Sarvam AI Launches New Language Model for Indian Languages
Sarvam AI has launched its new language model, Sarvam-1, focused on Indian languages. With 2 billion parameters, the open-source model supports 10 languages, including Hindi, Tamil, and Bengali, along with English. Sarvam-1 is designed for multilingual tasks and trained on Sarvam-2T, a dataset of 2 trillion tokens. It outperforms larger models like Meta’s Llama-3 and Google’s Gemma-2 in efficiency and accuracy benchmarks, especially in Indian languages. Available on Hugging Face, Sarvam-1 is well-suited for practical applications and edge devices, addressing the growing need for AI in India.
Related Posts
If you like this post, please share your feedback in the comments section below so that we will upload more posts like this.
Responses