Обложка канала

Spark in me - Internet, data science, math, deep learning, philosophy. Страница 12

2440 @snakers4

Канал про интересные мне темы - интернет - статистика - наука о данных Без рекламы и буллшита.

  • Spark in me - Internet, data science, math, deep learning, philosophy

    Mick Gordon vs Bethesda This letter is just pure gold. I participated in similar corporate shit fuckery a couple of times, and the pain is too real. medium.com/@mickgo…266b27ce
    My full statement regarding DOOM Eternal

    This statement is issued in response to id Software Studio Director Marty Stratton’s “DOOM Eternal OST Open Letter”, published on Reddit.

    Medium
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Actually amazing codec by FAIR Links [1] [2] - 10-20x slower than OPUS on 1 CPU thread - RTF of ~1x or ~10x (only decoding) - Trained ... on 8 x A100 - Is an encoder-decoder model - Streamable for 24 kHz, non-streamable for 48 kHz - What is the deal if it is 10-100x slower? It has similar quality for ~6 kbps bitrate compared to 24 kbps bitrate of OPUS. Opus degrades drastically with 6-12 kbps
  • Spark in me - Internet, data science, math, deep learning, philosophy

  • Реклама

  • Spark in me - Internet, data science, math, deep learning, philosophy

    Looks like GitHub exists in a parallel reality While it supports 3 out of 5 BRICS countries now, it still looks like the amount of financial support is mostly peanuts. Also if you explore the most "used" software, of course it will be endless JS packages. github.blog/2022-11…n-source
    Bringing greater financial sustainability to open source communities | The GitHub Blog

    We know that companies benefit from open source. That's why we’re making it easier for companies to financially support projects.

    The GitHub Blog
  • Spark in me - Internet, data science, math, deep learning, philosophy

    MlPerf Training v2.1 Results New Nvidia press releases and ML Perf benchmarks also seem more and more hypocritical each time. Notable: - No consumer GPUs at all - Highly sublinear scaling for many tasks (4-8 GPUs seem most efficient) - No proper apples-to-apples H100 test - AMD EPYC Vendor lock in place - "Open" tests are barely there - As predicted this is a playground for hardware vendors - On RNNT and Imagenet H100 2x faster, on BERT - 3-4x
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Голос нейронных овощей

    Клип на песню "Голос Овощей" группы "Бездна анального угнетения" сгенерирован с помощью нейросети Stable Diffusion

    YouTube
  • Spark in me - Internet, data science, math, deep learning, philosophy

    This is so nice When you also find out that their power generation is dominated by coal ... And they are one of the biggest uranium exporters ... Well you get the idea, right?
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Honest Government Ad | Visit Western Australia

    The Western Australien Government has made a tourism ad and it’s surprisingly honest and informative. 👉 Ways you can support us to keep making videos: 🔹 Become a Patron: https://www.patreon.com/thejuicemedia 🔹 Tip us on PayPal: https://www.paypal.me/thejuicemedia 🔹 Other options: https://www.thejuicemedia.com/support 👉 Speak up for the Burrup: saveoursonglines.good.do 👉 MERCH: https://shop.thejuicemedia.com 👉 PODCAST: https://www.thejuicemedia.com/podcast 👉 PG VERSIONS: https://www.thejuicemedia.com/teachers 👉 CREDITS 🔹 Produced by Patrons of The Juice Media 🔹 Written by Giordano for The Juice Media 🔹 Acted by Ellen Burbidge: https://www.youtube.com/ellenburbidge 🔹 Voice by Lucy Cahill 🔹 Animations by Brent Cataldo 🔹 Thanks to the Australia Institute for research assistance: https://www.youtube.com/c/TheAusInstitute 🔹 Thanks to Save Our Songlines for Murujuga Photos 🔹 Thanks to Gerard Mazza for filming Murujuga background 🔹 Thanks to Greenpeace for footage 🔹 Thanks to Ac Jamel for Perth drone footage: https://www.youtube.com/watch?v=dJ2zKtQdXOA 🔹 Thanks to Miklovin for Perth drone footage: https://www.youtube.com/watch?v=VVAAnr7jT8c 🔹 Soundtrack by Original_Soundtrack: https://audiojungle.net/item/a-happy/34676340

    YouTube
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Open Source Collective ⛓ Read between the lines ... only people from more equal countries will be accepted. ⚠️ Also kind of unsettling ... that the majority of OSC community supported FOSS projects (most of them from OECD countries) ... cannot appoint 2 maintainers. ☭ This is just hilarious that they have to outdo themselves in double-speak to make the FOSS community put their own chains.
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Choking Off China’s Access to the Future of AI New U.S. Export Controls on AI and Semiconductors Mark a Transformation of U.S. Technology Competition with China www.csis.org/analysi…uture-ai PS Wake up Mr. Anderson I like the American arrogance kind of assuming that Taiwan is not China, that sanctions against Russia are very effective and the fact that all of these actions are lawful and not an act of open warfare. The good thing is that the longer they are in denial, the easier it is for the rest of the world to topple the "chokeholds" over.
    Choking Off China’s Access to the Future of AI

