Try it out!
Grok voice early beta is now available on the @Grok app.
This is early beta, so expect issues (that will be resolved fast), but it’s still awesome.
Ting Chen
291 posts
- You can try grok vision in voice mode by turning on the camera too!
- Our goal is bit stream in and bit stream outYou can also upload any image to Grok, including memes, and it will explain what they mean x.com/i/grok/share/4…
- Introducing SimCLR: a Simple framework for Contrastive Learning of Representations. SimCLR advances previous SOTA in self-supervised and semi-supervised learning on ImageNet by 7-10% (see next). arxiv.org/abs/2002.05709 Joint work with @skornblith @mo_norouzi @geoffreyhinton.
- A new update coming up soon.And Grok said, “Let there be voice,” and there was voice.
- Have you wondered why object detection, unlike classification, has so many sophisticated algorithms? With Pix2Seq (arxiv.org/abs/2109.10852), we simply cast object detection as a language modeling task conditioned on pixels! (with @srbhsxn, Lala Li, @fleet_dj, @geoffreyhinton)
- We have recently added multilingual and camera support to the voice mode. Give it a try!It’s awesome
- Yes, the native voice experience is coming to Grok soon! Let us know what specific features you want to see (or hear)!Woah! That was one of the most unexpectedly rewarding hours of my life. Instead of passively listening to an audiobook about physics while doing errands as usual I had an hour long back-and-forth conversation with Ara from Grok 3 about a bunch of scientific topics. We started
- Just a beginning. Multimodal understanding and generation capabilities will be rapidly improving. DM open, come and join us!
- SimCLRv2: an improved self-supervised approach for semi-supervised learning. On ImageNet with 1% of the labels, it achieves 76.6% top-1, a 22% relative improvement over previous SOTA. arxiv.org/abs/2006.10029 Joint work with @skornblith, @kswersk, @mo_norouzi, @geoffreyhinton
- 📢Introducing Pix2Seq-D, a generalist framework casting panoptic segmentation as a discrete data generation task conditioned on pixels. Works for both images and videos, with minimal task engineering. arxiv.org/abs/2210.06366 work w/ Lala Li, @srbhsxn @geoffreyhinton @fleet_dj
00:00 - Happy to share that we've open-sourced both the code and pretrained models for SimCLR (a simple framework for contrastive learning of visual representations): github.com/google-researc… joint work with @skornblith, @mo_norouzi and @geoffreyhinton.
- Can we solve a (large) portion of vision tasks by simply formulating it as translating raw pixels into tokens/bits with higher level abstraction? A question that took a 2-year journey to get an answer: oh sure, if you know how to train a really good generative model🥳
- A few months ago, we made a decision to focus on autoregressive modeling for lots of good reasons like reusing the scalable LLM training and inference stacks here at @xai, but also met with many challenges. It is really amazing to see how far we are able to push it forward!Earlier today, we released a new model, code-named Aurora, that gives Grok the ability to generate extremely photorealistic images (and in the future, even edit them). It's free to use for all of 𝕏, try it out and send us what you're creating! This model was trained entirely
