Sebastian Raschka

@rasbt

Asst. Prof. of Statistics . Machine Learning & Deep Learning researcher. Author of 'Python Machine Learning' & open source contributor.

Madison, Wisconsin
sebastianraschka.com
सहभागी: ऑक्टोबर २०१२
१३२ छायाचित्रे किंवा व्हिडिओ छायाचित्रे आणि व्हिडिओज

ट्विट्स

आपण @rasbt यांना अवरोधित केले आहे

आपणास निश्चितपणे ही ट्विट्स पहायची आहेत? तसे केल्याने @rasbt यांना अनब्लॉक केले जाणार नाही.

  1. पिन केलेल्या ट्विट्स
    ५ जून, २०१९

    Just reorganized and uploaded all the TensorFlow and PyTorch models and methods I implemented for teaching in a fresh GitHub repo -- 80 Jupyter notebooks in total :)

    ३२ replies ८२१ पुन्हा ट्विट्स २,९६९ पसंत्या
    पूर्ववत् करा
  2. ९ एप्रि

    Just uploaded the introduction to convolutional neural networks part 2 (2/2) -- explaining the different Inception architectures and transfer learning

    ३ replies २८ पुन्हा ट्विट्स १६३ पसंत्या
    पूर्ववत् करा
  3. यांनी पुन्हा ट्विट केले
    ८ एप्रि

    Come chat with , , and myself on Friday at 9am where we'll be answering your GPU-accelerated ML questions. This is a great chance to chat in real time about topics like forest inference, distributed ML, and interop with libraries like CuPy and PyTorch.

    Join "Connect with the Experts"! Live interactive discussion with engineers about the latest GPU-accelerated machine learning in , this Friday (April 10th) at 9AM. Free registration! (But please register 24 hours in advance!) Session CWE21728 -
    हा थ्रेड दाखवा
    १ reply ७ पुन्हा ट्विट्स १८ पसंत्या
    पूर्ववत् करा
  4. यांनी पुन्हा ट्विट केले
    ७ एप्रि

    One thing we have been doing lately: an explicit 'Limitations' section in the Discussion section of our papers. Really like it in terms of communicating the main advance. Reviewers seem to like it too.

    १७ replies २७ पुन्हा ट्विट्स २०५ पसंत्या
    हा थ्रेड दाखवा
    हा थ्रेड दाखवा
    पूर्ववत् करा
  5. ७ एप्रि

    Just recorded the Intro to Convolutional Neural Networks part 2, covering padding, spatial dropout, and common CNN architectures: VGG-16, ResNets, and fully convolutional nets:

    ३३ पुन्हा ट्विट्स १७६ पसंत्या
    पूर्ववत् करा
  6. यांनी पुन्हा ट्विट केले
    ६ एप्रि

    Mimicry: a lightweight library aimed towards the reproducibility of GAN research - standardized implementations of popular GANs - baseline scores of GANs for comparison - a framework for GAN training boilerplate code

    १ reply ७० पुन्हा ट्विट्स २४९ पसंत्या
    पूर्ववत् करा
  7. यांनी पुन्हा ट्विट केले
    ६ एप्रि

    Our basic goodness shines as we take on the challenges of . professor Richard Davidson, founder of , reflects on how is an act of generosity and compassion toward others.

    ११ पुन्हा ट्विट्स २१ पसंत्या
    पूर्ववत् करा
  8. यांनी पुन्हा ट्विट केले
    ६ एप्रि

    RAPIDS 0.13 release is all systems go. The Great libcudf++ port is complete, 1.0 integrated, cuDF/cuGraph much faster, cuML adds new multi-node algos, cuSignal is official, and much more from the community. Read it all here -

    १ reply ३९ पुन्हा ट्विट्स ९० पसंत्या
    पूर्ववत् करा
  9. ६ एप्रि

    "Effective PyTorch" -- A nice GitHub repo with some useful tips on "Optimizing runtime with TorchScript" and "Numerical stability in PyTorch"

