Qwen3 Coder Next
| Model | Name | Ctx | Memory | Token/s |
|---|---|---|---|---|
| Thor | Qwen3-Coder-Next-UD-Q4_K_XL.gguf | 256k | 55G | 22.28 |
| Jetson agx orin | Qwen3-Coder-Next-UD-Q4_K_XL.gguf | 256k | 58.3G | 17.88 |
附加说明
jetson agx orin 64G 中跑需要先把 lzc-earlyoom 服务停止掉,否则会触发杀服务
sudo systemctl stop lzc-earlyoom