Skip to main content

Posts

Showing posts from February 26, 2023

Show HN: gpt-graph. A simple, GPT-3 text to entity-relation graph generator https://ift.tt/4VXdS3j

Show HN: gpt-graph. A simple, GPT-3 text to entity-relation graph generator Hi HN! This is a simple text to entity-relation graph generator, powered by gpt-3 davinci model. The purpose is to feed it actual written data, to obtain a graph representation of entities and relationships mentioned in the text. Also, being able to identify entity attributes like gender, size, age ... My initial goal, was to make it able to process a large amount of text into a big single graph. The problem being the 4000 token limit the model has, I decided to take the approach of feeding the text in batches, and try to merge the incoming graph with the existing information each time. This is done by comparing the incoming node labels with those already in the graph, adding the new information to the existing nodes. This works somewhat, but sometimes entities get duplicated if they are mentioned slightly differently in the text. The comparation method could use some improvement clearly. A nice feature, is tha

Show HN: Community Building as a Service” https://ift.tt/MZn6JmR

Show HN: Community Building as a Service” Launched this week. Our team are experts in building online and offline communities from scratch. We've been helping our friends build this for free in the past. What we offer is best practices and resources for onboarding new members, moderating, and and growing online communities on Slack or Discord. We would love to get honest feedback and criticisms. https://ift.tt/IWZjJcz March 4, 2023 at 05:36AM

Show HN: Zipslicer, a library for loading LLM checkpoints on consumer hardware https://ift.tt/7idfN5M

Show HN: Zipslicer, a library for loading LLM checkpoints on consumer hardware This is a low-level opensource library I developed for my own use and decided to share, as it makes it possible to process large checkpoints of neural networks without renting high-RAM instances, on a regular PC. It replaces torch.load() with a custom function that produces a dictionary that materializes tensors on the fly. Compared to other solutions it doesn't require sharding or re-encoding checkpoints and uses them completely as-is. It is a foundation to make it possible to run inference and compress language models and other large models one layer at a time - in principle, even one tensor at a time. I describe the rationale and technical details of the library's design in the blogpost: https://ift.tt/tydo8Uh https://ift.tt/OPafXBl March 4, 2023 at 01:29AM

Take Muni’s Safety Survey!

Take Muni’s Safety Survey! By Greer Cowan Everyone should feel safe on Muni. Help make Muni safer by taking the SFMTA’s survey about personal safety and harassment in the Muni system.   As part of the MuniSafe Safety Equity Initiative launched in August 2022, the SFMTA has partnered with the UCLA Institute of Transportation Studies to better understand Muni riders’ experiences and develop safety recommendations, tools and policies aimed at preventing harassment and assault on Muni.  Take the 5-minute survey   Survey information will also help the SFMTA understand Muni customers’ specific safety needs and challenges, and help us identify valuable trends and patterns so we know where, when and how harassment shows up in the Muni system.  Incidents often go unreported, but the SFMTA is working to change that with new reporting options. As of October 2022, Muni customers can report incidents of  harassment by calling 311, using the 311 mobile app or using the Muni Feedback form at

Show HN: Hire from Communities (Fromcommunity.com) https://ift.tt/0saFKhu

Show HN: Hire from Communities (Fromcommunity.com) Get your best resources from well-known communities within the industry, such as forums, online groups, or active community channels, where they have built a reputation for their skills and contributions. We are happy to get feedback and join forces :) https://ift.tt/4nWvd86 March 2, 2023 at 10:49PM

Show HN: Sort Any Awesome List by GitHub Stars https://ift.tt/M0Ulptr

Show HN: Sort Any Awesome List by GitHub Stars As a CS undergrad, I self-taught many topics not covered by the school. Typically awesome list is where I started with. I spent my second half in college researching computer vision and machine learning. CVPR accepts over 2000 papers every year, not to mention arXiv. The awesome lists curated by the community usually serve as checklists when I was doing literature surveys. But since the number of the list items was overwhelming, I had to prioritize which papers to read first. Many of them had been uploaded to arXiv in the last few months, so I had no clue how many citations they would get. When learning web development, I also had difficulty finding the best package for the tasks like data validation and client-side routing. To make a decision, I had to browse through GitHub repositories or read the documentation. I wish I could have my awesome lists sorted, desirably by popularity. This led to the creation of my project: a web app that re

Show HN: Mathesar – open-source collaborative UI for Postgres databases https://ift.tt/9NYqOGh

Show HN: Mathesar – open-source collaborative UI for Postgres databases Hi HN! We just released the public alpha version of Mathesar ( https://mathesar.org/ , code: https://ift.tt/l5SbsAI ). Mathesar is an open source tool that provides a spreadsheet-like interface to a PostgreSQL database. I was originally inspired by wanting to build something like Dabble DB. I was in awe of their user experience for working with relational data. There’s plenty of “relational spreadsheet” software out there, but I haven’t been able to find anything with a comparable UX since Twitter shut Dabble DB down. We're a non-profit project. The core team is based out of a US 501(c)(3). Features: * Built on Postgres : Connect to an existing Postgres database or set one up from scratch. * Utilizes Postgres Features : Mathesar’s UI uses Postgres features. e.g. "Links" in the UI are foreign keys in the database. * Set up Data Models : Easily create and update Postgres schemas and tables. * Data Entry

