491–500 of 909 posts
Earlier quoted context omitted.
I do it because 1) I am fascinated that I can and 2) at some point the online models will be enshitified — and I can then permanently fall back on my last good local version.
love the first and am sad you’re going to be right about the second
Funny how that works.
Model cards, for the people interested in the guts: https://cdn.openai.com/pdf/419b6906-9da6-406c-a19d-1bb078ac7... In my mind, I’m comparing the model architecture they describe to what the leading open-weights models (Deepseek, Qwen, GLM, Kimi) have been doing. Honestly, it just seems “ok” at a technical level: - both models use standard Grouped-Query Attention (64 query heads, 8 KV heads). The card talks about how…
I don't know how to ask this without being direct and dumb: Where do I get a layman's introduction to LLMs that could work me up to understanding every term and concept you just discussed? Either specific videos, or if nothing else, a reliable Youtube channel?
Sorry to ask what is possibly a dumb question, but is this effectively the whole kit and kaboodle, for free, downloadable without any guardrails? I often thought that a worrying vector was how well LLMs could answer downright terrifying questions very effectively. However the guardrails existed with the big online services to prevent those questions being asked. I guess they were always unleashed with other open sour…
These models still have guardrails. Even locally they won't tell you how to make bombs or write pornographic short stories.
The lede is being missed imo. gpt-oss:20b is a top ten model (on MMLU (right behind Gemini-2.5-Pro) and I just ran it locally on my Macbook Air M3 from last year. I've been experimenting with a lot of local models, both on my laptop and on my phone (Pixel 9 Pro), and I figured we'd be here in a year or two. But no, we're here today. A basically frontier model, running for the cost of electricity (free with a rounding…
How up to date are you on current open weights models? After playing around with it for a few hours I find it to be nowhere near as good as Qwen3-30B-A3B. The world knowledge is severely lacking in particular.
Just posted my initial impressions, took a couple of hours to write them up because there's a lot in this release! https://simonwillison.net/2025/Aug/5/gpt-oss/ TLDR: I think OpenAI may have taken the medal for best available open weight model back from the Chinese AI labs. Will be interesting to see if independent benchmarks resolve in that direction as well. The 20B model runs on my Mac laptop using less than 15GB…
I’m also very interested to know how well these models handle tool calling as I haven’t been able to make it work after playing with them for a few hours. Looks promising tho.
The lede is being missed imo. gpt-oss:20b is a top ten model (on MMLU (right behind Gemini-2.5-Pro) and I just ran it locally on my Macbook Air M3 from last year. I've been experimenting with a lot of local models, both on my laptop and on my phone (Pixel 9 Pro), and I figured we'd be here in a year or two. But no, we're here today. A basically frontier model, running for the cost of electricity (free with a rounding…
Seeing a 20B model competing with o3's performance is mind blowing like just a year ago, most of us would've called this impossible - not just the intelligence leap, but getting this level of capability in such a compact size. I think that the point that makes me more excited is that we can train trillion-parameter giants and distill them down to just billions without losing the magic. Imagine coding with Claude 4 Op…
But it does not actually compete with o3 performance. Not even close. As usual, the metrics are bullshit. You don't know how good the model actually is until you grill it yourself.
Orthogonal, but I just wanted to say how awesome Ollama is. It took 2 seconds to find the model and a minute to download and now I'm using it. Kudos to that team.
All the real heavy lifting is done by llama.cpp, and for the distribution, by HuggingFace.
Text only, when local multimodal became table stakes last year.
The community can always figure out hooking it up to other modalities. Native might be better, but no native multimodal model is very competitive yet, so better to take a competitive model and latch on vision/audio
Can this be done by a third party or would it have to be OpenAI?
Update: it seems to be completely useless for translation. It either refuses, outputs garbage, or changes the meaning completely for completely innocuous content. This already is a massive red flag.