cheptsov

Double that. The entire article reads to me as handpicked and manipulative.

It’s called handpicking

The author claims that everyone refers to Telegram as an encrypted messenger, but he only provides a single example to support that. I quickly checked Google News and couldn't find any media on the first page that did the same. It feels like a manipulation.

UPDATE: anyone who downvote, I invite to check for themselves.

Just a few known media:

1. https://www.aljazeera.com/amp/news/2024/8/25/telegram-messag...

2. https://www.washingtonpost.com/technology/2024/08/25/durov-t...

3. https://www.businessinsider.com/telegram-ceo-pavel-durov-arr...

4. https://www.theguardian.com/media/article/2024/aug/24/telegr...

However, indeed, I‘ve seen a few media that call it encrypted. This include France24, POLITICO, and The Times.

That’s on our list! We’re actually planning to expand our examples page with a lot more frameworks! Contributors are very welcome to our repo: https://github.com/dstackai/dstack/tree/master/examples/

Oops, thanks for sharing this! Gonna update it shortly.

Hi, I’m one of the contributors to dstack! Yes, you’re absolutely right—the title is a bit off. It should be "K8S for AI Teams Only." And yes, there is a lot of focus on LLMs specifically. If you have any other feedback, please share it!

I’m not a fan of Musk but it is so funny to see that many haters here.

Profs?

The one you disagree with but why do you think all people think the same way as you?

I entirely agree. I’m quite sure Musk has a very strong stance on ethics, but it would be great to hear about it more clearly, and ideally not just through words, but through actual actions.

Ahhhh

huggingface_hub.utils._errors.HfHubHTTPError: 502 Server Error: Bad Gateway for url: https://huggingface.co/api/models/meta-llama/Meta-Llama-3.1-...

Error: DownloadError

They have shared the link to GitHub [1], HuggingFace repo [2], and the paper [3]:

1. https://github.com/aiola-lab/whisper-medusa

2. https://huggingface.co/aiola/whisper-medusa-v1

3. https://paperswithcode.com/method/multi-head-attention

I wonder if the project makes it possible to run macOS inside a container on a macOS host?

If this receives support for an Apple GPU, it will be incredibly significant!

I wonder why AWQ is not mentioned. It’s pretty popular and I always was curious how it is different from GPTQ.

Sounds like an extremely complex technical problem. I also suggest to look at the use cases when this is needed. One of the problems is that loading weights into the GPU will be so slow that it will be really hard to share the GPU between different processes - causing long time to offload and load. Would love to learn more about what you do.

JetBrains PyCharm has this out of the box. VSCode has a nice extension [1]

1. https://marketplace.visualstudio.com/items?itemName=GitHub.v...

Except EU doesn’t have big tech.

Just curious what their agenda can be…

This seems like a very socialistic perspective to me, which might not align well with big tech and innovation. I suggest re-reading "Atlas Shrugged" for this topic.

Unfortunately true, but one can also think that it’s a consequence and not the root cause. Why VC should invest if the environment isn’t supportive? A vicious circle :)

Europe is an amazing place, which is why I moved here in the first place. But for techies, it can be pretty challenging. The issue isn't with the techies themselves, though. The problem is that the environment here holds back big tech, making Europe heavily dependent on the US.

It's socialism versus capitalism at its finest. The problem is that socialism risks regulating big tech out of existence. Is that an issue? Absolutely. By stifling big tech, Europe becomes heavily dependent on the US, which ultimately hurts our quality of life. As an entrepreneur living in Europe, the best thing you can do is move to the US.

I’m living in Europe, I'm deeply disappointed by the current situation. The problems run much deeper than just regulations; they extend far beyond politics.

1. VCs outright avoid investing in deep tech, with only rare exceptions.

2. Founders overwhelmingly choose to build small, sustainable companies, steering clear of big tech.

3. Employees consistently prefer consulting jobs and value vacation days over equity.

4. The bureaucracy startups face when incorporating or raising funds is staggering (Germany, I'm looking at you).

While this may seem beneficial from a social perspective, it creates the worst possible environment for tech startups. I have immense respect for the few European startups that manage to survive and thrive despite these obstacles.

Can't wait to see how open-source tools for LLMs emerge! Big congrats to DVC for jumping into the LLM scene. Definitely excited to try it out, especially with LLMs running on dstack

Thank you! This link is very helpful.

Would love to give it a try! Thanks for answering my question.

Wow, somebody doesn’t like Docker enough to downvote my question.

Sounds really awesome. Any chance someone can suggest if this works also inside a Docker container?