How many params do y’all think Gecko has
@mattlynley
-
PaLM 2 Multiple Sizes Released Gecko Runs Locally
By
–
Waiting for more details on this but looks like there will be multiple sizes for PaLM 2. No param counts mentioned but that Gecko can run on a local device.
-
Hugging Face Transformers Agents: Safe Code Execution Explained
By
–
"This code is then executed with our small Python interpreter on the set of inputs passed along with your tools. We hear you screaming 'Arbitrary code execution!' in the back, but let us explain why that is not the case." https://
huggingface.co/docs/transform
ers/transformers_agents
… -
Startup Funding Shifts: Post-LLM Hype Era Begins Now
By
–
The question now is: who else? So many startups raised at unicorn levels. While LLMs haven't permeated everything yet—and that'll take a VERY long time—that hype train has stalled and eyes have shifted to more LLM-focused startups.
-
W&B and LLM Stack Leaders Navigate Post-GPT Era
By
–
W&B stuck the landing in the post-GPT era, along with several other hot startups like MosaicML and Anyscale that proved to be an important part of the LLM-powered stack (in particular for training and inference, respectively, based on people I've talked to).
-
W&B Raises Capital Driven by Machine Learning and LLM Growth
By
–
W&B didn't need to raise the round, obviously. It's grown substantially on the back of the machine learning—including LLMs. OpenAI is a customer of W&B. So it was a situation where it might as well ask for a big number, and if it didn't stick, that's fine.
-
Weights & Biases $2B valuation amid AI startup funding trends
By
–
The best example of this I've heard is Weights & Biases, which held some discussions earlier this year for a round that would value it at $2 billion. The conversations that picked up was that the price was high—even though startups like Pinecone were fetching 200x+ ARR multiples.
-
ML Startups Pre-GPT Era: VC Shift to LLM-Powered Ventures
By
–
Today's issue of Supervised is about startups that jumped on the machine learning hype train… before the GPT era of machine learning. 2020-22 investments with eye-popping valuations in ML startups. Now VCs have shifted eyes to LLM-powered startups.
-
RedPajamas and INCITE Models: New Open Source LLM Generation
By
–
And of course LLaMA isn't the only one. RedPajamas, from @togethercompute
, is based on the LLaMA training set and will likely breed a whole new generation of LLaMA-like OSS models—as well as its own RedPajama-INCITE 3B/7B models (keep an eye on that 3B one!!) -
Maximizing Power in Smaller AI Models for Local Deployment
By
–
The goal here is to squeeze as much power out of as small a model as possible. The smaller the model, the easier it is to update and tune—as well as run it cheaply (or even locally on a device).