Partha Talukdar
@partha_p_t
Gemini Multilinguality. Languages@ Google DeepMind India
Great to see improved Indic language tokenization & processing capabilities in Gemini & Gemma models helping the ecosystem build further. Congratulations to @SarvamAI on the Sarvam-translate launch! #AIForAll
Congratulations to @SarvamAI on advancing AI-powered translation 🎊 Built with Gemma 3 4B-IT, Sarvam-Translate supports 22 Indian languages handling document translation, markdown, OCR, and more. deepmind.google/models/gemma/g…
It's extremely gratifying to see so many contributors to Gemini 2.5 from @GoogleDeepMind India! Sip filter coffee (w/ plant-based milk, of course) as you pave the path to AGI, can't think of a better deal 🙂 (Btw, we are growing, app link coming soon!) arxiv.org/abs/2507.06261
Bragging time: incredibly proud of my fantastic nephew @_Trinav on his very first self-composed set of piano pieces. His passion and perseverance for music over the 14+ yrs journey is truly inspiring! Please take a listen open.spotify.com/track/13oUhedp… youtube.com/@trinavtalukda…
Google DeepMind Tokyo is growing again! We're building the future of AI with Project Astra. Join my team! We're hiring a Research Engineer to lead the challenge from Tokyo 🗼 job-boards.greenhouse.io/deepmind/jobs/… #GoogleDeepMind #AI #Jobs #Tokyo
Google DeepMind India is hiring for research scientist role in multicultural & multimodal modeling. Strong candidates with proven research experience are encouraged to apply I shall be at #icassp2025 Hyderabad on Apr 8, happy to meet and chat, pls DM job-boards.greenhouse.io/deepmind/jobs/…
Thanks @Harman26Singh for your contributions! Among the many many gifts Gemma 3 brings, a more effective multilingual tokenizer, including for Indic languages, is one. Please check it out! blog.google/technology/dev…
Excited for Gemma 3 and to have contributed a bit to the effort! It’s a really strong model with great multilingual performance—including in Indian languages! 👇👇 Report link: storage.googleapis.com/deepmind-media…
Looking forward to seeing what people build with this unique multimodal geo-anchored dataset. This is Phase 1 data and more coming.
Project Vaani has released new open-source data from its Phase 1 which comprises of images, audio, and transcribed speech, encompassing 54 languages across 80 districts in 12 states of India. The dataset can be accessed here: vaani.iisc.ac.in bhashini.gov.in/sahyogi-projec…
Great to see this resource getting released for 115 very low data-resource languages. Please check it out. #AIForAll
😼SMOL DATA ALERT! 😼Anouncing SMOL, a professionally-translated dataset for 115 very low-resource languages! Paper: arxiv.org/pdf/2502.12301 Huggingface: huggingface.co/datasets/googl…
Today, we have open-sourced the complete code for CUBE, including i) cube extraction code, ii) cultural diversity evaluations using Vendi score, and iii) CUBE dataset. Check it out! Github: github.com/google-deepmin… Thanks for all the interest, patience and discussions during…
🚨 New paper alert 🚨: T2I models are about to revolutionize industries...but do they work equally well for all cultures ? 🌏 What is the *cultural diversity* of images generated for under-specified prompts? 🤔 Our latest research delves into this! arxiv.org/abs/2407.06863
Heartiest congratulations to all three winners for their ground-breaking work! I met David during his @iiscbangalore visit a few years back. We discussed GCNs and proteins over chai on campus which led to a collaboration on these topics :) biorxiv.org/content/10.110…
”David Baker, Demis Hassabis and John Jumper, your ground-breaking work in computational protein design and protein structure prediction has revolutionised these fields. It has opened up completely new possibilities to design proteins that have never been seen before, and we now…
I'll be presenting this on 13th Dec (Fri) at #NeurIPS2024 📍West Ballroom #5108 🕦 11 AM - 2PM I am excited to discuss CUBE: 1) cultural competence evaluation 2) our new T2I evaluation dimension for diversity, 3) pareto fronts and how T2I models are getting worse wrt diversity
🚨 New paper alert 🚨: T2I models are about to revolutionize industries...but do they work equally well for all cultures ? 🌏 What is the *cultural diversity* of images generated for under-specified prompts? 🤔 Our latest research delves into this! arxiv.org/abs/2407.06863
Application are now open for the Predoc 2025 program at Google DeepMind India. If you are pre-PhD (i.e., current undergrad or masters students or recent geaduates) and interested in research, then this should be a no-brainier. Please apply by Dec 18, 2024 google.com/about/careers/…
🚀 Join the Gemini Multilinguality team @GoogleDeepMind 🌐 We’re looking for researchers passionate about making LLMs helpful for all. Dramatically improve model quality, coverage, and cultural relevance across hundreds of languages. #NLProc #MultilingualAI #i18n #LLMs…
There are a couple of full-time openings for pre-doctoral research associates in my group. One of them is specifically for candidates interested in ensuring that large language (and vision) models are geo-culturally inclusive. Email me if you'd be interested.
Fantastic opportunities with an incredible team. Please apply.
Excited to share that the Machine Learning and Optimization team at @GoogleDeepMind India is hiring Research Scientists and Research Engineers! If you're passionate about cutting-edge AI research and building efficient, elastic, and safe LLMs, we'd love to hear from you. Check…
On Teacher’s Day, I remember Miss Begum, who left this world a few days back. I am forever grateful to her for the curiosity she nurtured & encouragement to push the boundary. I am fortunate to have found a teacher like her & wish the same for all. Thank you Miss for everything.
BoF session on Indian languages and LLMs at #ACL2024. Thanks everyone for a great discussion.
Our Technology, AI, Society, & Culture (TASC) team at Google Research (@GoogleAI) has an opening for a Visiting Researcher position in the Society Centered AI & ML org, starting 2025. Our topic of research is on culturally aware AI. Here's some of our recent work on this topic: