sosuke
My machine has an RTX 4060 Ti 16 GB on PCI 3.0, Threadripper 1950X with AVX2, and 128 GB quad-channel DDR4-2133
I stopped after hitting a successful 84-minute autonomous run in OMP. Just 4 bugs in the result found by a frontier-ish model. Hit 770.83 tokens/sec on prefill and 15.85 tokens/sec at 8k context and 11 tokens/sec over the full run at 128k context.
The link is to the main findings and attempts at improving performance. In the end the Laguna architecture didn't lend itself to the same kind of dissection as Qwen3.8 and I've taken it as far as I'm willing to go. It's all dropped into the performance and research folders.
I love competition in the open weight world and Strata has been such a boon to my own local development I was greedy for more! Hope this helps someone, or you find it interesting at least.
The only time I remember the newest model making an obvious regression was when the first rolled out MoE. Super speed update but each request had less intelligence at hand. We’re way past that now
To Read a Book, to Collect a Book
You Can Start Building LLM Skills Before You Know the Whole Shape
Models Have Blind Spots: Debugging Unfamiliar Code with a Multi-LLM Loop
Bone Keeper AI Assisted Feature Film
Edit: remember not to down vote ideas you disagree with. I think it was only down vote things that lower the discourse
The Workflows I Created Using LLMs Became My Ecosystem
Show HN: Agent Historic Philosophical Persona Routing and Prompts
https://github.com/barretts/AgentHistoric
I've been building this for a while. The core of it is I've assigned different tasks in a software engineering job to different philosophers. The prompts are written toward their thinking by "them" but the keywords are generic. I absolutely love using this prompt system. I share it every chance I get. I wrote a whitepaper about it. Maybe ya'll would be interested in it. I find my experience is...
The working memory it holds is still extremely small compared to what we would need for regular open ended tasks.
Yes there are outliers and I'm not being specific enough but I can't type that much right now.
Feature request: Exclude books already in shelf. This is harder I'm sure. I've got 1146 books in my Read shelf.
But if you don't already know the question your asking is not at all something I could distill down into a sentence or to that would make sense to a lay-person. Even then I know I couldn't distill it at all sorry.
Edit: I found this link I referenced above on quantized models by bartowski on huggingface https://huggingface.co/bartowski/Qwen2.5-Coder-14B-GGUF#whic...
Location: Austin, TX
Remote: Preferred remote but all things are negotiable
Willing to relocate: Not really
Technologies: JavaScript, TypeScript, Node, GIT, NPM, React, Web Components, Webpack, Build configuration, Jest, Mocha, ESLint, CSS, SASS, Storybook, Monorepo configuration, TestCafe, Unit testing, End-to-end testing, Integration testing, Backbone, Marionette, jQuery, Bootstrap, Google MUI React UI Components, MySQL, Postgres SQL, SQL Server, C#, Python, Ruby, Rake, Docker, TeamCity, Jenkins, Amazon Web Services, Terraform, PowerShell, Visual Studio, EC2, S3, DynamoDB, Lambda, API Gateway, Cloud Development Kit, EC2 Image Builder, HTML, Internationalization i18n, Localization l10n, Visual Diff testing, Bash scripting, Objective-C, Android Java, Cordova, KendoUI, Flash, Flex, ActionScript
Résumé/CV: https://barrettsonntag.com/Sonntag_Barrett_resume.pdf
Email: [email protected]
Is it better that I list everything or just a few things? I'm always conflicted. What if I said TypeScript and React but left out JavaScript? That could mean I never worked with JavaScript right? Which sounds crazy but code camps produce React developers that have no experience outside that framework.You have a dozen people who would give you emotional support but you don't ask for any because you aren't sure you could give emotional support in return.
You have a dozen people you think might give you emotional support but you don't want to burden them with your problems.
You have a dozen people that would give you emotional support, maybe once, but you are too afraid to burn the single emotional support coupon until its more important.
Location: Austin, TX
Remote: Preferred remote but all things are negotiable
Willing to relocate: Not really
Technologies: JavaScript, TypeScript, Node, GIT, NPM, React, Web Components, Webpack, Build configuration, Jest, Mocha, ESLint, CSS, SASS, Storybook, Monorepo configuration, TestCafe, Unit testing, End-to-end testing, Integration testing, Backbone, Marionette, jQuery, Bootstrap, Google MUI React UI Components, MySQL, Postgres SQL, SQL Server, C#, Python, Ruby, Rake, Docker, TeamCity, Jenkins, Amazon Web Services, Terraform, PowerShell, Visual Studio, EC2, S3, DynamoDB, Lambda, API Gateway, Cloud Development Kit, EC2 Image Builder, HTML, Internationalization i18n, Localization l10n, Web Accessibility Initiative Accessible Rich Internet Applications Suite WAI-ARIA, Visual Diff testing, Bash scripting, Objective-C, Android Java, Cordova, KendoUI, Flash, Flex, ActionScript
Résumé/CV: https://barrettsonntag.com/Sonntag_Barrett_resume.docx
Email: [email protected]I would say that math is unambiguous.
Measurements are becoming more unambiguous. The accepted measurement of a kilogram equalling s Avogadro’s constant. Something like the number of silicon atoms in a 93mm sphere?
But the definitions of best and reasonable are only accepted based off precedent and we’ve seen in recent history precedent in law isn’t as reliable or defined as we may have thought.
Is there any language or any level of language that can remain defined across time and culture?
So long as legalese requires interpretation to determine intent and outcome it can be expected be incomprehensible.
Location: Austin, TX
Remote: Preferred remote but all things are negotiable
Willing to relocate: Not really
Technologies: JavaScript, TypeScript, Node, GIT, NPM, React, Web Components, Webpack, Build configuration, Jest, Mocha, ESLint, CSS, SASS, Storybook, Monorepo configuration, TestCafe, Unit testing, End-to-end testing, Integration testing, Backbone, Marionette, jQuery, Bootstrap, Google MUI React UI Components, MySQL, Postgres SQL, SQL Server, C#, Python, Ruby, Rake, Docker, TeamCity, Jenkins, Amazon Web Services, Terraform, PowerShell, Visual Studio, EC2, S3, DynamoDB, Lambda, API Gateway, Cloud Development Kit, EC2 Image Builder, HTML, Internationalization i18n, Localization l10n, Web Accessibility Initiative Accessible Rich Internet Applications Suite WAI-ARIA, Visual Diff testing, Bash scripting, Objective-C, Android Java, Cordova, KendoUI, Flash, Flex, ActionScript
Résumé/CV: https://barrettsonntag.com/Sonntag_Barrett_resume.docx
Email: [email protected]
---------My wide breath of skills work on mobile and web development projects stems from my passion as a creator. It never is a question if something is possible but only if we’ve got the will to make it reality. Though I started this career on my own being self-taught I was drawn to leading and mentoring other developers to reach their own potential. There is no greater satisfaction than celebrating a successful launch as a team and to see us grow together. I love working with executives, clients and project managers to architect end to end software solutions that will exceed expectations. Whether the project is building in-house tooling or customer facing applications it is crucial to understand how the user will interact with it and ensure the technology stack will effectively deliver the best possible experience.
Over the last 20 years I’ve built, tested and released products using the languages, libraries or technologies listed above. They aren’t listed with any sorting in mind.
(unabridged resume) https://barrettsonntag.com/Sonntag_Barrett_resume_extended.d...
It would be nice to dream of it happening sometime soon.