Hi, there is something I might consider it to be wrong. When doing the post-pretraining, the command is:
CUDA_VISIBLE_DEVICES= python main.py --config cfgs//post_pretrain.yaml --exp_name --finetune_model
Why there is no need to use the --ckpts parameter, does that mean I'm training a total new model from the beginning?
Hope you can help me to figure it out.
@CGuangyan-BIT
Hi, there is something I might consider it to be wrong. When doing the post-pretraining, the command is: CUDA_VISIBLE_DEVICES= python main.py --config cfgs//post_pretrain.yaml --exp_name --finetune_model
Why there is no need to use the --ckpts parameter, does that mean I'm training a total new model from the beginning?
Hope you can help me to figure it out.
@CGuangyan-BIT