ChChwang

RPN的loss 和最后的loss是2部分。怎么训练的呢，end-2-end 将2部分的4个loss相加训练？还是 two step 交替训练rpn和最后的分类回归？ why __C.TRAIN.HAS_RPN=False?

the datasets are too big to be downloaded

3

the datasets are too big to be downloaded

batch_size 无论设置多大，都会OOM是什么原因呢

3

ResourceExhaustedError (see above for traceback): OOM when allocating tensor with shape[12928,768] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc If you want to see a list of allocated tensors when...

bert部分的参数是否finetune训练了？

1

训练的时候，bert部分的参数是否也finetune训练了？ 测试的时候，是用的原始的bert输出向量。


tf 1.3.0 cannot find -ltensorflow_framework

6

cannot find -ltensorflow_framework ?

create_synthtext_tfrecord.py error

3

NameError: global name '_create_samples_of_an_image' is not defined the function is right?

how to run batch inference

4

how to run batch inference

Does anyone achieve distributed training? multi machine with multi gpus. how to set TF_CONFIG? thanks