CPM Bee 微调时设置 half 出现 CUDA 报错，不设置 half 则 assert 报错

CPM 使用微调脚本训练， 不开启 --use-delta 这一选项，则出现如下错误：
Traceback (most recent call last):
File "finetune_cpm_bee.py", line 503, in
main()
File "finetune_cpm_bee.py", line 499, in main
finetune(args, tokenizer, model, optimizer, lr_scheduler, optim_manager)
File "finetune_cpm_bee.py", line 364, in finetune
optim_manager.step()
File "/ms_test2/miniconda3/envs/ms1.11/lib/python3.7/site-packages/bmtrain/optim/optim_manager.py", line 131, in step
optimizer.step(scale=self.loss_scale)
File "/ms_test2/miniconda3/envs/ms1.11/lib/python3.7/site-packages/torch/optim/optimizer.py", line 109, in wrapper
return func(*args, **kwargs)
File "/ms_test2/miniconda3/envs/ms1.11/lib/python3.7/site-packages/torch/autograd/grad_mode.py", line 27, in decorate_context
return func(*args, **kwargs)
File "/ms_test2/miniconda3/envs/ms1.11/lib/python3.7/site-packages/bmtrain/optim/adam_offload.py", line 77, in step
state["_grad_fp16"] = torch.empty(p.size(), dtype=torch.float16, pin_memory=True) # on host
RuntimeError: CUDA error: OS call failed or operation not supported on this OS
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.

Start with finetune_cpm_bee.py, especially main and finetune around the reported optimizer step, and reproduce the run with and without half and --use-delta. Trace whether the failures come from the finetuning path or the bmtrain optimizer, and consider the issue complete when both reported configurations run without the CUDA or assertion errors.