SlickGPT is a light-weight "use-your-own-API-key" web client for the OpenAI API written in Svelte. It offers GPT-4 integration, a userless share feature and other superpowers.
I'm investigating adding more models for my slickgpt fork, and found that:
If I start a new chat session with model A, then the model gets deleted in openai.ts (rather than setting the hidden: true flag), then the web frontend hangs and becomes unusable.
I'm investigating adding more models for my slickgpt fork, and found that:
If I start a new chat session with model A, then the model gets deleted in openai.ts (rather than setting the
hidden: true
flag), then the web frontend hangs and becomes unusable.To mitigate this, I proposed a commit to switch back to the default model (gpt3.5) when the model used for the chat session is deleted. https://github.com/xmoiduts/slick-zy-gpt/commit/68d8f0c373cd76b775e8219274c5b6b726936bff