Yiman-GO

shape not found when finetune

1

I got an error when I use . ./experiments/scripts/train_faster_rcnn.sh 0 pascal_voc res101 USE_GPU_NMS False . The error is as follows: Traceback (most recent call last): File "./tools/trainval_net.py", line 110, in...

is there anyone have successfully load the pretrained model in code?

7

is there anyone have successfully load the pretrained model in code? I have no idea about it. I have tried : saver1 = tf.train.import_meta_graph(pretrained_meta) saver1.restore(sess, pretrained_ckpt) but it tells me...

The MACs result doesn't contain q@k and attn@V

1

The calculation workload of matrix multiplication in the attention modules(q*k and attn*V) which is nonnegligible, is not included in torch_pruning.utils.count_ops_and_params。 torchprofile can calculate the macs of q@k and k@v.