Sidenoun

Founder. Developer. Artist.

[Sidenoun @sidenoun]

It's built different.

Build Your Noun 👉 https://t.co/rDqFpKzaid

[Sidenoun @sidenoun]

Does anyone know the o1 rate limits for @GitHub — I see it as “custom”, but wondered where I can find that information.

[Sidenoun @sidenoun]

@lgrammel @github It’s Azure, but with GitHub models endpoint. However, there’s no model in that endpoint that comes before /chat or /chat/completions. I’ve made my own provider using the Azure example and hardcoded URL, but wondered whether there was an easier options to assign the whole URL.

[Sidenoun @sidenoun]

Hi @lgrammel

Is it possible to have the option to adjust the URL for a provider? Not talking ${options.baseURL}, but more removing things like ${modelId} in ... `${options.baseURL}/${modelId}${path}?...`

Trying to implement @GitHub models, but it doesn't require the ${modelId}

[Sidenoun @sidenoun]

@hc_dsn @midjourney Love it, very useful.
Heads up, a typo on the —no description pill. 👍

[Sidenoun @sidenoun]

@OpenAIDevs Are there any plans for Context Caching?

[Sidenoun @sidenoun]

Just waiting for @perplexity_ai to add @OpenAI o1 🤗

[Sidenoun @sidenoun]

Are you able to use fine tuned models in Gemini with context caching? @OfficialLoganK 🤔

[Sidenoun @sidenoun]

@simonw @OpenAI I guess with cheaper API token pricing, it is somewhat possible? Probably why there’s a race to the bottom.

[Sidenoun @sidenoun]

@OfficialLoganK When? 👀

[Sidenoun @sidenoun]

@jxmnop The race to zero for API pricing’s has also been a contributing factor.

[Sidenoun @sidenoun]

@OfficialLoganK Will the new experimental models have a lower threshold for the required context caching tokens input? 👀

[Sidenoun @sidenoun]

Hi @OfficialLoganK 👋

Can we get a reduced Gemini storage cost for Context Caching and also a reduction in the minimum input token count, which is currently at 32,768

[Sidenoun @sidenoun]

I’ve got notifications turned on for @OfficialLoganK — the man, the myth, the legend.

[Sidenoun @sidenoun]

For sure if @wickedmishra makes a component kit of framer motions etc, most would pay for it 👌

[Sidenoun @sidenoun]

@punk4156 Man, @benbodhi is a legend 🙌

[Sidenoun @sidenoun]

@OfficialLoganK Any luck? 👀

[Sidenoun @sidenoun]

@OfficialLoganK Hero, thanks!

[Sidenoun @sidenoun]

Happy Nouniversary, frens! 🥳🎉

[Sidenoun @sidenoun]

@googledevs Planning to add support for @Ollama also, that’s cool.

[ollama @ollama]

@sidenoun ❤️ time flies when you're enjoying it! Ollama is just getting started. 🫡