Skip to content
- Sarvam AI has launched a new language model for 10 Indian languages.
- The latest language model of Sarvam AI was trained with NVIDIA’s NeMo framework on the Yotta Shakti Cloud using HGX H100 systems.
- The new AI model called Sarvam-1 supports Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Oriya, Punjabi, Tamil, and Telugu languages.
- Sarvam 2B launched its first foundational AI model in August 2024.
- This new AI model has been developed with 2 billion parameters. Parameter count is used to indicate the complexity of an AI model.
- Sarvam-1 and Phi-3 Mini AI models fall under the category of small language models (SLMs).
- Sarvam-1 is more efficient in handling Indic language scripts as compared to previous LLMs by using minimal tokens per word.
- Sarvam-1 is more efficient, with inference speeds that are 4-6 times faster than larger models like Gemma-2-9B and Llama-3.1-8B.
error: Content is protected !!