    Download the Report Introduction On October 7, 2022, the Biden administration announced a new export controls policy on artificial intelligence (AI) and semiconductor technologies to China. These new

    www.csis.org
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Shorter version
  • Spark in me - Internet, data science, math, deep learning, philosophy

    ML Data Fair Use ... is still a grey zone and and a form of Western exploitation and molding the whole world to their image / system / etc:
    By offer­ing Copi­lot as an alter­na­tive inter­face to a large body of open-source code, Microsoft is doing more than sev­er­ing the legal rela­tion­ship between open-source authors and users. Arguably, Microsoft is cre­at­ing a new walled gar­den that will inhibit pro­gram­mers from dis­cov­er­ing tra­di­tional open-source com­mu­ni­ties. Or at the very least, remove any incen­tive to do so. Over time, this process will starve these com­mu­ni­ties. User atten­tion and engage­ment will be shifted into the walled gar­den of Copi­lot and away from the open-source projects them­selves—away from their source repos, their issue track­ers, their mail­ing lists, their dis­cus­sion boards. This shift in energy will be a painful, per­ma­nent loss to open source.
    githubcopilotinvestigation.com… Unless you have some corporate / deep pocket sponsors and / or you are doing something truly unique, there is LESS AND LESS incentive to publish FOSS. I like the story of libraries like nmslib ... where they get all the accolades possible, but no visible financial benefits. The system just does not function this way.
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Other Digests 2022-10 Link #digest
    Other Digest 2022-10

    ____________________________________________________________________________ # Blogs The Real Name Fallacy - https://coralproject.net/blog/the-real-name-fallacy/ The Next Century of Computing - https://bzogrammer.substack.com/p/the-next-century-of-computing SKAdNetwork 4.0: what problems does it solve? - https://mobiledevmemo.com/what-problems-with-skadnetwork-4-0-solve/ Amazon Web Services Infrastructure Inefficiencies Cause Cuts To Twitch Driving Creators To Google’s YouTube - https://semianalysis.substack.com/p/amazon…

    Telegraph
  • Spark in me - Internet, data science, math, deep learning, philosophy

    ML Digest 2022-10 📌 Digest link 📌 This month spotlight goes to: - The State of AI Report itself - My Lazy Take on State of AI Report 2022 #digest
    Digest 2022-10

    ____________________________________________________________________________ # ML Whisper English benchmarks - https://twitter.com/lunixbochs/status/1574848899897884672 Listen to our major Text to Speech upgrades for 64 bit devices. - https://android-developers.googleblog.com/2022/09/listen-to-our-major-text-to-speech-upgrades-for-64-bit-devices.html AudioGen: Textually Guided Audio Generation - https://felixkreuk.github.io/text2audio_arxiv_samples/ The Illustrated Stable Diffusion - https://jalammar.githu…

    Telegraph
  • Spark in me - Internet, data science, math, deep learning, philosophy

    PyTorch 1.13: beta versions of functorch and improved support for Apple’s new M1 chips - torchaudio seems to be kind of frozen - CUDA 10.2 and 11.3 deprecated - BetterTransformer - import functorch without installing another package - Native builds for Apple's new M1 chip as a beta feature, - Python 3.11 preview - Default ONNX opset updated to 14 Looks like they are trimming the fat since the migration to the Linux foundation. Links: - https://pytorch.org/blog/PyTorch-1.13-release/ - github.com/pytorch…/v1.13.0
    PyTorch 1.13 release, including beta versions of functorch and improved support for Apple’s new M1 chips.

    We are excited to announce the release of PyTorch® 1.13 (release note)! This includes Stable versions of BetterTransformer. We deprecated CUDA 10.2 and 11.3 and completed migration of CUDA 11.6 and 11.7. Beta includes improved support for Apple M1 chips and functorch, a library that offers composable vmap (vectorization) and autodiff transforms, being included in-tree with the PyTorch release. This release is composed of over 3,749 commits and 467 contributors since 1.12.1. We want to sincerely thank our dedicated community for your contributions.

    pytorch.org
  • Реклама

  • Spark in me - Internet, data science, math, deep learning, philosophy

    How the Might Have Fallen

    This downturn in semis is different from past cycles in that each segment seems to be taking a beating at different times. This cycle is serial, not parallel, and we probably need to see all of tha…

    Digits to Dollars
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Yeah, looks like OEM model won
  • Spark in me - Internet, data science, math, deep learning, philosophy

    Why pay "Open"AI when another company boasts a valuation of 50% of your market cap and publishes a better model ... for free?