Show HN: notesbash – A notes management TUI written in bash
2 by klimperfix | 0 comments on Hacker News.
Hello everybody! We (lukeflo and me) have made a little notes management tui written in bash based on fff by Dylan Araps, which we've improved a lot to make it a notes management tui. It uses your editor of choice to edit and create notes, you can fuzzy search your notes and search them by tags you give them by writing them in the markdown yaml or emacs org header. All in all it turned out to be a surprisingly good project that we now want to refine after just releasing v1.0, so any feedback, suggestions or criticism is very welcome!
Saturday, 3 August 2024
Friday, 2 August 2024
New Show Hacker News story: latest news
Show HN: Valkey-Operator Kubernetes Operator for Valkey (Redis Fork)
4 by GrayTShirt | 0 comments on Hacker News.
Hi there, I've been working on a Kubernetes Operator to deploy and manage Valkey Cluster ( https://valkey.io ). I was unsatisfied with the existing options available, and felt that it didn't hurt to create a new one. I know a lot of folks are going to make the case that the Bitnami helm chart ( https://ift.tt/O9BQrAn... ) is good enough, or just use a cloud provider option is ideal. And for some use cases that makes sense, but for a lot of other use cases, especially testing, and multi-tenant SaaS offerings I think an Operator provides a reliable mechanism to satisfy the middle ground.
4 by GrayTShirt | 0 comments on Hacker News.
Hi there, I've been working on a Kubernetes Operator to deploy and manage Valkey Cluster ( https://valkey.io ). I was unsatisfied with the existing options available, and felt that it didn't hurt to create a new one. I know a lot of folks are going to make the case that the Bitnami helm chart ( https://ift.tt/O9BQrAn... ) is good enough, or just use a cloud provider option is ideal. And for some use cases that makes sense, but for a lot of other use cases, especially testing, and multi-tenant SaaS offerings I think an Operator provides a reliable mechanism to satisfy the middle ground.
New Show Hacker News story: latest news
Show HN: Ell – A command-line interface for LLMs written in Bash
13 by simonmysun | 0 comments on Hacker News.
Hi HN! I've created a CLI tool called "ell" that allows you to interact with LLMs directly from your terminal. Designed with the Unix philosophy in mind, ell is simple, modular, and extensible. You can easily pipe input and output to integrate with other tools. Its templates and hook-based plugins enable you to customize and extend its functionality to suit any needs. Check out the README for usage instructions and examples. I developed this tool because existing solutions often felt too heavy, with many dependencies, or they weren't friendly to piping and customization. I, on the contrary, wrote in almost pure Bash with least dependencies. Additionally, I found a lack of tools that could read past terminal output as context. Imagine encountering an issue in your terminal and being able to directly ask an LLM for help with a simple command—this is now possible with ell (see the demo video). Known limitations: - To maintain simplicity and efficiency, jq is used for JSON parsing. - Cannot avoid curl to sending HTTPS requests. If only there were SSL / TLS support in `/dev/tcp/`! - Perl is used to handle terminal escape sequences because regex in Bash does not support looking around. - Markdown syntax highlighting is not perfect due to the need for streaming output. It relies on a simple state machine instead of a full parser, which may produce falsy results. - Other known issues are listed in Github Issues. Please help add more! I welcome any criticism and suggestions, whether it's about the idea or code!
13 by simonmysun | 0 comments on Hacker News.
Hi HN! I've created a CLI tool called "ell" that allows you to interact with LLMs directly from your terminal. Designed with the Unix philosophy in mind, ell is simple, modular, and extensible. You can easily pipe input and output to integrate with other tools. Its templates and hook-based plugins enable you to customize and extend its functionality to suit any needs. Check out the README for usage instructions and examples. I developed this tool because existing solutions often felt too heavy, with many dependencies, or they weren't friendly to piping and customization. I, on the contrary, wrote in almost pure Bash with least dependencies. Additionally, I found a lack of tools that could read past terminal output as context. Imagine encountering an issue in your terminal and being able to directly ask an LLM for help with a simple command—this is now possible with ell (see the demo video). Known limitations: - To maintain simplicity and efficiency, jq is used for JSON parsing. - Cannot avoid curl to sending HTTPS requests. If only there were SSL / TLS support in `/dev/tcp/`! - Perl is used to handle terminal escape sequences because regex in Bash does not support looking around. - Markdown syntax highlighting is not perfect due to the need for streaming output. It relies on a simple state machine instead of a full parser, which may produce falsy results. - Other known issues are listed in Github Issues. Please help add more! I welcome any criticism and suggestions, whether it's about the idea or code!
