Show HN: Turn your available ingredients into recipes https://bit.ly/49zf6ny

Show HN: Turn your available ingredients into recipes https://bit.ly/42LqeLE February 21, 2024 at 09:36AM

Show HN: Deploy highly available infra to EC2 with Docker-compose and CDK https://bit.ly/3OMazpO

Show HN: Deploy highly available infra to EC2 with Docker-compose and CDK I created a CDK deploy that uses docker-rollout [1][2] to deploy highly available infrastructure to EC2 using only autoscaling groups. It is not super polished but it is a complete example, so it could be useful if you are considering hosting on EC2. Rolling out deploys involves updating one file on S3 and running one script. Ironically after all that setup, I decided to give Linode with k8s a try [3] :-) (due to aws' high costs of egress and NAT gws / IPv4 tax on AWS, and the fact that some apps that I want to run are easier to deploy with helm). More notes: * I did try ECS and Fargate, which are nice, but also come with associated costs and a bunch of complexity. At that point, I rather spend time directly with k8s, which should make my localhost parity way higher, and hosting somewhere more affordable. * I tried both Pulumi and Terraform. I have mixed feelings about them. I ended up using CDK because it _felt_ like the nicer development experience (except when CloudFormation fails and it kind of hides the reason why, sigh ... fishing for logs on CloudWatch is such a drag!). * I tried to add some NACL rules since I ended up running the thing on a public VPC. I couldn't make it work but at that time I had already decided to host elsewhere so I left it like that :-). I did succeed on adding support for AWS WAF. Sadly, the cdk currently doesn't have high level support for WAF so it was not as nice to setup. -- 1: https://bit.ly/3JL4lVn 2: https://bit.ly/3HEK6pL 3: https://bit.ly/3OMaAtS... https://bit.ly/49nPgD3 February 21, 2024 at 02:59AM

Show HN: I built my own CDN to host over 20k websites https://bit.ly/3UIxDth

Show HN: I built my own CDN to host over 20k websites https://bit.ly/3SKTltW February 20, 2024 at 06:19PM

Show HN: I made a site that creates the documentation of your code https://bit.ly/49ioE6v

Show HN: I made a site that creates the documentation of your code https://bit.ly/48lHHeY February 20, 2024 at 09:20AM

Show HN: I Made an RSS to Tweet Generator in 2 Hours with ChatGPT https://bit.ly/3I7FetY

Show HN: I Made an RSS to Tweet Generator in 2 Hours with ChatGPT Hey HN. I've made a tool that parses RSS feeds and creates ready-to-post Tweets with ChatGPT. https://bit.ly/49CpIlp February 20, 2024 at 08:57AM

Show HN: Is It FDR? https://bit.ly/3SI0lYB

Show HN: Is It FDR? Is it authentic audio recording from FDR, or AI voice clone? https://bit.ly/3SO0mds February 20, 2024 at 01:15AM

Show HN: We have built an API to extract data from unstructured documents https://bit.ly/3T4J6Sz

Show HN: We have built an API to extract data from unstructured documents https://bit.ly/48iX1c1 February 19, 2024 at 11:46PM

Show HN: RAM Prices https://bit.ly/3uFvALU

Show HN: RAM Prices I was inspired by this discussion https://bit.ly/3I2svIG about diskprices.com last month, and decided to go ahead and make a site for RAM. It's my first time building anything like this! Any tips / suggestions / calls for complete overhaul are welcome :) https://bit.ly/3uGO3HZ February 19, 2024 at 03:39AM

Show HN: I built a vector database API on Cloudflare https://bit.ly/49FBDis

Show HN: I built a vector database API on Cloudflare https://bit.ly/49D7PTE February 19, 2024 at 05:24AM

Show HN: Like HN, but for Science https://bit.ly/3usfLbu

Show HN: Like HN, but for Science https://bit.ly/49kDFVt February 18, 2024 at 10:30PM

Show HN: FakeMyDb – Fake data generator for your RDBMS https://bit.ly/4bJhLMW

Show HN: FakeMyDb – Fake data generator for your RDBMS Introducing fake data generator for your relational database. Are you sick of manually entering data into your database for every darn table that you have before you could test, debug and deploy? With this tool, all you have to do is click a few buttons and then copy the data to your database! easy-peasy-lemon-squeezy. https://bit.ly/4bKOa5R February 18, 2024 at 03:47AM

Show HN: I create a free website for download YouTube transcript, subtitle https://bit.ly/3OOzsAX

Show HN: I create a free website for download YouTube transcript, subtitle https://bit.ly/42FE4iN February 18, 2024 at 10:41AM

Show HN: I Built an Open Source API with Insanely Fast Whisper and Fly GPUs https://bit.ly/3UMJPsS

