yolov3 yolov4 channel and layer pruning, Knowledge Distillation 层剪枝，通道剪枝，知识蒸馏
Apache License 2.0
1.5k
stars
446
forks
source link
Traceback (most recent call last): File "slim_prune.py", line 153, in <module> prune_and_eval(model, CBL_idx, CBLidx2mask) File "slim_prune.py", line 145, in prune_and_eval bn_module.weight.data.mul_(mask) RuntimeError: output with shape [256] doesn't match the broadcast shape [1, 256] #111
merge the mask of layers connected to shortcut!
Traceback (most recent call last):
File "slim_prune.py", line 153, in
prune_and_eval(model, CBL_idx, CBLidx2mask)
File "slim_prune.py", line 145, in prune_and_eval
bnmodule.weight.data.mul(mask)
RuntimeError: output with shape [256] doesn't match the broadcast shape [1, 256]
zbyuan
commented
3 years ago
tanluren大佬，您好，我修改了yolov3的cfg文件，加入了SE注意力机制，并且稀疏训练成功。 但在slim_prune.py剪枝过程中，出现了以下错误：
merge the mask of layers connected to shortcut! Traceback (most recent call last): File "slim_prune.py", line 153, in
prune_and_eval(model, CBL_idx, CBLidx2mask)
File "slim_prune.py", line 145, in prune_and_eval
bnmodule.weight.data.mul(mask)
RuntimeError: output with shape [256] doesn't match the broadcast shape [1, 256]
请问大佬，您知道这是原因吗？