amasky / ram

Recurrent Models of Visual Attention (RAM) with Chainer
MIT License
44 stars 8 forks source link

Does this repo only work for Grayscale Images? #9

Open Deccan12 opened 3 years ago

Deccan12 commented 3 years ago

I can run the repo for grayscale images but for colored images I am getting this error, I do not understand what I am doing wrong?

Loading Data Shape (12271, 100, 100, 3) First time reading data, saving to pickle to speed up next time test: 33%|██████████████████████████████████████▍ | 1000/3068 [00:00<00:00, 2205.01it/s] Traceback (most recent call last): File "train.py", line 310, in loss, acc = test(test_data, test_targets) File "train.py", line 302, in test model(x[i:i+batchsize], t[i:i+batchsize], train=False) File "C:\ram-master\ram.py", line 63, in call self.forward(x, train, action=False, init_l=l) File "C:\ram-master\ram.py", line 128, in forward hg = F.relu(self.emb_x(rho)) File "C:\envs\RAMCHAINER1\lib\site-packages\chainer\links\connection\linear.py", line 93, in call return linear.linear(x, self.W, self.b) File "C:\envs\RAMCHAINER1\lib\site-packages\chainer\functions\connection\linear.py", line 107, in linear return LinearFunction()(x, W, b) File "C:\envs\RAMCHAINER1\lib\site-packages\chainer\function.py", line 189, in call self._check_data_type_forward(in_data) File "C:\envs\RAMCHAINER1\lib\site-packages\chainer\function.py", line 273, in _check_data_type_forward self.check_type_forward(in_type) File "C:\envs\RAMCHAINER1\lib\site-packages\chainer\functions\connection\linear.py", line 23, in check_type_forward type_check.prod(x_type.shape[1:]) == w_type.shape[1], File "C:\envs\RAMCHAINER1\lib\site-packages\chainer\utils\type_check.py", line 487, in expect expr.expect() File "C:\envs\RAMCHAINER1\lib\site-packages\chainer\utils\type_check.py", line 449, in expect '{0} {1} {2}'.format(left, self.inv, right)) chainer.utils.type_check.InvalidType: Invalid operation is performed in: LinearFunction (Forward)

Expect: prod(in_types[0].shape[1:]) == in_types[1].shape[1] Actual: 576 != 192