HN user

shallow-mind

341 karma
Posts37
Comments5
View on HN
qwen.ai 27d ago

Qwen-AgentWorld: Language World Models for General Agents

shallow-mind
2pts0
github.com 1mo ago

Doclang-Project/Doclang

shallow-mind
2pts1
research.ibm.com 1mo ago

Granite Libraries and Project Granite Switch – IBM Research

shallow-mind
3pts0
developers.googleblog.com 2mo ago

Genkit Middleware: Intercept, extend, and harden your agentic apps Blog

shallow-mind
2pts0
huggingface.co 2mo ago

Granite 4.1 LLMs: How They're Built

shallow-mind
1pts0
mcosovic.github.io 7mo ago

JuliaGrid: Julia Power System Simulation

shallow-mind
4pts0
www.ibm.com 7mo ago

IBM Bob: Shift left for resilient AI with security-first principles

shallow-mind
1pts0
lennyforlibraries.org 8mo ago

Lenny for Libraries

shallow-mind
1pts0
jules.google 9mo ago

Practical Examples and Scripting – Jules

shallow-mind
2pts0
huggingface.co 9mo ago

Paper Page – Regression Language Models for Code

shallow-mind
1pts0
techcommunity.microsoft.com 11mo ago

Azure Linux with OS Guard: Secure, Immutable, and Open-Source Container Host

shallow-mind
1pts0
github.com 1y ago

MCP Toolbox for Databases – open-source MCP server for databases

shallow-mind
3pts0
almalinux.org 1y ago

AlmaLinux OS – Forever-Free Enterprise-Grade Operating System

shallow-mind
4pts0
www.phoronix.com 1y ago

New Set of Serious Security Vulnerabilities Exposed for Intel and Arm CPUs

shallow-mind
4pts0
www.cursor.com 1y ago

Cursor for Students – Cursor – The AI Code Editor

shallow-mind
1pts0
research.ibm.com 1y ago

Bamba: An open-source LLM that crosses a transformer with an SSM

shallow-mind
207pts69
www.wanaku.ai 1y ago

Wanaku: The First Open-Source MCP Router – Wanaku.ai

shallow-mind
2pts0
red-hat-ai-innovation-team.github.io 1y ago

Red Hat AI Innovation Team – R1-Like Reasoning

shallow-mind
3pts0
github.com 1y ago

Oumi-AI/oumi: Everything you need to build foundation models, end-to-end

shallow-mind
14pts0
www.ibm.com 1y ago

IBM Granite 3.1: powerful performance, longer context and more

shallow-mind
2pts0
www.redhat.com 1y ago

RHEL AI 1.3 adds the Granite 3.0 8B model

shallow-mind
3pts0
lfaidata.foundation 1y ago

LF AI and Data Announces VLLM as Its Latest Incubation Project – LFAI and Data

shallow-mind
2pts0
samba.plus 1y ago

SerNet secures funding for Samba project from Sovereign Tech Fund

shallow-mind
2pts0
www.undeadly.org 2y ago

RIP Dhclient(8)

shallow-mind
2pts0
www.redhat.com 2y ago

Getting Hands-On with Red Hat OpenShift Virtualization

shallow-mind
3pts0
blog.podman.io 2y ago

Podman 5.0 has been released

shallow-mind
77pts31
lore.kernel.org 2y ago

Hans Reiser on ReiserFS Deprecation

shallow-mind
4pts0
blog.rust-embedded.org 2y ago

Embedded-hal v1.0 now released

shallow-mind
3pts0
github.com 2y ago

FreeRDP 3.0.0-C0

shallow-mind
1pts0
gitlab.freedesktop.org 2y ago

PipeWire 1.0.0

shallow-mind
218pts103

I actually works already, just add new agent definition:

--- "Local QWEN": { "command": "/usr/local/bin/qwen", "args": [ "--experimental-acp" ], "env": {} }, ---

The awful performance you can be stuck with when you can only use emulators and v8 is going to come back to haunt some of you, I believe. It is indeed easy to sell something like this, as it looks easy to work with, the same way that no-code looks easy. But inevitably even native performance is not nearly enough and you need to employ algorithms that use 256- and 512-bit vector operations. At that point, we are talking about running natively on the CPU, and for that you need something that isn't just a web app.

I remember I recently saw a lecture by an industry professional who took a Python algorithm and improved it 100 000x, in parts. That is, in stages he explained how to utilize the hardware on that particular machine, to improve the algorithm so much that it was hard to believe. Anyone have a link? I looked and I could not find it. It was a recent thing that I believe I saw on this site.

When it comes to configuring web services: It is extremely important to configure them in such a way that you can fully or partially cache content, reducing costs. I get it, if you are a startup you need to move fast, but with the way things are going these days, you should have someone on your team who can at the very least deal with the basics of caching.

You also cannot pre-initialize v8 as far as I know, so you inevitably end up putting it inside an emulator anyway, and guess what happens to the performance then? I could be wrong.

The awful performance you can be stuck with when you can only use emulators and v8 is going to come back to haunt some of you, I believe. It is indeed easy to sell something like this, as it looks easy to work with, the same way that no-code looks easy. But inevitably even native performance is not nearly enough and you need to employ algorithms that use 256- and 512-bit vector operations. At that point, we are talking about running natively on the CPU, and for that you need something that isn't just a web app.

I remember I recently saw a lecture by an industry professional who took a Python algorithm and improved it 100 000x, in parts. That is, in stages he explained how to utilize the hardware on that particular machine, to improve the algorithm so much that it was hard to believe. Anyone have a link? I looked and I could not find it. It was a recent thing that I believe I saw on this site.

When it comes to configuring web services: It is extremely important to configure them in such a way that you can fully or partially cache content, reducing costs. I get it, if you are a startup you need to move fast, but with the way things are going these days, you should have someone on your team who can at the very least deal with the basics of caching.

You also cannot pre-initialize v8 as far as I know, so you inevitably end up putting it inside an emulator anyway, and guess what happens to the performance then? I could be wrong.

The awful performance you can be stuck with when you can only use emulators and v8 is going to come back to haunt some of you, I believe. It is indeed easy to sell something like this, as it looks easy to work with, the same way that no-code looks easy. But inevitably even native performance is not nearly enough and you need to employ algorithms that use 256- and 512-bit vector operations. At that point, we are talking about running natively on the CPU, and for that you need something that isn't just a web app.

I remember I recently saw a lecture by an industry professional who took a Python algorithm and improved it 100 000x, in parts. That is, in stages he explained how to utilize the hardware on that particular machine, to improve the algorithm so much that it was hard to believe. Anyone have a link? I looked and I could not find it. It was a recent thing that I believe I saw on this site.

When it comes to configuring web services: It is extremely important to configure them in such a way that you can fully or partially cache content, reducing costs. I get it, if you are a startup you need to move fast, but with the way things are going these days, you should have someone on your team who can at the very least deal with the basics of caching.

You also cannot pre-initialize v8 as far as I know, so you inevitably end up putting it inside an emulator anyway, and guess what happens to the performance then? I could be wrong.