Datasets:

ArXiv:

Want up to 10,000 FREE A40 hours to train models with? #GPUPOOR contest

#2
by rmonsurate - opened

Let me know how you plan to use this dataset to beat SOTA in any one benchmark. Let me know why you deserve the free compute hours. Specifically:

  1. What has your team built before?
  2. How will you leverage Red Pajama (what's your plan and how many ISOFLOPS do you estimate you'll need)
  3. Share your Github or Linkedin profile

Sign up or log in to comment