# UNQ_C1 : Error in post processing portion

Lab ID: gtizubcyvtpv

I get this error. My weights and cam are of shape (1024,).


My debug:

Weights Shape: (1024,)
cam is :::: [<tf.Tensor ‘mul_5120:0’ shape=(?, ?, ?, 1024) dtype=float32>
<tf.Tensor ‘mul_5121:0’ shape=(?, ?, ?, 1024) dtype=float32>
<tf.Tensor ‘mul_5122:0’ shape=(?, ?, ?, 1024) dtype=float32> …
<tf.Tensor ‘mul_6141:0’ shape=(?, ?, ?, 1024) dtype=float32>
<tf.Tensor ‘mul_6142:0’ shape=(?, ?, ?, 1024) dtype=float32>
<tf.Tensor ‘mul_6143:0’ shape=(?, ?, ?, 1024) dtype=float32>] and its shape is (1024,)


Program Output

TypeError Traceback (most recent call last)
in ()
2 ### this cell is NOT editable
3 reference = np.load(‘reference_cam.npy’)
----> 4 cam = grad_cam_test(grad_cam, im_path, mean, std, load_image_normalize, model, reference)

~/work/W3A1/public_tests.py in grad_cam_test(target, im_path, mean, std, load_image_normalize, model, reference)
11 im = grad_cam_test_case(target, im_path, mean, std, load_image_normalize, model)
12
—> 13 cam = target(model, im, 5, ‘conv5_block16_concat’) # Mass is class 5
14
15 # Loads reference CAM to compare our implementation with.

in grad_cam(input_model, image, category_index, layer_name)
132 # We’ll take care of the postprocessing.
133 H, W = image.shape[1], image.shape[2]
→ 134 cam = np.maximum(cam, 0) # ReLU so we only get positive importance
135 cam = cv2.resize(cam, (W, H), cv2.INTER_NEAREST)
136 cam = cam / cam.max()

/opt/conda/lib/python3.6/site-packages/tensorflow/python/framework/ops.py in bool(self)
604 TypeError.
605 “”"
→ 606 raise TypeError("Using a tf.Tensor as a Python bool is not allowed. "
607 "Use if t is not None: instead of if t: to test if a "
608 "tensor is defined, and use TensorFlow ops such as "

TypeError: Using a tf.Tensor as a Python bool is not allowed. Use if t is not None: instead of if t: to test if a tensor is defined, and use TensorFlow ops such as tf.cond to execute subgraphs conditioned on the value of a tensor.

Would like to close this. It is subsumed under another post I just created.

No need for any action at this point.

– Jaidev

Thank you for letting us know @getjaidev.