sugarforever/chat-ollama

Llama3.1:405b 对话 不工作

caih1943 opened this issue · 5 comments

caih1943 commented

Llama3.1:405b 对话 不工作 ：
Ollama call failed with status code 500: llama runner process has terminated: error loading model: unable to allocate backend buffer

satrong commented

你通过 ollama 提供的终端命令能进行对话吗？

caih1943 commented

我刚试了一下，同样Error: llama runner process has terminated: error loading model: unable to allocate backend buffer

satrong commented

所以这与 chatollama 无关了。会不会是你的模型太大了，内存不够用？

caihengsheng commented

32Gb RAM also faced this problem, unable to run it

satrong commented

32GB 内存带不动 405b 吧？