szesiongteo

4 ms·

1. 1.▲
   Demo: Code generation and editing via streaming speech recognition(gist.github.com)

   1 points by szesiongteo 3y ago|1 comments
2. 2.▲
   by szesiongteo 3y ago

   I created this Proof of Concept (PoC) and video demonstrate how prompt routing can be used to make ChatGPT dynamically select the targeted prompt, achieving better results. The concept involves categorizing the initial prompt using ChatGPT
3. 3.▲
   by szesiongteo 3y ago

   It doesn't show up, I suppose OpenAI kept the custom instruction (SYSTEM role instructions) at the server-side. I was referring to how to obtain the custom instructions to make the response human-like when using ChatGPT mobile app in v
4. 4.▲
   Ask HN: ChatGPT mobile app voice chat custom instructions

   3 points by szesiongteo 3y ago|3 comments
5. 5.▲
   by szesiongteo 3y ago

   Well, the thing with Google is whenever their service does not get enough users to reach their target revenue then they can suddenly shutdown in the following year. This is the reason why I would always go with AWS or Azure and GCP the last
6. 6.▲
   by szesiongteo 3y ago

   I think the cost calculation here does not reflect the actual scenario where most people face. In real world scenario, we don't get inputs queued up to millions and wait for the GPU to inference them continuously at 100% utilization. W