Odia Generative AI (@OdiaGenAI) header
Odia Generative AI (@OdiaGenAI) profile picture

Odia Generative AI and LLM Initiative

Bhubaneswar
Joined April 2023
Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)Photo by Odia Generative AI (@OdiaGenAI)
Thanks to @silogenai (@AMDSiloAI) & Aku Rouhe for releasing Qwen1.5 7B Odia Instruct โ€” advancing LLMs for the Odia language! Trained on Odia data + fine-tuned with 5 Odia instruct sets released by @OdiaGenAI & 1 English set. ๐Ÿ‘‰ Model: huggingface.co/silogen/Qwen1โ€ฆ
3
8
277
Odia Generative AI retweeted
๐Ÿš€ Thrilled to welcome the Artificial Intelligence Institute @UofSC (#AIISC) as our Academic Partner for #GenerativeAI Workshop 2025! ๐Ÿ™ Grateful to Prof. Amit Sheth (@amit_p) for his guidance & support odiagenai.org/workshop-2025 @iitbbs @KIITUniversity @OdiaVAcademy @tfugbbsr
1
3
5
271
Excited to share that our open-source initiative @OdiaGenAI by our team @sambit_ai @swateekj @soumendrak_ @Babunisatya has been featured by @timesofindia & @timestechies! ๐ŸŽ‰ Check out the article: timesofindia.indiatimes.com/โ€ฆ Thank you, @timestechies, for the coverage!
1
6
11
294
Glad to see the Odia Llama2 paper selected for the Association for Computing Machinery (ACM) Showcase on Kudos. Link: growkudos.com/publications/1โ€ฆ Congratulations to all the authors. @guneetsk99 @sambit_ai
3
1
7
990
Odia Generative AI retweeted
๐Ÿ”ฅ ๐‘๐ž๐ฅ๐ž๐š๐ฌ๐ข๐ง๐  ๐ˆ๐ง๐๐ข๐œ ๐†๐ž๐ฆ๐ฆ๐š 7๐/2๐ ๐ˆ๐ง๐ฌ๐ญ๐ซ๐ฎ๐œ๐ญ๐ข๐จ๐ง ๐ญ๐ฎ๐ง๐ž๐ ๐ฆ๐จ๐๐ž๐ฅ ๐จ๐ง 9 ๐ˆ๐ง๐๐ข๐š๐ง ๐‹๐š๐ง๐ ๐ฎ๐š๐ ๐ž๐ฌ โ€” ๐๐š๐ฏ๐š๐ซ๐š๐ฌ๐š ๐Ÿš€ We are thrilled to share ๐ŸŒŸ ๐๐š๐ฏ๐š๐ซ๐š๐ฌ๐š, a Gemma 7B & 2B instruction-tuned models in 9 Indian Languages - Perhaps this is the first Indic open instruction-tuned model trained in 9 Indian languages additionally English included. ๐Ÿ”ฅ๐๐š๐ฏ๐š๐ซ๐š๐ฌ๐š is a Gemma 7B & 2B SFT model using Gemma 7B & 2B base models. Last week we released the Telugu Gemma 7B/ 2B SFT model using curated Telugu datasets from Telugu LLM Labs and we observed really good performance compared to Llama2-based models. ๐ŸŒ So, we thought why donโ€™t we scale up Gemma 7B & 2B models to multiple Indian languages and we went ahead with testing tokenizers of the following 9 Indian Languages and English Language. 1. Hindi 2. Telugu 3. Tamil 4. Malayalam 5. Kannada 6. Gujarati 7. Bengali 8. Punjabi 9. Odia 10. English โœจ We found the model to have the following capabilities: (X represents any other Indian language) 1. Instruction and Input in Native X language, Output in Native X language. 2. Instruction and Input in English language prompted to respond in Native X language, Output in Native X language. 3. Instruction in Native X language, Input in English language, and Output in Native X language. ๐Ÿ“Š๐“๐ซ๐š๐ข๐ง๐ข๐ง๐  ๐ƒ๐ž๐ญ๐š๐ข๐ฅ๐ฌ: 1. Single A100 machine which took approx. 36 hours for the 7B model and 15 hours for the 2B model. 2. Platform: E2E Networks Limited ๐Ÿ“ We have shared details on datasets, Examples of Reasoning, Translation, and Question Answering with Context in our blog post. ๐Ÿค The work would not have been possible without huge community effort from different languages and a huge shout out to each one of their work over the past few months showcasing the true OSS power. Following are details of contributors for the languages: 1. Hindi: @SarvamAI 2. Telugu: Telugu LLM Labs 3. Tamil: @abhinand58 4. Kannada: @adarshxs and the team at Tensonic 5. Malayalam: Vishnu Prasad J 6. Odia: @OdiaGenAI 7. Gujarati: Adarsh Shirawalmath and the team at Tensonic 8. Punjabi: HydraIndicLM 9. Bengali: HydraIndicLM ๐Ÿ‘ Special thanks toย @unslothai for simplifying the training and inference processes! ๐Ÿ”œ As we release these models, the next step is to create romanized datasets and we are working hard on evaluation datasets so that we can benchmark and improve on top of it. ๐Ÿค This work is done in collaboration withย @ramsri_gouthamย as part ofย the Telugu LLM Labsย independent initiative. ๐๐ฅ๐จ๐ ๐๐จ๐ฌ๐ญ: shorturl.at/jBQWY ๐‚๐จ๐๐ž๐๐š๐ฌ๐ž: shorturl.at/elxBF
13
46
11
218
30,760
Odia Generative AI retweeted
LLM Pre-training presentation by our interns Shakshi Panwar and Sanskruti Mishra youtu.be/yVWZTDFgHj8?si=zbE9โ€ฆ @Shantipriyapar3 @sambit_ai @odias_in_ai @soumendrak_
3
6
277
Thanks to Mohit Pandey @Analyticsindiam for the insightful article highlighting the necessity of an Indic LLM Leaderboard. We at @OdiaGenAI, are actively engaged in the development of the Indic Benchmark Dataset. analyticsindiamag.com/the-diโ€ฆ @sambit_ai @odias_in_ai
1
3
6
626
Sharing some of our @OdiaGenAI initiatives of building GenAI and LLM-based tools/technologies for the people of Odisha during a short interview at @kanak_news Thanks to @kalingashree @kanak_news Link: youtu.be/RL1IWQxDP0E?si=DC0dโ€ฆ @sambit_ai @odias_in_ai @soumendrak_ @BBSRBuzz
4
11
292
Odia Generative AI retweeted
I will be in the QA session!
Exciting news! ๐ŸŒŸ Join us for a QA Session with Dr. Swaroop Mishra @Swarooprm7, he'll be addressing queries on Instruction tuning and prompting large language models. ๐Ÿค–๐Ÿง  Don't miss out on this insightful session at the Generative AI and LLM Workshop by @OdiaGenAI. @odias_in_ai
1
3
26
3,758