Jonseed

@akaGik-jit I'm also having the issue that you mentioned, where stopping generate forever doesn't stop it, and it continues using old parameters. I'm not sure what is happening, but one...

I'm having this exact issue with a lora I just trained, and Flux Q8 GGUF. If I change to Flux NF4 or FP8, it looks fine. So it seems there...

I have also noticed this on my LG washer. It stays "spinning" long after completion of a cycle. I'm using the LG ThinQ integration. I have also noticed, generally, that...

@aaronbolton you might like my new [Ollama Node](https://github.com/Jonseed/Ollama-Node) for help expanding prompts in Invoke. By the way, I have also noticed that Invoke doesn't seem to free up the memory...

Any service that supports the OpenAI Chat Completions API seems like it could work, but other services might have a different formatting. Groq [supports](https://console.groq.com/docs/text-chat) the `response_format` parameter to turn on...

I have found that simple prompting to get JSON output from LLMs is very unreliable. Small models in particular are worse at this than large models. Forcing the API to...

I think the simpler JSON Mode would be sufficient to begin with, and pretty straightforward to implement. All of the latest OpenAI models support it, and it is turned on...

> Here's an example of using the plain json_object response_format with a descriptive prompt. I think it's reasonable: I think that's great! I think that covers most needs without getting...

> Just FYI, > The "v1" endpoint would be associated with using the OpenAI API object and the corresponding menu item for a local app like Ollama would be: `Local...

> You mentioned some other json node, does it handle this? I do like to make it so users don't have to download a bunch of other nodes in order...