Show HN: jSuites v4 - A library of ultra-light components and plugins free (MIT) https://ift.tt/dhvxQIo

Show HN: jSuites v4 - A library of ultra-light components and plugins free (MIT) https://ift.tt/u6fDKsL February 21, 2024 at 10:47PM

Show HN: LoraLand – 25 fine-tuned LLMs that beat GPT-4 https://ift.tt/JrxoWmb

Show HN: LoraLand – 25 fine-tuned LLMs that beat GPT-4 Hi all, today we're excited to launch LoraLand: 25 fine-tuned mistral-7b models that outperform #gpt4 on task-specific applications ranging from sentiment detection to question answering. All 25 fine-tuned models… - Outperform GPT-4, GPT-3.5-turbo, and mistral-7b-instruct for specific tasks - Are cost-effectively served from a single GPU through LoRAX - Were trained for less than $8 each on average You can prompt all of the fine-tuned models today and compare their results to mistral-7b-instruct in real time! We'd love to hear comments and feedback from the community https://ift.tt/0pXFhgf February 20, 2024 at 10:18PM

Show HN: Hyperdiv – Reactive, immediate-mode web UI framework for Python https://ift.tt/RFA9rTq

Show HN: Hyperdiv – Reactive, immediate-mode web UI framework for Python Hello HN, I'm releasing Hyperdiv ( https://hyperdiv.io ), a framework for rapidly developing reactive browser UIs in Python, with immediate-mode syntax and using Shoelace ( https://shoelace.style ) as its built-in component system. This short coding video will give you a good idea of what it is: https://www.youtube.com/watch?v=4XJKfxaqvGE I wrote a brief article about the motivation and approach: https://ift.tt/mP3sXFk Hyperdiv doesn't aim to compete with serious full-stack frameworks. The core aim was to make it easy and fast to prototype apps and build UI-based tools. I was originally motivated by internal tools at work -- feeling the need to quickly put together UI-based tools to share with both technical and non-technical coworkers, without having to stand up and maintain a full internal stack. This is my first major open source release. I really appreciate your feedback and support. - Marius https://ift.tt/OHvQmzS February 20, 2024 at 07:53PM

Show HN: DMARC Checker https://ift.tt/uy2wWKR

Show HN: DMARC Checker https://ift.tt/FA7kVjp February 20, 2024 at 10:22PM

Show HN: CaveRibbon (SFCave Remake) https://ift.tt/9xyetF1

Show HN: CaveRibbon (SFCave Remake) https://ift.tt/kpP90FG February 19, 2024 at 05:02PM

Show HN: Tl;Dr Voters – Simplifying Democracy with AI Powered Ballot Summaries https://ift.tt/97Ge5Hs

Show HN: Tl;Dr Voters – Simplifying Democracy with AI Powered Ballot Summaries https://ift.tt/OqWymu6 February 19, 2024 at 10:54PM

Show HN: Like HN, but for Science https://ift.tt/ZCjsSwD

Show HN: Like HN, but for Science https://ift.tt/Lp1vTod February 19, 2024 at 03:00AM

Show HN: The History Chronicle – daily historical facts in newspaper form https://ift.tt/qt8fc4u

Show HN: The History Chronicle – daily historical facts in newspaper form https://ift.tt/AW0U1oe February 19, 2024 at 01:56AM

Show HN: Caps-log (Captain's log) – A small TUI journaling tool https://ift.tt/6yWGEsp

Show HN: Caps-log (Captain's log) – A small TUI journaling tool Caps-log is a compact TUI (Text-based User Interface) journaling application crafted in C++ and leveraging the FTXUI library for its terminal interface. It allows users to save daily log entries as simple markdown files, making it an appealing tool for those who prefer working within a terminal environment. The interface is designed with a calendar feature that stands out by marking the days associated with a log entry. Furthermore, it can accentuate days based on specific 'tags' or 'sections' identified in the logs, which are either markdown list items starting with '*' or level one headers. In addition to these features, caps-log includes password protection for your entries and offers a somewhat 'hacky' (for now) method for remote storage. This is achieved by integrating with a pre-configured Git repository, enabling remote storage via a git remote. https://ift.tt/7K0Bf4h February 19, 2024 at 01:07AM

Show HN: I Built an Open Source API with Insanely Fast Whisper and Fly GPUs https://ift.tt/KS8EXQA

