Here's info on the pre-training emissions of the Gemma 2 series of models, from a paper we released a few months ago (end of July 2024):
GENERATIVE AI
-
Gemma 2 Model Series: Emissions Estimates for AI Training
By
–
In July 2024, we released emissions estimates for pre-training the Gemma 2 model series (2B, 9B, and 27B parameter models, so ~12X, ~43X, and ~128X larger than the 213M parameter models you were looking at, and respectively trained on 2T, 8T, and 13T tokens of data). We estimate
-
Microsoft Copilot introduces Think Deeper for enhanced reasoning capabilities
By
–
"When you Think Deeper with Copilot you can get help working through a tricky coding or math problem, or even planning for complex life events. With better reasoning capability, Think Deeper spends more time considering your scenario and replies with a more comprehensive… pic.twitter.com/F2xWSARnHI
— 🚨 AI News | TestingCatalog (@testingcatalog) 8 octobre 2024"When you Think Deeper with Copilot you can get help working through a tricky coding or math problem, or even planning for complex life events. With better reasoning capability, Think Deeper spends more time considering your scenario and replies with a more comprehensive
-
Microsoft Copilot Introduces Vision and Think Deeper Reasoning Features
By
–
Finally had a chance to read the latest MS Copilot announcement properly. It turns out that it will have its own Copilot Labs program with 2 pilot features: Copilot Vision and Think Deeper. Think Deeper is essentially an o1 with reasoning
-
Using ChatGPT Prompts to Enhance Student Engagement and Lesson Planning
By
–
Elevate Course Engagement Prompts boost student interest • Utilize ChatGPT prompts
• Create interactive content
• Streamline lesson planning Read more: https://
buff.ly/4ezpkqO -
Runway Gen-3 Alpha Turbo Adds First and Last Frame Control
By
–
Runway got First and Last frame support on Gen-3 Alpha Turbo 👀👀👀
— 🚨 AI News | TestingCatalog (@testingcatalog) 8 octobre 2024
This feature allows users to select two different images that would define how the generated video should start and end. https://t.co/fe1Wm3ai38 pic.twitter.com/LaLDg05OxjRunway got First and Last frame support on Gen-3 Alpha Turbo This feature allows users to select two different images that would define how the generated video should start and end.
-
Zillow Acquires Virtual Staging AI for Real Estate Listing Images
By
–
'Zillow Group has acquired Virtual Staging AI (VSAI), a company whose technology helps home sellers, agents and photographers create digitally staged listing images without the need for physical furniture and decor.' GIF via Zillow https://t.co/PCtGOkbzdN pic.twitter.com/IudL8Ab8oz
— Jordan Novet (@jordannovet) 8 octobre 2024'Zillow Group has acquired Virtual Staging AI (VSAI), a company whose technology helps home sellers, agents and photographers create digitally staged listing images without the need for physical furniture and decor.' GIF via Zillow https://
zillowgroup.com/news/zillow-gr
oup-acquires-ai-company-offering-virtual-staging/
… -
Checkr Optimizes Hiring with Fine-Tuned Open Source Small Models
By
–
@checkr , Inc. is changing the way companies hire with #AI-driven automation and their doing it with small models (#SLMs)! Check out our latest case study to hear how Vlad Bukhin and the Checkr team fine-tuned small #opensource models that are more accurate, 30x #faster
-
Building Account Qualification Agents: Cognitive Architectures Explained
By
–
Agents are here, you just have to know how to build them Excited to share this VERY detailed blogpost from @unifygtm on how they built their recent account qualification agent This contains all sorts of nuggets like:
The three different cognitive architectures they considered -
Meta Movie Gen: 13B Audio Generation Model for Video Synthesis
By
–
As part of Meta Movie Gen, we trained a 13B parameter audio generation model that can take a video + optional text prompts to generate high quality audio — including ambient sound, foley & instrumental background music — all synced to the video. Details https://
go.fb.me/vgkswc