Hacker News: mmis1000

New comment by mmis1000 in "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"

mmis1000 — Tue, 29 Sep 2026 17:39:58 +0000

For my personal experience, antropic model have better user experience except for 4.7 and 4.8 though. 4.7 and 4.8 feels like expensive downgrade of 4.6 to me (I didn't know why these two should even exist)

However it's less willing to obey your instruction so it's less usable for general runtine flows.

Transformers now runs llama.cpp quants

mmis1000 — Wed, 23 Sep 2026 06:18:18 +0000

Article URL: https://huggingface.co/blog/transformers-llama-cpp-quants

Comments URL: https://news.ycombinator.com/item?id=49812339

Points: 4

# Comments: 0

New comment by mmis1000 in "OpenAI is well positioned to fast-follow Jev"

mmis1000 — Tue, 22 Sep 2026 20:34:09 +0000

Although the underlying model is unknown. If it expose input token count, the tokenizer may be probable though. Most tokenizer segemnts wildly different in CJK inputs. It can probably be used to fingerprint the tokenizer based on token count if it is using existing tokenizer.

New comment by mmis1000 in "OpenAI is well positioned to fast-follow Jev"

mmis1000 — Tue, 22 Sep 2026 20:19:29 +0000

The way jev actually internally work could be interesting though. I believe most llm are only tuned to return the first or second logits(or a few more) correctly as that is what the sampler would choose anyway. Do they alter existing model for better behavior across all options? Or they distilled one to have the proper behavior? We can only guess without the actual implementation.

New comment by mmis1000 in "OpenAI is well positioned to fast-follow Jev"

mmis1000 — Tue, 22 Sep 2026 19:53:01 +0000

You don't even bother text after the [a] at first place in this case

Your question is something like

anwser only a,b,c,d for following question a. b. c. d....

the model output possibility of next character a: 0.8 b: 0.7 c: 0.3 f: 0.2 d: 0.1

If the list contains option you did not provide. The model is confused anyway, it don't matter if you use grammer to filter out the bad option or not, the answer is screwed already.

New comment by mmis1000 in "OpenAI is well positioned to fast-follow Jev"

mmis1000 — Tue, 22 Sep 2026 19:19:36 +0000

If you really know what grammers did, grammer is a filter to mask out option llm provided but you don't like.

It does not change potential distribution in any means. It DROPS part of answer model returned directly.

The text generation model go wild because model relies on previous section it answered to continue later section. And because now it contain item model have no idea, it is completely screwed.

In the case you only require model to answer one of a,b,c,d and don't care about later segment at all. It don't really matter.

New comment by mmis1000 in "OpenAI is well positioned to fast-follow Jev"

mmis1000 — Tue, 22 Sep 2026 19:07:50 +0000

Fixed guard today is not very fixed. For ex, the safeguard qwen released is a full 4b llm model. It has no different to normal llm model arch except tuned for this specific purpose,

New comment by mmis1000 in "Windows 11's built-in Weather app wastes more than 1 GB of RAM"

mmis1000 — Sun, 09 Aug 2026 20:06:52 +0000

i mean webKit never handle viewport well when compare to firefox or chrome. it's willing to materialize a 10000*10000 texture when firefox and chrome will think that is too dumb and only load the visible part. It's not a good target if you want to compare windows and mac software quality.

New comment by mmis1000 in "Google will expand age checks on Android worldwide till the end of the year"

mmis1000 — Thu, 30 Jul 2026 14:41:03 +0000

I doubt the purpose is even age verification. They already can know I have a visa card which is credit card that can be only applied by an adult by looking at the BIN. Why they need my id card if the purpose is actually age verification?

New comment by mmis1000 in "HomeLab #1: MikroTik as a Home Router"

mmis1000 — Mon, 20 Jul 2026 05:11:51 +0000

Plus the routeros supports exporting all settings directly as text file. so you can really export the whole system, send it to llm and ask it to edit. install the new rule.

in worst case you just install the old rule back if it fuck up, no really damage can happen.