Show HN: Try out the new ChatGPT API on Promptly https://ift.tt/CGURjTE

Show HN: Try out the new ChatGPT API on Promptly Hey HN Community, We're excited to announce the integration of the newly launched ChatGPT API into Promptly - a platform designed to make prompt management and sharing a breeze for developers. With Promptly, you can easily test out different prompts and model parameters for various providers, and quickly share prompt snippets together with parameters and generated output. It's like CodePen or JSFiddle, but for prompts! In addition to that, Promptly also allows you to create high-level endpoints on top of provider APIs (such as Open AI, DreamStudio, and more) with templated and versioned prompts. And with built-in caching for endpoints, you can save on Open AI costs and improve latency. Today, we're thrilled to add the ChatGPT API to our platform. So head on over to Promptly and try it out for yourself! It's easy, intuitive, and completely free to use. Check it out at https://ift.tt/Frp8kus We can't wait to see what am

Show HN: Supaglue – open-source unified API https://ift.tt/DcPntXv

Show HN: Supaglue – open-source unified API Last month, we open-sourced Supaglue (https://ift.tt/1GIAkxg) as a developer toolkit for building customer-facing Salesforce integrations. Since then, we've been iterating on our approach with users and are re-launching Supaglue as an open source unified API, starting with CRMs. With our re-launched public alpha you can: - Interact with HubSpot and Salesforce through a unified REST API - Sync data from HubSpot and Salesforce into a local Postgres cache and make reads against a normalized CRM schema - Make writes (POST and PATCH) to HubSpot and Salesforce using the unified API - Open source MIT license so anyone can self-host for free Today we support reads, creates, and updates for Accounts, Contacts, Leads, and Opportunities for HubSpot and Salesforce. In the coming weeks, we plan to add more connectors and features like configuring sync frequencies, webhooks for notifications, sync rate limiting/throttling. We'd love to hear your th

Show HN: Rapid Kubernetes Controller Development with Tilt and Silly Drawings https://ift.tt/nmr1tox

Show HN: Rapid Kubernetes Controller Development with Tilt and Silly Drawings Like the says. I'm describing how to use Tilt to speed up the dev cycle of testing and coding Kubernetes controllers. Tilt uses a built in local registry to push images into and can use hot swapping to kickly deploy processes in containers. It's pretty neat. :) The Tiltfile can be a bit daunting but I'm trying to break it down as much as possible. :) I hope this helps! Good luck! https://ift.tt/E3r24xv March 1, 2023 at 06:08PM

Show HN: Graph-based roadmap to learn anything 10x faster https://ift.tt/UxLgdjK

Show HN: Graph-based roadmap to learn anything 10x faster Hey guys, I am founder of Neuton. I have built roadmap playground that guide student to learn skills faster with the help of progress tracking and interactive graph-based roadmap. To read more: https://ift.tt/MoVqLCY... I would love to get some feedbacks. https://ift.tt/c4pm6PD March 1, 2023 at 07:55PM

Show HN: Dak – a Lisp-like language that transpiles to JavaScript https://ift.tt/LS8WgPw

Show HN: Dak – a Lisp-like language that transpiles to JavaScript Hi HN, author here. Happy to answer any questions. I had an itch to make a lisp like language that was a thin layer on top JavaScript. Something that could leverage the thriving ecosystem that exists around JavaScript. It's brittle, hot off the oven. Besides being a fan of parenthesis, I think macros fill in a gap that the JavaScript ecosystem today fills in with one-off compilers, bundler plugins and such. Macros can't do everything, but for example I think they have the potential to enable things like JSX, Solid and Svelte style libraries. Take the tour to get a feel for what it can do and play with the live code in your browser! https://ift.tt/jrqY9eQ February 27, 2023 at 07:08PM

Show HN: Crul – Query Any Webpage or API https://ift.tt/SAP7KgZ

Show HN: Crul – Query Any Webpage or API Hi HN, we’re Carl and Nic, the creators of crul ( https://www.crul.com ), and we’ve been hard at work for the last year and a half building our dream of turning the web into a dataset. In a nutshell crul is a tool for querying and building web and api data feeds from anywhere to anywhere. With crul you can crawl and transform web pages into csv tables, explore and dynamically query APIs, filter and organize data, and push data sets to third party data lakes and analytics tools. Here’s a demo video, we’ve been told Nic sounds like John Mayer (lol) ( https://ift.tt/ec8b6Qx ) We’ve personally struggled wrangling data from the web using puppeteer/playwright/selenium, jq or cobbling together python scripts, client libraries, and schedulers to consume APIs. The reality is that shit is hard, doesn’t scale (classic blocking for-loop or async saturation), and comes with thorny maintenance/security issues. The tools we love to hate. Crul’s value prop is s

Show HN: DbDeclare – A Python declarative layer for your database https://ift.tt/rqwMl7T

Show HN: DbDeclare – A Python declarative layer for your database Hi HN! I made and just published v0.0.1 of DbDeclare. I use Python a lot, and interact with Postgres a lot. I like using SQLAlchemy, and I love Alembic. Those wonderful tools primarily operate on tables, though, and I often find myself writing custom code to declare what databases, roles, schemas, privileges, etc. I want, and I have a hard time updating them reliably and in a repeatable fashion. That's where DbDeclare aims to help: declare what you want in your cluster (in addition to SQLAlchemy-defined tables and columns) in-code, alongside your tables. There is a lot this can't do yet (thus the v0.0.1), but I think there's a decent foundation here to build on and eventually have really nice features like autogenerating change statements between your in-code definition and what is actually in your database cluster (like Alembic). This is also my first attempt at building an open-source project, so I'm su

Show HN: Sup – one liners to distribute binaries https://ift.tt/jqnmp6P

Show HN: Sup – one liners to distribute binaries Hello all, I found myself rewriting the same shell scripts to distribute my binaries over and again, so I decided to put them in a repository and serve them with a prettier URL - to have nicer READMEs for my projects. I'll add some other tasks like installing/uninstalling on Windows, but in general ideas are welcome. Maybe somebody else finds them useful! https://ift.tt/94rsRCf February 27, 2023 at 01:46PM