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://bit.ly/3weFJiY 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://bit.ly/42GNs5J February 18, 2024 at 09:18AM

Show HN: Chat-focused RAG with automated memory management https://bit.ly/3wjg6xJ

Show HN: Chat-focused RAG with automated memory management https://bit.ly/3witW3g February 17, 2024 at 06:01AM

Show HN: A Flask-Based Internet Radio Player Built in Hylang https://bit.ly/42H6BnY

Show HN: A Flask-Based Internet Radio Player Built in Hylang I ported this small project to Hylang a few weeks ago. It calls out to mpd and mpc, but the most exciting part is that it is written in Hylang, a Lisp that compiles down to Python. It pairs well with [1]. Be sure to use the direct links. [1] https://bit.ly/49hXI6V https://bit.ly/42KmGth February 16, 2024 at 10:30PM

Show HN: Magika: AI powered fast and efficient file type identification https://bit.ly/42EsjZV

Show HN: Magika: AI powered fast and efficient file type identification We are very excited to announce the release of Magika our AI powered fast and efficient file type identification lib and tool - https://bit.ly/3uAeFu3 Thanks to its optimized Keras model, large scale training dataset, and Onnx Magika massively outperform other file identification tools while be very fast even on CPU. Magika python code and model is open sourced on Github: https://bit.ly/42EskwX and we also provide an experimental TFJS based npm package https://bit.ly/3I1VyfB With the team we hope you will find Magika useful for your own projects. Let us know what you think or if you have any question! February 16, 2024 at 02:08AM

Show HN: Old-school Unix talk for the web https://bit.ly/48jo1YX

Show HN: Old-school Unix talk for the web Posted 12 years ago, but (a) it's been completely rewritten, and (b) it looks awesomely retro now, instead of just some textarea boxes. We would love feedback and PRs. Please note that if you're testing it "solo" - your second browser window should be in an incognito window. https://bit.ly/3wjIXBS February 15, 2024 at 10:40PM

Show HN: Erogen AI – Safe, ethical, and accessible romantic AI companionship https://bit.ly/3OHBxyE

Show HN: Erogen AI – Safe, ethical, and accessible romantic AI companionship Like many others on the recent internet, I've found the idea of AI companionship to be sort of an attractive idea. At the very least, some element of AI companionship seems inevitable in our daily lives after 5-10 years, whether its through video games or on our phones / in our eyes and ears, etc. I had been playing around with locally hosted options for AI chatbot companions but found the interfaces clunky, and also I was really disheartened by the direction of the content and quality I was seeing on existing sites online. I saw some that were completely SFW-handicapped so as to sprout entire communities around filter-breaking, making safeguard nullification the standard mode by your users (dangerous) simply to get some affection (which seems unreasonable given how much violence and sex we're exposed to in TV, movies, and online). On the other end of the spectrum, I saw a few sites that were completely off the rails with its content free-for-all so as to potentially give AI companionship a really bad name. So I challenged myself to create a more polished experience that would be safe, ethical, and accessible to the mainstream. You can find the site at https://bit.ly/3I3gyT8 (nsfw) I shared this site on Reddit a few months ago asking for feedback on the beta version and it's been slowly gaining traction. I just made significant AI speed improvements for site users this month, added the ability for the AIs to change their avatar images dynamically based on your conversation, and implemented customizable AI voices (still adding options), and I'm ready to unveil the new website to the general public in its more complete form. A few upcoming features I'm excited about include long-term memory and speech-to-text for full voice conversations. I'm also exploring avatar generation on the fly as well as moving avatars. The name "Erogen" and its content is meant to be very much oriented towards romance, love, and NSFW - I think that safe romantic and adult roleplay with a premium feel will be essential components towards the success of AI companionship in the long-run, so I'm dedicating my efforts towards building around that type of content in the most transparently direct and reasonable way possible. I'm eager to listen to the community. I'd love for folks to try it out for free and provide feedback! I'm proud of how far its come already (AI has certainly given me coding and content management superpowers), but I'm still improving many things so please let me know if you run into any issues. https://bit.ly/3I3gyT8 February 15, 2024 at 11:13PM

Show HN: IaC to Launch Palworld Servers on Fargate Spot https://bit.ly/3wfjyJI

Show HN: IaC to Launch Palworld Servers on Fargate Spot https://bit.ly/3wjXeP7 February 15, 2024 at 05:47AM

Show HN: NeuralFlow – Visualize the intermediate output of Mistral 7B https://bit.ly/3I2sl4m

Show HN: NeuralFlow – Visualize the intermediate output of Mistral 7B https://bit.ly/3UBuawx February 15, 2024 at 04:29AM