New Show Hacker News story: latest news
Show HN: French National Assembly Simulator
3 by yassinsiouda | 0 comments on Hacker News.
Interactive simulator that lets you test your legislative skills by submitting a law for consideration by virtual deputies. Unfortunately in french only Stack: - Svelte - LangChainJs - llama3.1:8b - a graphics card (if it's slow, it's because there's only one)
3 by yassinsiouda | 0 comments on Hacker News.
Interactive simulator that lets you test your legislative skills by submitting a law for consideration by virtual deputies. Unfortunately in french only Stack: - Svelte - LangChainJs - llama3.1:8b - a graphics card (if it's slow, it's because there's only one)
New Show Hacker News story: latest news
Show HN: Three.js mirror cube that stays in sync
7 by mackopes | 1 comments on Hacker News.
Hello HN, Mirror Cube is a Three.js based 3D model of a Rubiks cube variant. It keeps unscrambling itself doing random moves, but the moves are (ehm.. should be) synced across all devices. I built this over one week and I thought you might find it interesting. Let me know what you think and if you find any bugs! Martin
7 by mackopes | 1 comments on Hacker News.
Hello HN, Mirror Cube is a Three.js based 3D model of a Rubiks cube variant. It keeps unscrambling itself doing random moves, but the moves are (ehm.. should be) synced across all devices. I built this over one week and I thought you might find it interesting. Let me know what you think and if you find any bugs! Martin
Thursday, 1 August 2024
New Show Hacker News story: latest news
Show HN: We open sourced Verbis, a local AI assistant with SaaS connectors
2 by alexmavr | 0 comments on Hacker News.
Hi everyone, Verbis is an AI assistant for MacOS that connects to applications such as Gmail, Gdrive, Outlook, Slack. I built Verbis to empower users with an LLM assistant without having to send any of their data to a third party, while also having the ease of installation of a MacOS application. I've seen a number of similar projects in this space: GPT4All, Swirl, PrivateGPT and more. However, there were one or more gaps in each of them: - You have to choose your model of choice, or even run it locally via ollama or llama.cpp - Pulling data from third party applications is often unsupported, or requires a complex service account token flow - Running the project requires having docker installed I wanted to make Verbis as easy as possible to install and set up, so we use our own choices of models for generation, and user-scoped OAuth tokens to allow integrations with third-party applications to be as simple as a consent screen. From a technical viewpoint, we have settled on Mistral 7B as the generation model, nomic-embed-text as the embedding model, Weaviate as the vector store, and ms-marco-MiniLM-L-12-v2 as the reranker model. On our next release we will be likely defaulting to LLama 3.1 8B for systems with more than 8G unified memory. Let me know if you have any feedback! I'd love to learn more about use cases where a fully private RAG pipeline makes sense
2 by alexmavr | 0 comments on Hacker News.
Hi everyone, Verbis is an AI assistant for MacOS that connects to applications such as Gmail, Gdrive, Outlook, Slack. I built Verbis to empower users with an LLM assistant without having to send any of their data to a third party, while also having the ease of installation of a MacOS application. I've seen a number of similar projects in this space: GPT4All, Swirl, PrivateGPT and more. However, there were one or more gaps in each of them: - You have to choose your model of choice, or even run it locally via ollama or llama.cpp - Pulling data from third party applications is often unsupported, or requires a complex service account token flow - Running the project requires having docker installed I wanted to make Verbis as easy as possible to install and set up, so we use our own choices of models for generation, and user-scoped OAuth tokens to allow integrations with third-party applications to be as simple as a consent screen. From a technical viewpoint, we have settled on Mistral 7B as the generation model, nomic-embed-text as the embedding model, Weaviate as the vector store, and ms-marco-MiniLM-L-12-v2 as the reranker model. On our next release we will be likely defaulting to LLama 3.1 8B for systems with more than 8G unified memory. Let me know if you have any feedback! I'd love to learn more about use cases where a fully private RAG pipeline makes sense
Subscribe to:
Posts (Atom)