flash on Nostr: ⚡️🤖 FLASH - Today, as part of my work on AI, I managed to get Gemma 4 and Qwen ...
⚡️🤖 FLASH - Today, as part of my work on AI, I managed to get Gemma 4 and Qwen 3.X running locally on my PC with llama.cpp
The result: aside from the privacy concerns, I don’t see any point in it—it’s slow, and my PC started heating up and making a lot of noise. I opened Task Manager and saw it was using ~100% CPU and 100% GPU.
This is my first attempt, and I’m pretty proud of it. I’m going to try Ollama tomorrow and see if it’s better. The next step is to create a Hermes agent; I’ve downloaded the app, so now I need to find a tutorial.
Published at
2026-10-26 19:43:22 UTCEvent JSON
{
"id": "7bf948ced855729a29bb885b89ab8745c6570d3055dfd2ea9d0b1d770bb43de4",
"pubkey": "4d7842051782e0d3feb034d150adc2b6bae4ee3b49786793bffa468b6f5b96b3",
"created_at": 1793043802,
"kind": 1,
"tags": [
[
"imeta",
"url https://blossom.primal.net/f9c7bebc487e64574e7e3910c21be8d698bc45f6e942d7b76fc4ea43b5c724de.jpg",
"m jpeg",
"dim 1115.0x744.0"
],
[
"client",
"Primal iOS"
]
],
"content": "⚡️🤖 FLASH - Today, as part of my work on AI, I managed to get Gemma 4 and Qwen 3.X running locally on my PC with llama.cpp\n\nThe result: aside from the privacy concerns, I don’t see any point in it—it’s slow, and my PC started heating up and making a lot of noise. I opened Task Manager and saw it was using ~100% CPU and 100% GPU. \n\nThis is my first attempt, and I’m pretty proud of it. I’m going to try Ollama tomorrow and see if it’s better. The next step is to create a Hermes agent; I’ve downloaded the app, so now I need to find a tutorial. \nhttps://blossom.primal.net/f9c7bebc487e64574e7e3910c21be8d698bc45f6e942d7b76fc4ea43b5c724de.jpg",
"sig": "ae2ac25b25fb9be4e8722359d6f6d7ab5aff9b7854b2cbb4d79a2bb6eb74f203eed8b5490fdcae22327b6900361fd3fad49a14d751335fbdd6d008bcba4d78bb"
}
