Open curran opened 1 year ago
Idea: hitting the Escape key during generation can send another request to the server to cancel the ongoing generation. We may need to generate an ID to specify which ongoing generation to terminate.
This may not seem like a big deal with GPT3.5 Turbo, as it's so insanely fast, but as soon as we switch to the slower (but better quality) GPT4, this could be a serious usability issue.
From https://github.com/openai/openai-node#streaming-responses
If you need to cancel a stream, you can break from the loop or call stream.controller.abort().
Some thoughts:
Done in #316
Reopening this as it's not currently functioning. I think we took it out at some point.
As a user leveraging the AI assist feature, when I can see that the AI is going in the wrong direction or has provided enough output, I want to be able to stop the generation, so that I don't need to wait for it to complete and can seamlessly continue working.