PaLM 2 (L) also does not outperform GPT-4 on an MMLU 5-shot, but it does outperform LLaMA-I, a fine-tuned model for LLaMA. Here's what I pulled from the technical reports for LLaMA, PaLM 2, and GPT-4.
AI
-
Amazon Security Acquisition: Potential M&A Targets Analysis
By
–
Will Amazon go security shopping? Consider the take-out targets https://
thetechnologyletter.com/the-posts/will
-amazon-go-security-shopping
… $AMZN $MSFT $GOOGL #investing #MandA #stocks #cybersecurity #cloudcomputing -
Small Language Models: 3B and 7B Parameters Emerge in Open Source
By
–
We are seeing a lot of models pop up in the 3B and 7B param range, incl RedPajamas-INCITE 3B from Together that is based on the RedPajamas dataset. That dataset is designed to replicate the LLaMA dataset. I've heard Meta is also exploring 3B as an option for its fully OSS model.
-
Google PaLM 2 Comes in Four Sizes with Mobile Capabilities
By
–
When it comes to PaLM 2, Google said it would come in four sizes. It didn't disclose parameter count, but it did give us some tea leaves to evaluate, the first of which is that Gecko (its smallest) can run on a mobile device offline.
-
Meta Lattice Architecture Evolution for Business Growth
By
–
Going forward, we’ll continue iterating on Meta Lattice. This new architecture creates a more nimble system — one that’s more adaptable to broader market changes, can quickly utilize new AI innovations & operates more efficiently to deliver results that help businesses grow. 5/5
-
High-Capacity Ads System for Better Performance and Optimization
By
–
Better Performance
Using a high-capacity architecture allows our ads system to more broadly & deeply understand new concepts and relationships in data and benefits advertisers through joint optimization of a large number of goals. 2/5 -
Meta Lattice improves AI compute efficiency and innovation agility
By
–
Improved AI efficiency
We expect that transitioning to Meta Lattice will enable our fleet to improve compute efficiency. Maintaining & advancing fewer, but more powerful models also allows us to be more nimble with future innovations. 3/5 -
Meta Lattice: Faster Market Adaptation Through Generalized Learning
By
–
Faster adaptability to the shifting market landscape
Meta Lattice is capable of generalizing its learnings across domains and objectives. Fewer models also means we can proactively and efficiently update models and adapt to the fast-evolving market landscape. 4/5 -
Meta Lattice: New Model Architecture Improves Ads System Performance
By
–
Today we’re sharing details on Meta Lattice, a new model architecture that improves Meta’s ads systems performance and efficiency.
— AI at Meta (@AIatMeta) 12 mai 2023
More on this new work ➡️ https://t.co/0KXgAbk248
Three ways that this new work is enhancing our ads system 🧵 pic.twitter.com/Jt7t1ZkDyLToday we’re sharing details on Meta Lattice, a new model architecture that improves Meta’s ads systems performance and efficiency. More on this new work https://
bit.ly/3W07lkO Three ways that this new work is enhancing our ads system -
LangChain’s Challenge Converting Framework into Business Model
By
–
On the first one, no one disputes that LangChain is a transformative technology—especially as we move to more agent-based workflows. But LangChain is also a Startup That Has Raised Real Capital, and it faces an uphill climb to convert a beloved framework into a business.