#hide
! [ -e /content ] && pip install -Uqq fastbook
import fastbook
fastbook.setup_book()

#hide
from fastbook import *

path = untar_data(URLs.PETS)/'images'
def is_cat(x): return x[0].isupper()
dls = ImageDataLoaders.from_name_func(
    path, get_image_files(path), valid_pct=0.2, seed=21,
    label_func=is_cat, item_tfms=Resize(224))
learn = vision_learner(dls, resnet34, metrics=error_rate)
learn.fine_tune(1)

img = PILImage.create(image_cat())
x, = first(dls.test_dl([img]))

class Hook():
    def hook_func(self, m, i, o): self.stored = o.detach().clone()

hook_output = Hook()
hook = learn.model[0].register_forward_hook(hook_output.hook_func)

with torch.no_grad(): output = learn.model.eval()(x)

act = hook_output.stored[0]

F.softmax(output, dim=-1)

dls.vocab

x.shape

cam_map = torch.einsum('ck,kij->cij', learn.model[1][-1].weight, act)
cam_map.shape

x_dec = TensorImage(dls.train.decode((x,))[0][0])
_,ax = plt.subplots()
x_dec.show(ctx=ax)
ax.imshow(cam_map[1].detach().cpu(), alpha=0.6, extent=(0,224,224,0),
              interpolation='bilinear', cmap='magma');

hook.remove()

class Hook():
    def __init__(self, m):
        self.hook = m.register_forward_hook(self.hook_func)   
    def hook_func(self, m, i, o): self.stored = o.detach().clone()
    def __enter__(self, *args): return self
    def __exit__(self, *args): self.hook.remove()

with Hook(learn.model[0]) as hook:
    with torch.no_grad(): output = learn.model.eval()(x.cuda())
    act = hook.stored

class HookBwd():
    def __init__(self, m):
        self.hook = m.register_backward_hook(self.hook_func)   
    def hook_func(self, m, gi, go): self.stored = go[0].detach().clone()
    def __enter__(self, *args): return self
    def __exit__(self, *args): self.hook.remove()

cls = 1
with HookBwd(learn.model[0]) as hookg:
    with Hook(learn.model[0]) as hook:
        output = learn.model.eval()(x.cuda())
        act = hook.stored
    output[0,cls].backward()
    grad = hookg.stored

w = grad[0].mean(dim=[1,2], keepdim=True)
cam_map = (w * act[0]).sum(0)

_,ax = plt.subplots()
x_dec.show(ctx=ax)
ax.imshow(cam_map.detach().cpu(), alpha=0.6, extent=(0,224,224,0),
              interpolation='bilinear', cmap='magma');

with HookBwd(learn.model[0][-2]) as hookg:
    with Hook(learn.model[0][-2]) as hook:
        output = learn.model.eval()(x.cuda())
        act = hook.stored
    output[0,cls].backward()
    grad = hookg.stored

w = grad[0].mean(dim=[1,2], keepdim=True)
cam_map = (w * act[0]).sum(0)

_,ax = plt.subplots()
x_dec.show(ctx=ax)
ax.imshow(cam_map.detach().cpu(), alpha=0.6, extent=(0,224,224,0),
              interpolation='bilinear', cmap='magma');