dbcSep03

训练时，每一轮的时间越来越长，如何解决？

1

 重新训练后 时间减少后，又逐渐增加 谢谢解答！

sft_train

1

使用huggingface实现的sft_train.py 中有实现对应的embeeding和encoder冻结么？

Some NCCL operations have failed or timed out.

6

rank1]:[E ProcessGroupNCCL.cpp:537] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data. [rank1]:[E ProcessGroupNCCL.cpp:543] To avoid...

你好！想问一下开集检测的逻辑，是在coco上训练后，在lvis-v1的验证集上验证的吗？谢谢 有相关的配置文件么？想学习一下，谢谢