Do you think indian ecosystem allows for existence of a foundational model in text-to-image space?
What if its possible to build an image generation model which can be about 8x cheaper to be built and can be about 30% better than StableDiffusion or Midjourney in terms of output coherence, image quality etc
Fascinating!
Do you think indian ecosystem allows for existence of a foundational model in text-to-image space?
What if its possible to build an image generation model which can be about 8x cheaper to be built and can be about 30% better than StableDiffusion or Midjourney in terms of output coherence, image quality etc