灵哥教你玩转Llama3：解决NCCL缺失问题

解决NCCL缺失问题

有很多小伙伴私信灵哥: 官网的llama3模型用不了, 运行会报错

灵哥按照官网给的启动命令， 确实启动不了

torchrun --nproc_per_node 1 example_chat_completion.py \
    --ckpt_dir Meta-Llama-3-8B-Instruct/ \
    --tokenizer_path Meta-Llama-3-8B-Instruct/tokenizer.model \
    --max_seq_len 512 --max_batch_size 6

报错如下

Traceback (most recent call last):
  File "D:\AI\llama3\example_chat_completion.py", line 84, in <module>
    fire.Fire(main)
  File "D:\dev\env\python3.11\Lib\site-packages\fire\core.py", line 143, in Fire
    component_trace = _Fire(component, args, parsed_flag_args, context, name)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "D:\dev\env\python3.11\Lib\site-packages\fire\core.py", line 477, in _Fire
    component, remaining_args = _CallAndUpdateTrace(
                                ^^^^^^^^^^^^^^^^^^^^
  File "D:\dev\env\python3.11\Lib\site-packages\fire\core.py", line 693, in _C