Microsoft releases Phi-2, an AI language model the company says outperforms Mistral and Llama 2 at 7B and 13B parameters and is small enough to run on a phone
Carl Franzen / VentureBeat :
VentureBeat Carl Franzen
Related Coverage
- Phi-2: The surprising power of small language models Microsoft Research · Alyssa Hughes
- Microsoft Packs Power Into Its New Small AI Model Phi-2 Decrypt · Jose Antonio Lanz
- Phi-2's Debut: Challenging Conventional AI Model Dominance WinBuzzer · Luke Jones
- Microsoft Releases Phi-2, Outperforms Gemini Nano, Mistral 7B, and Llama 2 Models Analytics India Magazine · Siddharth Jindal
- Microsoft debuts 2.7B-parameter Phi-2 model that outperforms many larger language models SiliconANGLE · Mike Wheatley
- Microsoft Says Its Homegrown AI Model Outperforms Meta's Llama 2, Runs On Phones The Information · Aaron Holmes
- Wouldn't it be great to host an LLM model on your own laptop or even cell phone? This could help with a ton of cybersecurity regulations and signifiantly improve the OS functionalities. … Sohrab Rahimi
- Microsoft introduce Phi-2 a new SLM (small language Model) of 2.7B parameters that outperforms models of 7 to 13B on some reasoning and language understanding capabilities. … Stephane Requena
- Another LLM (actually “small” language model) is released. Phi-2 from Microsoft. 2.7B size, you can run it on a laptop or mobile. … Jia Li
- Phi-2: The surprising power of small language models Hacker News
Discussion
-
@vitor_dlucca
Vitor de Lucca
on threads
Yesterday, new Mistral killed Gemini Pro. Today, Microsoft showed us that GPT4 Turbo with proper prompting is better than Gemini Ultra. Microsoft also launched Phi-2. A small LLM (2.7b) that beat Gemini Nano (3.2b), Mistral 7b, and beat llama 70b on coding.
-
@msftresearch
@msftresearch
on x
Today, we share our teams' latest contributions, Phi-2 and promptbase. Phi-2 outperforms other existing small language models, yet it's small enough to run on a laptop or mobile device. https://www.microsoft.com/...
-
@sebastienbubeck
Sebastien Bubeck
on x
Phi-2 numbers, finally! We're seeing a consistent ranking: phi-2 outperforms Mistral 7B & Gemini Nano 2* (*on their reported benchmarks) and is roughly comparable to Llama 2-70B (sometimes better, sometimes worse). Beyond benchmarks, playing with the models tells a similar story.…
-
@marktenenholtz
Mark Tenenholtz
on x
Microsoft just released Phi-2 (2.7B) which seems to come close to Llama 70B. It's trained on 1.4T tokens of heavily filtered web data and a lot of GPT 3.5/4 outputs. Knowledge distillation really does work, and combining it with other high quality data could be the key for OSS. […
-
@andrewcurran_
Andrew Curran
on x
Phi-2 was trained on 1.4T tokens, the training took 14 days on 96 A100 GPUs. They claim to have already surpassed Google Nano despite being a smaller size. [image]
-
@linusekenstam
Linus
on x
Phi-2 A new large language model that beats its rivaling 25x larger counterpart. What else? It runs on your mobile device. Truly personal LLM's that work offline using edge compute is here https://www.microsoft.com/... [image]
-
@erichorvitz
Eric Horvitz
on x
More https://www.microsoft.com/... on promptbase & advances in prompting, including how we steered GPT-4 to record performance on the Measuring Massive Multitask Language Understanding (MMLU) benchmark, spanning 57 areas @satyanadella @kevin_scott @peteratmsr @HarshaNori @MSFTRes…
-
@emostaque
@emostaque
on x
Smol models ftw [image]
-
@richinseattle
Richard Johnson
on x
Contrary to rumors, I assure you the Phi-2 model is downloadable, but it is against the license to redistribute it. You can login to Azure ML Studio (free with a basic Azure account), select it from the Model Catalog, and then download the files from the Artifacts tab. [image]
-
@sebastienbubeck
Sebastien Bubeck
on x
Sorry I know it's a bit confusing: to download phi-2 go to Azure AI Studio, find the phi-2 page and click on the “artifacts” tab. See picture. [image]
-
@satyanadella
Satya Nadella
on x
From new best-in-class small language models to state-of-the-art prompting techniques, we're excited to share these innovations and put them in the hands of researchers and developers.
-
@alphasignalai
Lior
on x
Microsoft just released Phi-2, a 2.7B LLM that rivals the 25x bigger LLaMa-2 70B. The best part? The model is small enough to run on a laptop or mobile device. Trained on 1.4T tokens: mixture of synthetic & web datasets, it beats Mistral 7B and Llama-2-70B model on muti-step... […
-
@simonw
Simon Willison
on x
Anyone figured out how to run Microsoft's new Phi-2 LLM on a Mac yet? https://www.microsoft.com/...