New comment by mmis1000 in "Apple's new SpeechAnalyzer API, benchmarked against Whisper and its predecessor"

mmis1000 — Mon, 13 Jul 2026 18:15:18 +0000

Every single asr model I tested so far did not support timestamps properly though. Some use external aligner to create timestamp, but the accuracy is still much inferior than whipser in case the audio is noisy.

New comment by mmis1000 in "Since Chromium 148, Math.tanh is now fingerprintable to link underlying OS"

mmis1000 — Sun, 12 Jul 2026 23:20:30 +0000

Chrome is the only browser that preserve unused bit in value NaN through non JITed mode as far as I remember. And that bit become 0 when code get JITed.

New comment by mmis1000 in "Since Chromium 148, Math.tanh is now fingerprintable to link underlying OS"

mmis1000 — Sun, 12 Jul 2026 23:14:39 +0000

browser report where your browser window (not position in tab) is even page itself never need to use it.

https://developer.mozilla.org/zh-CN/docs/Web/API/Window/scre...

New comment by mmis1000 in "Claude Fable 5"

mmis1000 — Tue, 09 Jun 2026 19:18:54 +0000

If calling somebody with phone is still 'dialing' someone even there is nothing round on smart phone. Then why not?

New comment by mmis1000 in "Does anybody like React?"

mmis1000 — Tue, 26 May 2026 09:06:45 +0000

Vue solve single source of truth much better than react though. The shape of derived data and source data (or even external data) are the same and interchangeable. So you can write an api that works with both without handling implementation quirks at all.

The words borrowed from someone else:

The react is more about view.

The vue is more about reactivity

New comment by mmis1000 in "Agents need control flow, not more prompts"

mmis1000 — Thu, 07 May 2026 21:02:30 +0000

> This started breaking down after ~30 files.

Codex's short context and todolist system combined somehow helps here though. Because of the frequent compact. The model was forced to recheck what todo list item has not done yet and what workflow skill it has to use. I used to left it for multi hour to do a big clean up and it finished without obvious issues.

New comment by mmis1000 in "Taking on CUDA with ROCm: 'One Step After Another'"

mmis1000 — Mon, 13 Apr 2026 08:01:17 +0000

Compatibility so foundation packages like torch onnx-runtime can run on AMD GPU without massive change in architecture. It's the biggest reason for those stuff that "only works on nvidia gpu". It's not faster if vulkan alternative exists, but at least it runs.

New comment by mmis1000 in "Taking on CUDA with ROCm: 'One Step After Another'"

mmis1000 — Mon, 13 Apr 2026 07:07:52 +0000

At least they finally do something this time. Now torch and whatever transformer stuff runs normally on windows/linux as long as you installed correct wheel from amd's own repository.

It's a huge step though.

New comment by mmis1000 in "Unsloth Studio"

mmis1000 — Wed, 18 Mar 2026 02:56:44 +0000

I think the website should probably mention those installation preset in unsloth pyproject.toml though. The website instruct you to install dependencies separately. But it turns out there are dedicated preset that install specific rocm/cuda/xpu version in the project.

New comment by mmis1000 in "Unsloth Studio"

mmis1000 — Wed, 18 Mar 2026 02:31:54 +0000

Uv helps you up though. Use a pyproject.toml and uv sync. Everything will be put into the venv only, nothing spread across the whole system.

The pyproject.toml can even handles build env for you, so you no longer need a setup.sh that installs 10 tool in specific order with specific flag to produce working environment. A single uv sync, and the job is done.

Plus the result is reproducible, so if this time uv sync work, then it also work next time.

Highly recommend if you are still on pip.

Note: Take a example that I used to install unsloth with rocm setup that based on unreleased git version dependencies and graphic card specific build flag, all of them can be handled with one command 'uv sync'. This will require a big pile of shell script if doing another way. https://github.com/unslothai/unsloth/issues/4280#issuecommen...