Working on a government phone.
There seems to be a big push to sell subscriptions for imaginarium and produce ai generated images. So I asked it to make photos from pictures of me and my girlfriend from my phone. I didn't realize that I needed to change the permissions for termux to access them. I also don't know what libraries or drivers the ai would use to do it.
I'm using ggml-org/llama.cpp compiled on termux. The llama-server tried to work but it didn't post readable text. Other than the terrible redundant response loops it actually works better than llama3 on my phone! Cheers 🥂🍻 I'm planning on working on a small nerdy data set and this is a great bootstrap. Id like to know what version of opencv and other libraries I would need for this model to fulfill what it talks about.