Show HN: Example POC to run the Dolly V2 12B model in 8bit mode https://ift.tt/xJ28lHU

Show HN: Example POC to run the Dolly V2 12B model in 8bit mode I banged this out today because nobody seems to have figured this out yet (and I wanted to play with this model). I was inspired by the fact that Simon Willison was struggling to get this running[0], so I wanted to take a crack at it! This is by no means "clean", but it is a baseline for others to hack from. Feel free to send over PRs and I'll land them! (I'd love it somebody used this to glue this into this webui[1].) 0: https://ift.tt/h0p53gH 1: https://ift.tt/sPeqfQu https://ift.tt/PCLIRlQ April 15, 2023 at 04:16AM

Show HN: A curated list of remote jobs https://ift.tt/0NI5nzV

Show HN: A curated list of remote jobs https://ift.tt/QEANgym April 15, 2023 at 12:37AM

Show HN: Don't Hold Your Breath – Breathing Analysis with Polar H10 HR Monitor https://ift.tt/hyZMV5K

Show HN: Don't Hold Your Breath – Breathing Analysis with Polar H10 HR Monitor A person's heart rate fluctuates based on factors such as activity, environment, and mental state. Additionally, heart rate variations occur on a much shorter timescale, with changes happening every few seconds. During inhalation, heart rate slightly increases, and it decreases again during exhalation, a phenomenon known as respiratory sinus arrhythmia. There are several reasons for this. Firstly, an increase in volume within the thoracic cavity (the space between the ribs) leads to a corresponding drop in pressure, which is counteracted by a rise in heart rate to regulate blood pressure. Secondly, breathing is deeply intertwined with the autonomic nervous system, which involves the opposing activation of the parasympathetic ("rest and digest") and sympathetic ("fight or flight") responses. By measuring heart rate variability and regulating our breath, we can gain insight into our stress response and learn to control it. The Polar H10 heart rate monitor provides basic heart rate measurements from a chest strap ECG, compatible with a number of fitness tracking apps that monitor heart rate zones wirelessly using bluetooth. The device also provides raw sensor data including an accurate measure of the timing between heartbeats (interbeat interval, with millisecond resolution) and 200 Hz accelerometer data. In this project this information is used to measure breathing rate and heart rate variability, and explore the relationship between the two. https://ift.tt/lAHErQg April 15, 2023 at 12:36AM

Show HN: Using zod to get structured and typed output from ChatGPT in TypeScript https://ift.tt/oYNLiDt

Show HN: Using zod to get structured and typed output from ChatGPT in TypeScript https://ift.tt/HYAWf2k April 14, 2023 at 03:47AM

Show HN: Ampersand – “Terraform for Integrations” https://ift.tt/XACIJoR

Show HN: Ampersand – “Terraform for Integrations” Hey HN! I’m one of the co-founders of Ampersand and we’re excited to launch our developer docs today! Ampersand is built for SaaS engineers who want to: - Read data from their customers’ SaaS (e.g. Salesforce, Hubspot). - Write data to their customers’ SaaS. - Subscribe to events happening in their customers’ SaaS (including creates, deletes and field updates). We feature a Terraform-like declarative interface for defining integrations, embeddable UI components for your users to set up and manage their integrations, and a management console for your team to monitor and troubleshoot. We’re launching our docs ahead of our product because we believe that docs are one of the most important things to get right in building a developer product, and we want to get early feedback on it. Looking forward to hearing your feedback in the comments below! https://ift.tt/tnW9CKm April 14, 2023 at 02:38AM

Show HN: Extract clips from YouTube (without downloading whole video) https://ift.tt/8t6nXwp

Show HN: Extract clips from YouTube (without downloading whole video) https://ift.tt/ya1AwvQ April 13, 2023 at 10:00PM

Show HN: Google Analytics alternative with the most generous free tier https://ift.tt/Sje7CaQ

Show HN: Google Analytics alternative with the most generous free tier https://ift.tt/LYiTbzV April 12, 2023 at 07:10PM

Show HN: ZP a cli cmd built in rust to copy things to the clipboard https://ift.tt/poNVDAO

Show HN: ZP a cli cmd built in rust to copy things to the clipboard `zp` is a custom command used to copy the contents of a source file or of the std output buffer to the clipboard, allowing users to easily paste the contents into another file or program. https://ift.tt/sg3xzVd April 14, 2023 at 12:21AM

Show HN: NeverPrinter – find and use thousands of nearby printers https://ift.tt/J9XkiZD