Show HN: I Built an Open Source API with Insanely Fast Whisper and Fly GPUs Hi HN! Since the launch of JigsawStack.com, we've been trying to dive deeper into fully managed AI APIs built and fine tuned for specific use cases. Audio/video transcription was one of the more basic things and we wanted the best open source model at this point it is OpenAI's whisper large v3 model based on the number of languages it supports and its accuracy. The thing is, the model is huge and requires tons of GPU power for it to run efficiently at scale. Even OpenAI doesn't provide an API for their best transcription model while only providing whisper v2 at a pretty high price. I tried running the whisper large v3 model on multiple cloud providers from Modal.com, Replicate, and Hugging faces dedicated interface and it takes a long time to transcribe any content about ~30mins long for 150mins of audio and this doesn't include the machine startup time for on-demand GPUs. Keeping in mind at JigsawStack we aim to return any heavy computation under 25s or 2mins for async cases and any basic computation under 2s. While exploring Replicate, I came across this project https://ift.tt/A40xZrI by Vaibhav Srivastav which optimises the hell out of this whisper large v3 model with a variety of techniques like batching and using FlashAttention 2. This reduces computation time by almost 30x, check out the amazing repo for more stats! Open source wins again!! First, we tried using Replicate's dedicated on-demand GPU service to run this model but that did not help, the cold startup/booting time alone of a GPU made the benefits of the optimised model pretty useless for our use case. Then we tried Hugging face and modal.com and we got the same results, with an A100 80GB GPU, we saw around an average of ~2mins start up time to load the machine and model image. It didn't make sense for us to have an always on GPU running due to the crazy high cost. At this point, I was inches away from giving up. The next day I got an email from Fly.io: "Congrats, Yoeven D Khemlani has GPU access!". I forgot the Fly started providing GPUs and I'm a big fan of their infra reliability and ease of deployment. We also run some of our GraphQL servers for JigsawStack on Fly's infra! I quickly picked up some Python and Docker by referring to a bunch of other Github repos and Fly's GPU tutorials, then wrote the API layer with the optimised version of Whisper 3 and deployed it on Fly's GPU machines. And wow the results were pretty amazing, the start up time of the machine on average was ~20 seconds compared to the other providers at ~2mins with all the performance benefits from the optimised whisper. I've added some more stats in the Github repo. The more interesting thing to me is the cost↓ Based on 10mins of audio: - OpenAI Whisper v2 API -> $0.06/10mins - Insanely Fast Whisper large v3 API on Fly GPU (Cold startup) -> ~$0.029/10mins - Insanely Fast Whisper large v3 API on Fly GPU (Warm startup) -> ~$0.011/10mins (Note: These are rough estimates I did by taking averages after running 5 rounds each) If you guys want to run this on any other GPU providers you can as long as they support Docker. We'll be optimising this more over the next few days specific to Fly's infrastructure allowing for global distributed instances of whisper and will soon be providing a fully managed API on JigsawStack.com. Stay tuned! https://ift.tt/fqmatjT February 18, 2024 at 01:48PM

Show HN: Programming is easier than you think https://ift.tt/iaqJCFW

Show HN: Programming is easier than you think https://ift.tt/BqjTktw February 17, 2024 at 10:23PM

Show HN: Polybar Module for Using AirPods https://ift.tt/xG3F0sR

Show HN: Polybar Module for Using AirPods https://ift.tt/EpMWac9 February 18, 2024 at 01:06AM

Show HN - tool that converts image receipts to Excel https://ift.tt/XfZ5PEU

Show HN - tool that converts image receipts to Excel Hey I'm excited to share my first project, Receipts2CSV, a web application designed to simplify bookkeeping by converting receipt images into CSV files. https://ift.tt/4irGm0x Problem Statement: Keeping track of expenses and managing receipts can be a tedious task, especially for small businesses and freelancers. Traditional methods involve manually entering data from receipts into spreadsheets, which is time-consuming and prone to errors. With Receipts2CSV, users can streamline this process by simply uploading images of their receipts and obtaining structured CSV files ready for import into accounting software. If you are lazy like I am, you could accumulate receipts in just one folder and re-run all images every time, remove duplicates and merge with older CSV to minimize looking through receipts on a monthly/annual basis. Questions for Validation: Do you find a receipt image to CSV converter useful? Would you consider using such a tool for your bookkeeping needs? Considering higher costs for AI models like GPT-4 Vision Preview, how are other indie hackers able to create and sustain offering free products like these? Do small products like these have a monetization market? If so, where do I begin? Curious to hear your candid thoughts about this web app. Should I explore it further or move to the next idea? Feel free to share your thoughts, suggestions, or any additional features you'd like to see in the product! Thank you for your valuable input and support! https://ift.tt/xWA3UwJ February 17, 2024 at 11:35PM

