Accelerate your research with Gemma 2, Google's 9B and 27B open weights models

O JavaScript não está habilitado em seu navegador. Em função disso, não é possível abrir este arquivo. Habilite o JavaScript e atualize a página.

Google is accelerating academic access for Gemma 2 models. These are 9B and 27B parameters, next-generation, faster, more powerful, and more efficient open weights models. Selected applicants will be awarded Google cloud credits (USD $10K per award) for novel academic research using Gemma 2. The application form will be open for 6 weeks (until Aug 9th 2024) from the launch date (June 27th 2024). Academics will be notified within four weeks of the application submission date.

By submitting this form, I acknowledge that my information will be used in accordance with Google's Privacy Policy and agree to Google's Terms of Service. I acknowledge that ideas contained in my application will not be treated as confidential or proprietary.

To request changes to or deletion of your data, and for any other queries, please contact gemma-academic-program@google.com.

Eligibility: Only individuals (faculty members, researchers, PhD scholars or equivalent) affiliated with a valid academic institution, or academic research organization can apply. Please note that cloud credits will be granted at Google's discretion.

First Name *

Last Name *

This email address will be used for all correspondence about the program.

Organization Name *

Organization Type *

University

Academic Research Lab

Outro:

Job Title / Role *

Faculty Member

Researcher Staff Member

Ph.D. Student

Outro:

Link(s) to your profile(s) *

e.g. Google Scholar / Website / DBLP

Research Interest(s) *

Evaluations, Benchmarks and Metrics for Multimodal LLMs

Ego-centric research for multimodal perception

Novel Applications of Multimodal LLMs (eg: creativity, scientific discovery)

Algorithmic Efficiency

Unlocking new capabilities through planning and reasoning

AI Alignment, Safety and Security

Multilingual / Multicultural Multimodal LLMs

Model compression for low-resource devices (on-device)

Outro:

Obrigatória

Project Abstract *

Please limit to 300 words. Note that this information will not be treated as confidential or proprietary.

Expected types of outcomes *

Please indicate the expected types of outcomes from the project. This could be publications, open-source code / data / models, tools, tech reports or any other artifact(s) that are relevant.

Country *

Enviar

Limpar formulário

Nunca envie senhas pelo Formulários Google.

Este formulário foi criado em Google.com. Privacidade e Termos

Formulários