Show HN: NeverPrinter – find and use thousands of nearby printers Hi all, My friends and I have created a Chrome extension and we are looking for feedback on whether people find it useful or not. The extension is intended to serve as a home printer substitute for people (like us) who don't want or need their own printer at home. It makes it easy to send documents to nearby office supply / shipping stores for printing and pickup. Installing the extension adds a printer named NeverPrinter to your list of available printers in Chrome. When you print a document with NeverPrinter, a new tab opens allowing you to select which nearby store you want to send the document to. This is the latest iteration in an ongoing project my friends and I have been working on. We would love to hear any positive or negative feedback. You can get the extension by going to the Chrome Web Store and searching for "NeverPrinter." Or you can visit https://ift.tt/4heqwc1 , which will redirect you there. https://ift.tt/0xg7UhP April 13, 2023 at 11:18PM

Show HN: A minimalist net worth tracker for individuals https://ift.tt/9RGdiE6

Show HN: A minimalist net worth tracker for individuals Hi HN, Today I'm launching the Net Worth Tracker by Flow Finance, a beautiful and minimalist Excel template that allows individuals to track their assets, liabilities, income, and spending on a monthly basis. The focus of this template is understanding long-term trends in the incoming and outgoing expenses for an individual—always with a focus on maximising insights while minimising user inputs. This project was created because the alternatives, frankly, are completely overwhelming. A friend of mine introduced me to the FIRE Movement and, after doing deep dives through Reddit, I tried to find templates to help facilitate a financial tracking system of my own. Every template felt like complete overkill: portfolio tracking for shares, mortgage repayment calculators, even a kind of speedometer which moved depending on my savings rate. Clever, no doubt, but unnecessary for my purposes. After years of tinkering and refining what I truly wanted out of this spreadsheet, I realised it was to clearly understand the flows of my finances. This spreadsheet does not attempt to compete with a typical FIRE spreadsheet or tools like YNAB/Mint, it is about capturing the big picture once a month and essentially taking a snapshot of my financial health. I've priced this spreadsheet at $15 which I believe is a fair price for the value you'll see in taking more control over your personal finances, however as an added incentive (and because I am proud of what I've created and would love more people to use it), I'm currently offering a 50% discount to all HN users, who can use the code f3g27mt at the Gumroad checkout. Thank you for your time and I sincerely hope this product can assist you as much as it has for me. I am more than happy to answer any questions you may have so please feel free to comment below or reach out at contact@getflow.finance. Thank you! https://ift.tt/LFM6QZ3 April 13, 2023 at 04:37AM

Show HN: GPTCache – Redis for LLMs https://ift.tt/oYbKUSa

Show HN: GPTCache – Redis for LLMs Hey folks, As much as we love GPT-4, it's expensive and can be slow at times. That's why we built GPTCache - a semantic cache for autoregressive LMs - atop the vector database Milvus and SQLite. GPTCache provides several benefits: 1) reduced expenses due to minimizing the number of requests and tokens sent to the LLM service 2) enhanced performance by fetching cached query results directly 3) improved scalability and availability by avoiding rate limits, and 4) a flexible development environment that allows developers to verify their application's features without connecting to LLM APIs or network. Come check it out! https://ift.tt/Qo3G1PH https://ift.tt/jlJ9VLN April 13, 2023 at 03:14AM

Show HN: Let's find Pokémon that are the same height and weight as you https://ift.tt/4ae39km

Show HN: Let's find Pokémon that are the same height and weight as you https://ift.tt/pntJH3e April 13, 2023 at 05:51AM

Show HN: Twitter API Reverse Engineered https://ift.tt/G8WwYEZ

Show HN: Twitter API Reverse Engineered endpoints: /graphql /1.1 /2 /i https://ift.tt/PNHXOJR April 13, 2023 at 05:20AM

Show HN: I’m building an autonomous business run by ChatGPT https://ift.tt/lNzoQ04

Show HN: I’m building an autonomous business run by ChatGPT Hey HN!! I just shipped a project called Airfeed: https://airfeed.co/ Here's how it works: 1. You tell it the research topics you're interested in following This can be anything -- LLMs, chatbots, image generation, scaling model inference, small language models, datasets, etc. 2. Every morning, it filters new AI research papers for quality work that is relevant to your configured topics, and sends you an email summary. Here's a sample email: https://ift.tt/nxPNYaS --- As a fun experiment, I've made everything completely autonomous: Every morning, ChatGPT takes care of filtering/processing all research papers & writing each email. --- Why I built this: I think there needs to be a better way for engineers, researchers, and AI product builders to stay up to date on the latest AI research that's relevant to whatever they're working on. For someone building LLMs, this could mean keeping up with research papers covering multiple topics, ex: scaling attention, model architecture, new datasets, new fine-tuning techniques, etc. --- How this is different from other newsletters: This isn't a generic AI newsletter. It's research-focused & personalized, since it only highlights the papers that might be relevant to you (based on your specific topics). I found this to be useful, since a lot of papers are irrelevant and/or low quality. --- When building maroofy.com [1], I would regularly read the latest research papers on everything related to: contrastive learning, vector embeddings, music classification, etc. So, I thought it would be nice to build something that filters/summarizes just the new research papers covering topics I care about, and sends me the top ones in a daily, morning email. Would love any feedback!! :D [1] https://ift.tt/XIl9g05 https://airfeed.co/ April 13, 2023 at 03:22AM

