omneity

21 ms·

301.
by omneity 2y ago
This is a pretty wild claim on the Eclipse foundation. Do you have a source to substantiate it?
302.
by omneity 2y ago
Take out the oxygen and the conditions are almost the same as a lunar lava tube. Trailblazing work truly.
303.
by omneity 2y ago
Fixed, thanks!
304.
An African Superstate: The East African Federation (2022)(karlsnotes.com)
2 points by omneity 2y ago|2 comments
305.
by omneity 2y ago
I like that the usb connector is vertical. I always found it more challenging than necessary to plug and unplug the cable when using a breadboard as its hard to get a grip on the cable and forces me to put the board on the breadboard's
306.
by omneity 2y ago
So am I correct in understanding what they really mean is 37 full forward passes per second? In which case, if the model weights are fitting in the VRAM and are already loaded, why does the bandwidth impact the rate of tok/s?
307.
by omneity 2y ago
I'm surprised at the simplicity of the formula in the paragraph below. Could someone explain the relationship between model size, memory bandwidth and token/s as they calculated here? > Taking LLaMA 3 70B as an example, in floa
308.
by omneity 2y ago
I sincerely hope there's no moonseed in it. https://en.wikipedia.org/wiki/Moonseed_(novel)
309.
by omneity 2y ago
One could argue flipping doesn't change X, Y nor Z coordinates :)
310.
by omneity 2y ago
Oh I guess it didn't load for me then. I was a bit surprised at the absence of it. Consider my peeve unpeeved :)
311.
by omneity 2y ago
Making the terminal more lively is a laudable goal. My only peeve is that a video or gif of those effects would have been most perfect.
312.
by omneity 2y ago
> ReactOS must take the record for software with the longest gestation in history. May I introduce you to GNU Hurd? First release happened 34 years ago. https://www.gnu.org/software/hurd/
313.
by omneity 2y ago
It actually is true. Running an OpenAI compatible server using llama.cpp is a one-liner. Check out the docker option if you don’t want to build/install llama.cpp. https://github.com/ggerganov/llama.cpp/tree&#x
314.
by omneity 2y ago
The post starts with a fascinating premise, but then falls short as it does not define search in the context of LLMs, nor does it explain how “Pfizer can access GPT-8 capabilities today with more inference compute”. I found it hard to fol
315.
by omneity 2y ago
The reporting on this article is just bad. The issue is not Qualcomm taking over Nuvia's business and license. According to Arm the Nuvia license is contractually tied to server/datacenter usage, while Qualcomm's acquisitio
316.
by omneity 2y ago
I read it as a double entendre joke about B.C. Whether it's appropriate on HN is another matter.
317.
by omneity 2y ago
Your comment has an unnecessary and overly negative tone to it that doesn't do this tech justice. These approaches are totally valid and can get you great results. An LLM is just a component in a pipeline. I deployed many of these in p
318.
by omneity 2y ago
Natural language search and answer generation _are_ completely separate. Search is often (but not exclusively) performed using cosine similarity over semantic vectors. Such vectors are produced using embedding models, which represent the me
319.
by omneity 2y ago
Do you know where does this cost lie? Say for an organization with no prior spending on "legacy" electronics based hardware. Is it the sunken cost in existing hardware, or developing the rest of the "value chain" to make
320.
by omneity 2y ago
Marketing is hardly something that can be done via an "add-on" or some mechanical approach. It's more of a mindset, getting yourself and what you're working on known is something you need to do throughout the entire proc
321.
by omneity 2y ago
Except the cost is only going down over time
322.
by omneity 2y ago
My understanding is that Stability was renting GPUs from AWS rather than buying them[0]. So they have no capacity per se. [0]: https://www.theregister.com/2024/04/03/stability_ai_bills/
323.
by omneity 2y ago
One way you could track updates to their ToS is via https://monitoro.co Disclosure: I work on it.
324.
by omneity 2y ago
try the latest Ableton Push, or NI Maschine+
325.
by omneity 2y ago
If you expect an OpenAI compatible API to use function calls, I don't think Ollama supports it yet (to be confirmed). However you can do it yourself using the appropriate tokens for the model. I know that Llama3, various Mistrals and C
326.
by omneity 2y ago
Awesome work! I tried it on a couple of pretty confusing examples and it worked out great. However you might want to build with production mode enabled, since your current build actually tries to connect to a Vite dev server on localhost:30
327.
by omneity 2y ago
This is trivially achievable with function calling, assuming the model you use supports this (which most models do at this point). Define a function `reportFactual(isFactual: boolean)` and you will get standardized, machine-readable answers
328.
by omneity 3y ago
The team has commented that they skipped a step in the microsoft release process (toxicity check). This is a nothingburger. https://twitter.com/WizardLM_AI/status/1780101465950105775
329.
by omneity 3y ago
> find an enterprise that doesn't shy away from big budgets How do you identify such enterprises?
330.
by omneity 3y ago
It seems to me Meta is simply following the old startup tenet: build something people want. They were well positioned to pull this off with their AI/ML research resources and tradition of openness, and happen to have one of the biggest

More ›