Show HN: Host a planet-scale geocoder for $10/mo https://ift.tt/IORLiE5

Show HN: Host a planet-scale geocoder for $10/mo For the uninitiated, a geocoder is maps-tech jargon for a search engine for addresses and points of interest. Geocoders are expensive to run. Like, really expensive. Like, $100+/month per instance expensive unless you go for a budget provider. I've been poking at this problem for about a month now and I think I've come up with something kind of cool. I'm calling it Airmail. Airmail's unique feature is that it can query against a remote index, e.g. on object storage or on a static site somewhere. This, along with low memory requirements mean it's about 10x cheaper to run an Airmail instance than anything else in this space that I'm aware of. It does great on 512MB of RAM and doesn't require any storage other than the root disk and remote index. So storage costs stay fixed as you scale horizontally. Pretty neat. Demo here: https://ift.tt/2XiAWKf Writeup: https://ift.tt/gUJOTce... Repository: https://ift.tt/XpoMWb4 https://ift.tt/TIoKXH0 February 16, 2024 at 09:51PM

Show HN: A real-time speech-language model for $10 of training https://ift.tt/6ekOitJ

Show HN: A real-time speech-language model for $10 of training https://tincans.ai/slm February 16, 2024 at 11:53PM

Show HN: Galactic Compass – an app that points to the galactic center https://ift.tt/Fc4hBZA

Show HN: Galactic Compass – an app that points to the galactic center I made a single-serving iPhone app with a floating green arrow that always points to the center of the Milky Way, 26,000 lights years away. It’s weirdly grounding. I’ve never made an app before, so here’s how it worked with ChatGPT. https://ift.tt/NgtbZBJ February 16, 2024 at 03:42AM

Show HN: BotArchive – Print Directly to Google Drive from Windows https://ift.tt/pstLoQk

Show HN: BotArchive – Print Directly to Google Drive from Windows I built BotArchive ( https://ift.tt/VPUtzjT ) to make it easy to save content to cloud storage: you connect it to your Google Drive, Microsoft OneDrive, Dropbox or Box, and then anything you send to it gets automatically saved in the attached storage location. I'd originally built it for Telegram and Slack, but I've been long interested in doing things with print: printing has long felt like one of the biggest underutilized output processes out there. Almost every piece of software has print integration, the output format is usually legible and for things like receipts I always find myself printing them out, and then copying them to Google Drive. So I built a custom IPP printer: no drivers needed on Windows, OSX or Android (haven't tried iOS yet but I assume its the same), easy configuration and prints PDFs that appear right in your cloud storage location of choice (handy replacement for the missing Google Cloud Print feature). Future work includes parsing the printed PDFs for things like conversion into a Kindle book, or extracting financial data. Feedback welcome! https://ift.tt/3KwWTDc February 16, 2024 at 01:02AM

Show HN: SlideCross – a combination of a crossword and Rubik's Cube https://ift.tt/SleiX3z

Show HN: SlideCross – a combination of a crossword and Rubik's Cube Hello! Try out a 3x3 puzzle and click the question mark in the top right for a tutorial. I'd really appreciate any feedback! Thanks https://slidecross.io February 16, 2024 at 12:11AM

Show HN: NeuralFlow – Visualize the intermediate output of Mistral 7B https://ift.tt/KvTDHqZ

Show HN: NeuralFlow – Visualize the intermediate output of Mistral 7B https://ift.tt/zUD9Ejy February 15, 2024 at 08:59AM

Show HN: Gitlab Meeting Simulator 2024 https://ift.tt/coXaGbY

Show HN: Gitlab Meeting Simulator 2024 Gitlab's meeting recordings on YouTube have tens of thousands of views by people pretending to work. Now you can appear to be in the meeting using your own webcam. https://ift.tt/NinH2x6 February 15, 2024 at 05:52AM