29 ms·

Open models like Llama make it pointless for the majority of companies to train from scratch. It was obvious this would happen.

• bjornsing 2y ago
  True. The hard part is timing it.
• 7734128 2y ago
  Inference should always be more significant than training in the end though.
  • Tepix 2y ago
    There are more options for inference.