    ६९ पुन्हा ट्विट्स ३९३ पसंत्या
    पूर्ववत् करा
  10. ४ एप्रि

    Excited to share that our paper "Machine Learning in Python" article was published today! In this article, , , and I discuss GPU-enabled machine learning and Deep Learning in Python -- the most recent trends and developments in ML+DL

    ११ replies २२० पुन्हा ट्विट्स ७२३ पसंत्या
    पूर्ववत् करा
  11. ४ एप्रि

    "Swift: Google’s bet on differentiable programming" via Nice writeup by & Interesting take on using an already popular lang with integrated differentiable programming capabilities for DL And Swift syntax really doesn't look too bad, actually:

    ४ replies ३९ पुन्हा ट्विट्स १९० पसंत्या
    पूर्ववत् करा
  12. ३ एप्रि

    "Weighted Random Search for CNN Hyperparameter Optimization" -- a simple method that seems to perform pretty well in practice.

    १३ पुन्हा ट्विट्स ६९ पसंत्या
    पूर्ववत् करा
  13. २ एप्रि

    Wow! The first time ever that I am impressed by German bureaucracy. But joking aside, that's really great.

    Germany's response is incredible. Self-employed business owner I know applied for government support on Sunday. Money is on his account on Tuesday. Wow. I hope other countries will understand they need to act fast. Support is needed now. Not in 2 months.
    ४ पुन्हा ट्विट्स १८ पसंत्या
    पूर्ववत् करा
  14. २ एप्रि

    Lecture 12 is live! "L12 Intro to Convolutional Neural Networks (Part 1)" -- There's lots to talk about regarding CNNs, the challenge was cherrypicking the most important topics ;)

    ५ replies ३१ पुन्हा ट्विट्स १६७ पसंत्या
    पूर्ववत् करा
  15. १ एप्रि

    Peer review forms the backbone of modern scientific manuscript eval. [...] does this protocol remain fit for purpose in 2020? [...] we answer this question in the negative (strong reject, high confidence) and propose instead State-Of-the-Art Review (SOAR)

    २ replies ६ पुन्हा ट्विट्स ३४ पसंत्या
    पूर्ववत् करा
  16. यांनी पुन्हा ट्विट केले
    १ एप्रि

    Nobody believes a model except the person that writes it down. Everyone believes the data, except the person that takes the measurements.

    As we are having a moment. So given people are paying attention to us, what’s the one epi methods lesson you would like the public to understand? What would be the top thing on your epi cheat sheet?
    ६ replies २०१ पुन्हा ट्विट्स ७५८ पसंत्या
    पूर्ववत् करा
  17. यांनी पुन्हा ट्विट केले
    ३१ मार्च

    is always for the common good. Quick search for COVID-19 papers:

    १ पुन्हा ट्विट ५ पसंत्या
    पूर्ववत् करा
  18. ३१ मार्च

    New week new recordings: Intro to Deep Learning L11: Learning rate decay, momentum learning, and adaptive learning

    २ replies ३७ पुन्हा ट्विट्स १६१ पसंत्या
    पूर्ववत् करा
  19. ३१ मार्च

    oops, wrong link!

    हा थ्रेड दाखवा
    हा थ्रेड दाखवा
    पूर्ववत् करा
  20. ३० मार्च

    torchlayers: an interesting new abstraction layer API on top of PyTorch, which is basically an "Sequential" object w/o the need to call it in the forward method. And, more interestingly, it performs automatic dimensionality and shape deduction:

    १ reply ८३ पुन्हा ट्विट्स ३४७ पसंत्या
    पूर्ववत् करा
  21. यांनी पुन्हा ट्विट केले
    २८ मार्च

    Poorly-used pie chart strikes again!

    २२ replies ७७ पुन्हा ट्विट्स २४६ पसंत्या
    पूर्ववत् करा

@rasbt यांनी अजूनही ट्विट केलेले नाही.

लोड करण्या करता काही वेळ लागू शकतो.

Twitter वरची क्षमता ओलांडली गेली आहे किंवा तात्पुरती अडचण अनुभवास येत आहे. पुन्हा प्रयत्न करा किंवा अधिक माहितीसाठी Twitter स्थिती येथे भेट द्या.

    आपण मला सुद्धा विचारू शकता

    ·