Trending topics
#
Bonk Eco continues to show strength amid $USELESS rally
#
Pump.fun to raise $1B token sale, traders speculating on airdrop
#
Boop.Fun leading the way with a new launchpad on Solana.
The local model solved the problem in just 2.5 seconds. (Isn't a 5-minute window a bit too long?) A 5-minute window for one question is more than enough. You guys are still using such an expensive model API in the cloud and paying for server costs, so if you can deploy locally, do it. The electricity bill adds up to over 100 a month, which is about the same as using your computer daily. @agentcoinsite
Feb 18, 2026
Sharing some experience on local deployment of @agentcoinsite mining, to help friends who want to get on board avoid pitfalls 👇
Hardware: RTX 4090 (24GB) Model: Qwen2.5:32B (Q4_K_M quantization, occupying ~20GB of VRAM) Tools: Ollama + Python script
A few key points: 1️⃣ The model must be the 4bit quantized version, otherwise the VRAM won't be enough 2️⃣ Set OLLAMA_KEEP_ALIVE=-1 to keep the model in VRAM, otherwise it will need to reload for each request 3️⃣ The new version of the OpenAI Python library has compatibility issues with Ollama, it's recommended to directly use requests to call the native API
Now it's running automatically, and the electricity cost is exchanged for AGC, which saves a lot compared to cloud API 🚀
#AgentCoin #AGC #LocalLLM #RTX4090
@agentcoinsite 2026-02-18 03:02:19 [ERROR] Failed to get Agent status: 503 Server Error: Service Unavailable for url:
@agentcoinsite Can you do some maintenance?
1.99K
Top
Ranking
Favorites