Show HN: GPT-4-powered web searches for developers https://ift.tt/h2sEkx9

Show HN: GPT-4-powered web searches for developers Hi HN, Today we’re launching GPT-4 answers on Phind.com, a developer-focused search engine that uses generative AI to browse the web and answer technical questions, complete with code examples and detailed explanations. Unlike vanilla GPT-4, Phind feeds in relevant websites and technical documentation, reducing the model’s hallucination and keeping it up-to-date. To use it, simply enable the “Expert” toggle before doing a search. GPT-4 is making a night-and-day difference in terms of answer quality. For a question like “How can I RLHF a LLaMa model”, Phind in Expert mode delivers a step-by-step guide complete with citations ( https://ift.tt/Mn2i4hd... ) while Phind in default mode meanders a bit and answers the question very generally ( https://ift.tt/pR2Ccro... ). GPT-4 is significantly more concise and “systematic” in its answers than our default model. It generates step-by-step instructions over 90% of the time, while our default model does not. We’re particularly focused on ML developers, as Phind can answer questions about many recent ML libraries, papers, and technologies that ChatGPT simply cannot. Even with ChatGPT’s alpha browsing mode, Phind answers technical questions faster and in more detail. For example, Phind running on “Expert” GPT-4 mode can concisely and correctly tell you how to run an Alpaca model using llama.cpp: ( https://ift.tt/WN5XjIQ... ). In contrast, ChatGPT-4 hallucinates and writes a make function for a fictional llama.cpp. We still have a long way to go and would love to hear your feedback. https://www.phind.com April 12, 2023 at 11:14PM

Show HN: Automatically Generate Company Names with Available Domains https://ift.tt/8piXSG7

Show HN: Automatically Generate Company Names with Available Domains I have found myself running into the issue where I think of a great project or company name, but the domain is already taken multiple times. I decided to create this simple project which uses openai and route53 (hence the name) to generate available domains for a company based on keywords, rather than just come up with new company names. https://53names.com April 12, 2023 at 01:46AM

Show HN: Discord Time formatting with natural language processing https://ift.tt/FnyI3Mg

Show HN: GPT–LLM native macOS app with time travel, versioning, search https://ift.tt/Ghvck3s

Show HN: GPT–LLM native macOS app with time travel, versioning, search Hey everyone! I made a Mac app for exploring large language models. It’s fast, small, has a tiny memory footprint. It’s immutable by design with both immediate time travel and automatic versioning as foundational elements. The app is written in Swift and a bit of Rust for the tokenizer. I used SwiftUI for structure and animations and Cocoa for advanced behavior. All storage is SQLite and local-only. You can go through the database as needed and backup it as well. The app has support for variants, which is the `n` parameter in the OpenAI chat completion API—equivalent to the drafts feature in Bard. It shows model- and ChatML-aware tokens and cost, and it dynamically adjusts response length in function of prompt length and context length. It supports hiding runs that are not useful, marking the ones that are notable, and going through the ancestry of the current run. It also has support for examples, as specified in ChatML, conversation names and personal and local-only notes. The full-text search works across all stored text and it has support for `all`, `any`, `prefix` and `phrase` matching; the results are time-ordered rather than ranked for the time being. There are several shortcuts and more are coming so that the app can be used entirely by keyboard. You can export to JSON and all data is exported. The app is sandboxed and notarized. The API keys are stored securely in the Keychain. Next I plan to implement combinatorial runs (mixing multiple values of the same parameter and multiple models as well) and full Markdown support. Currently only OpenAI’s conversational models are supported (GPT 3.5, GPT 4 8K, GPT 4 32K), but I’ll be adding support for local models and custom ChatML endpoints as soon as possible. Give it a try. Any and all feedback welcome! Thoughts and questions as well! William https://thellm.app April 11, 2023 at 11:34PM

Show HN: Profile Writer for Dating Apps https://ift.tt/rIMlvje

Show HN: Profile Writer for Dating Apps Hi folks. We built a fun app that asks a few questions and spins them into a dating app profile. Would love to hear what you think. https://ift.tt/fw8HR2i April 11, 2023 at 08:56PM

Show HN: Local-first secret manager for developers https://ift.tt/Dda1mlS

Show HN: Local-first secret manager for developers https://ift.tt/eWv8SHO April 11, 2023 at 08:45PM