尝试用GTX 1650跑本地大模型belike：

rt，显存只有4GB的GTX 1650，尝试了：

NAME |
ID |
SIZE |
MODIFIED |

huihui_ai/deepseek-r1-abliterated:1.5b |
d536977cfb9e |
1.1 GB |
3 hours ago |

huihui_ai/llama3.2-abliterate:3b |
2ffe056672b5 |
2.2 GB |
3 hours ago |

huihui_ai/qwen2.5-abliterate:3b |
7eac36b0fdba |
1.9 GB |
3 hours ago |

qwen2.5-coder:1.5b |
d7372fd82851 |
986 MB |
3 hours ago |

qwen2.5-coder:3b |
f72c60cabf62 |
1.9 GB |
3 hours ago |

其中qwen2.5出现一直重复问题，用opencode调用所有模型均有上下文丢失问题

还是没钱买新电脑，跑跑api得了（虽然api也很贵）