Was playing around with this model - why does it return 2 or 3 responses when I ask it for one? I asked it for a json response and it generates 2 or 3 at a time. What's with this.
Microsoft Phi-3 Cookbook
51–59 of 59 posts
Re: Microsoft Phi-3 Cookbook
#52Looks like some of the docs are generated by an llm. I see pictures with typos and imagined terms, incomplete texts etc., I wonder to what extent we can trust rest of the docs. https://github.com/microsoft/Phi-3CookBook/blob/main/md/04.F...
Gool for specific tucks!
Good for specific tasks?
Re: Microsoft Phi-3 Cookbook
#53https://huggingface.co/collections/microsoft/phi-3-6626e15e9... , all of these models except Phi-3 mini are new.
Looking at the benchmarks, it seems like Phi-3 Small (7B) marginally beats out Llama3-8B on most tasks, which is pretty exciting!
It also seems to be comparable to gpt 3.5 turbo which I feel hard to believe. People have obviously found out a way around these benchmarks.
Re: Microsoft Phi-3 Cookbook
#54I installed Phi:medium last night on my Mac using Ollama and, subjectively, it looks good. I was surprised of the claim the it was better than mistral-8x7B.
I largely ignore benchmarks now, but on the other hand, while trying many models myself is easy for simple tests, really using a LLM for an application is a lot of work.
Re: Microsoft Phi-3 Cookbook
#55Slightly off topic: what’s the reasonably smallest LLM model i can use to do language processing and rewriting of a large library of word documents? For the purposes of querying information and regurgitating out summaries or detailed information? My use case is very simple: take 1000 word documents filled with two to three pages of information and pictures. And then output a set of requested information via prompting…
Sounds like a good RAG use-case unless all 1k documents need to be comprehended simultaneously.
Look at H2O.ai: https://github.com/h2oai/h2ogpt
Re: Microsoft Phi-3 Cookbook
#56Earlier quoted context omitted.
Looking at the benchmarks, it seems like Phi-3 Small (7B) marginally beats out Llama3-8B on most tasks, which is pretty exciting!
Have you found either model to be good enough to do anything interesting, reliably?
Llama3-8B is adequate for non-technical summarization or simple categorization.
Re: Microsoft Phi-3 Cookbook
#57Earlier quoted context omitted.
In single core performance... In multi-core from what I've seen X Plus is faster than base M3.
M3 was last year though, M4 has just been released, though iPad only for now
And here's some real world benchmarks:
M4 iPad Pro: https://browser.geekbench.com/v6/cpu/6036233 S: ~3747 M: ~14740
Snapdragon X Elite: https://browser.geekbench.com/search?utf8=%E2%9C%93&q=snapdr... S: ~2400 M: ~14000
Note this is an M4 iPad Pro. I would imagine an M4 Mac would spec a little better, due to less heat/power constraints.
tl;dr: M4 stomps on the Elite in single-core with about 36% more performance. The baseline M4 and X Elite are about even on multi-core. X Elite has marginally better NPU TOPS performance if that's something that matters to you.
Re: Microsoft Phi-3 Cookbook
#58Earlier quoted context omitted.
But not same performance. One of the reason QC not releasing anything but controlled benchmarks is most likely the subpar performance of Windows on ARM. This will be the biggest hurdle for QC Elite chips, competing with M-seried which is design in tandem with MacOS.
Speculation.
More like not totally baseless rumors.
https://www.tomshardware.com/tech-industry/qualcomm-faces-be...
Re: Microsoft Phi-3 Cookbook
#59Looks like some of the docs are generated by an llm. I see pictures with typos and imagined terms, incomplete texts etc., I wonder to what extent we can trust rest of the docs. https://github.com/microsoft/Phi-3CookBook/blob/main/md/04.F...
https://github.com/microsoft/Phi-3CookBook/commit/ba688b9a35...
Scroll down to the end and the removed text is totally suspect. I wouldn't be to surprised if all of this was generated by an LLM then anything strange was edited by a human. Another reason not to leave everything to the LLM.