Listen to this story
|
Ola Krutrim, the AI initiative by Ola, is set to launch its standalone mobile app, as announced by chief Bhavish Aggarwal.
The app comes with significant advancements, including a notable decrease in the Time it Takes to Generate the First word of a response (TTFT), from 22 seconds at its initial launch to a swift 0.3 seconds presently, with further enhancements anticipated. Aggarwal also hinted at forthcoming improvements in an upcoming detailed blog post.
Aggarwal recently revealed that Krutrim has achieved a major breakthrough by operating on its independent cloud infrastructure, signifying a move away from reliance on external cloud providers like AWS or Azure. He emphasised ongoing efforts by the Krutrim team to enhance both the model itself and its infrastructure.
Intel recently disclosed that Ola Krutrim is leveraging Intel Gaudi 2 clusters for pre-training and fine-tuning its foundational models, boasting industry-leading price/performance ratios across ten languages.
Moreover, Krutrim is actively pre-training an expanded foundational model on Intel Gaudi 2 clusters, further elevating its AI capabilities.
A few days ago, Krutrim announced its partnership with Databricks to improve its foundational language model, particularly for Indian languages, aiming to enhance AI solutions in India.
“The Krutrim model was launched using our platform,” said Naveen Rao, VP of generative AI at Databricks, during an exclusive interview with AIM.
Ola Krutrim has been quite obsessed with developing its own foundational model from scratch, despite rumours that it is being built on fine-tuned models such as Llama-2, Mistral, Claude-3 or even the most recent, DBRX.
Launched in December last year, Krutrim has been lauded as “India’s first full-stack AI” solution, showcasing prowess in understanding and generating content across multiple Indian languages, including Marathi, Hindi, Bengali, Tamil, Kannada, Telugu, Odia, Gujarati, and Malayalam, with claims of superiority over GPT-4 in Indic languages