Closed wmx-github closed 2 months ago
我huggingface 模型 Qwen1.5-0.5B-Chat 生成 qwen.rkllm push到板子上: `orangepi@orangepi5b:~/workspace/llm$ ls
librkllmrt.so llm_demo qwen.rkllm 然后 export LD_LIBRARY_PATH=./
然后
orangepi@orangepi5b:~/workspace/llm$ ./llm_demo ./qwen.rkllm
rkllm init start E RKNN: [00:03:36.400] failed to convert handle(1020) to fd, ret: -1, errno: 24, errstr: Too many open files Segmentation fault (core dumped)
` 怎么解决
In terminal ulimit -n 10096
我huggingface 模型 Qwen1.5-0.5B-Chat 生成 qwen.rkllm push到板子上: `orangepi@orangepi5b:~/workspace/llm$ ls
librkllmrt.so llm_demo qwen.rkllm
然后
export LD_LIBRARY_PATH=./orangepi@orangepi5b:~/workspace/llm$ ./llm_demo ./qwen.rkllm
rkllm init start E RKNN: [00:03:36.400] failed to convert handle(1020) to fd, ret: -1, errno: 24, errstr: Too many open files Segmentation fault (core dumped)
` 怎么解决