Hacker News: chisleu

New comment by chisleu in "Great resources for self-hosting AI hardware"

chisleu — Thu, 09 Apr 2026 16:21:18 +0000

Many are investing in RTX6000pro hardware but hitting deployment walls. This guide covers everything from simple 4x builds to scaling up to 16 cards with PCIe switches.

Great resources for self-hosting AI hardware

chisleu — Thu, 09 Apr 2026 16:21:18 +0000

Article URL: https://github.com/voipmonitor/rtx6kpro

Comments URL: https://news.ycombinator.com/item?id=47705632

Points: 2

# Comments: 1

New comment by chisleu in "Google ADK-Go"

chisleu — Sun, 09 Nov 2025 22:23:46 +0000

An open-source, code-first Go toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control.

Google ADK-Go

chisleu — Sun, 09 Nov 2025 22:23:46 +0000

Article URL: https://github.com/google/adk-go

Comments URL: https://news.ycombinator.com/item?id=45869793

Points: 2

# Comments: 1

New comment by chisleu in "Hacking India's largest automaker: Tata Motors"

chisleu — Sat, 01 Nov 2025 14:10:13 +0000

Total tangent, but I got to ride in some of these on a recent trip to India and I was really impressed with the build quality and utilitarian usefulness of the design.

New comment by chisleu in "Qwen3-Omni: Native Omni AI model for text, image and video"

chisleu — Mon, 22 Sep 2025 18:45:15 +0000

Here is the demo video on it. The video w/ sound input -> sound output while doing translation from the video to another language was the most impressive display I've seen yet.

https://www.youtube.com/watch?v=_zdOrPju4_g

New comment by chisleu in "iPhone Air"

chisleu — Wed, 10 Sep 2025 18:01:00 +0000

Because of the prompt processing speed, small models like Qwen 3 coder 30b a3b are the sweet spot for mac platform right now. Which means a 32 or 64GB mac is all you need to use Cline or your favorite agent locally.

New comment by chisleu in "GLM 4.5 with Claude Code"

chisleu — Sat, 06 Sep 2025 02:07:27 +0000

I've been using GLM 4.5 and GLM 4.5 Air for a while now. The Air model is light enough to run on a macbook pro and is useful for Cline. I can run the full GLM model on my Mac Studio, but the TPS is so slow that it's only useful for chatting. So I hooked up with openrouter to try but didn't have the same success. Any of the open weight models I try with open router give sub standard results. I get better results from Qwen 3 coder 30b a3b locally than I get from Qwen 3 Coder 480b through open router.

I'm really concerned that some of the providers are using quantized versions of the models so they can run more models per card and larger batches of inference.

New comment by chisleu in "Updates to Consumer Terms and Privacy Policy"

chisleu — Fri, 29 Aug 2025 13:11:47 +0000

This is going to improve the quality of LLM responses for users. I'm for this.

New comment by chisleu in "AWS in 2025: Stuff you think you know that's now wrong"

chisleu — Sun, 24 Aug 2025 19:49:06 +0000

> but eventually they did it

They can do this with manual partitioning indeed. I've done it before, but it's not ideal because the auto partitioner will scale beyond almost anything AWS will give you with manual partitioning unless you have 24/7 workloads.

> you can be throttled for more than a day before it kicks in

I expect that this would depends on your use case. If you are dropping content you need to scale out to tons of readers, that is absolutely the case. If you are dropping tons of content with well distributed reads, then the auto partitioner is The Way.

New comment by chisleu in "AWS in 2025: Stuff you think you know that's now wrong"

chisleu — Sun, 24 Aug 2025 19:43:52 +0000

and indeed the bucket is not separate from the object key. the API separates it logically "for humans" but it's all one big string

New comment by chisleu in "AWS in 2025: Stuff you think you know that's now wrong"

chisleu — Wed, 20 Aug 2025 17:03:57 +0000

> You don’t have to randomize the first part of your object keys to ensure they get spread around and avoid hotspots.

As of when? According to internal support, this is still required as of 1.5 years ago.

New comment by chisleu in "What could have been"

chisleu — Mon, 18 Aug 2025 22:55:14 +0000

/agree

We are in the infancy of LLM technology.

New comment by chisleu in "GPT-5 vs. Sonnet: Complex Agentic Coding"

chisleu — Fri, 08 Aug 2025 16:41:07 +0000

How was he doing "complex agentic coding" when the APIs have such extreme context and throughput limitations?

New comment by chisleu in "Leonardo Chiariglione – Co-founder of MPEG"

chisleu — Thu, 07 Aug 2025 12:32:00 +0000

holy shit it does. The scene with him inventing the new compression algorithm basically foreshadowed the gooning to follow local LLM availability.

New comment by chisleu in "Claude Opus 4.1"

chisleu — Wed, 06 Aug 2025 06:27:03 +0000

I use opus or gemini 2.5 pro for plan mode and sonnet for act mode in Cline. https://cline.bot

It's my experience that Opus is better at solving architectural challenges where sonnet struggles.

New comment by chisleu in "Kimi-K2 Tech Report [pdf]"

chisleu — Wed, 23 Jul 2025 23:24:50 +0000

It looks like qwen3-coder is going to steal K2's thunder in terms of agentic coding use.

New comment by chisleu in "Qwen3-Coder: Agentic coding in the world"

chisleu — Wed, 23 Jul 2025 17:12:59 +0000

It's 480B params, not 480GB. The 4 bit version of this is 270GB. I believe it's trained at bf16, so you need over a TB of memory to operate the model at bf16. No one should be trying to replace claude with a quantized 8 bit or 4 bit model. It's simply not possible. Also, this model isn't going to be as versed as Claude at certain libraries and languages. I have something written entirely my claude which uses the Fyne library extensively in golang for UI. Claude knows it inside and out as it's all vibe coded, but the 4 bit Qwen3 coder just hallucinated functions and parameters that don't exist because it wasn't willing to admit it didn't know what it was doing. Definitely don't judge a model by it's quant is all I'm saying.

New comment by chisleu in "Qwen3-Coder: Agentic coding in the world"

chisleu — Wed, 23 Jul 2025 15:38:35 +0000

A Mac Studio 512GB can run it in 4bit quantization. I'm excited to see unsloth dynamic quants for this today.

New comment by chisleu in "Qwen3-Coder: Agentic coding in the world"

chisleu — Wed, 23 Jul 2025 13:15:57 +0000

I tried using the "fp8" model through hyperbolic but I question if it was even that model. It was basically useless through hyperbolic.

I downloaded the 4bit quant to my mac studio 512GB. 7-8 minutes until first tokens with a big Cline prompt for it to chew on. Performance is exceptional. It nailed all the tool calls, loaded my memory bank, and reasoned about a golang code base well enough to write a blog post on the topic: https://convergence.ninja/post/blogs/000016-ForeverFantasyFr...

Writing blog posts is one of the tests I use for these models. It is a very involved process including a Q&A phase, drafting phase, approval, and deployment. The filenames follow a certain pattern. The file has to be uploaded to s3 in a certain location to trigger the deployment. It's a complex custom task that I automated.

Even the 4bit model was capable of this, but was incapable of actually working on my code, prefering to halucinate methods that would be convenient rather than admitting it didn't know what it was doing. This is the 4 bit "lobotomized" model though. I'm excited to see how it performs at full power.