blog/tuning-vector-recall-while-gener... [4]: https://github.com/postgresml/postgresml/tree/master/pgml-sd...

Python SDK for PostgresML with scalable LLM embedding memory and text generation

montanalow — Fri, 02 Jun 2023 16:21:49 +0000

Article URL: https://postgresml.org/blog/introducing-postgresml-python-sdk-build-end-to-end-vector-search-applications-without-openai-and-pinecone

Comments URL: https://news.ycombinator.com/item?id=36167144

Points: 3

# Comments: 1

New comment by montanalow in "PostgresML raises $4.7M to launch serverless AI app databases based on Postgres"

montanalow — Thu, 25 May 2023 16:42:53 +0000

Full disclaimer, I work on PostgresML, and I'm not a MindsDB expert.

They both do ML in the database, but we've been at least as focused on scalability for Postgres workloads as ML functionality, with our PgCat project. It's a pooler that gives us load balancing, sharding, failover etc to handle large clusters of many machines for application and inference workloads to scale horizontally.

OTOH MindsDB interconnects with just about every data source out there, which you may consider an advantage.