HERIUN
diff = y_pred - Y # Back prop (chain rule) d_l2 = diff * sigmoid_prime(l2) 1. 위 부분에서 diff를 그대로 쓰면 안될거 같습니다. diff가 loss를 의미하는데, 저 자리에는 diff가 아니라...
PyTorch also has F.nll_loss() function that computes the negative **loss** likelihood. = > PyTorch also has F.nll_loss() function that computes the negative **log** likelihood.
in voc.py def random_scale(self, img, boxes): WHY img = cv2.resize(img, dsize=(int(w * scale), **h**), interpolation=cv2.INTER_LINEAR) I suggest img = cv2.resize(img, dsize=(int(w * scale), **int(h * scale)**), interpolation=cv2.INTER_LINEAR)
W&B logging TypeError: list indices must be integers or slices, not float error.
in past there are japanese model existed.. reupload~ https://drive.google.com/file/d/1KQINqyobor1OFpke6VrlG2MCHL7V-bV8/view?usp=sharing but not good as i expected...
1. coords are not in [0,1]^2 square.. just coords/img_size(=1024) for normalization(0~1). it is ok? https://github.com/ymy-k/Hi-SAM/blob/e1ecd86dfcd49202dc369a52e7f1c138d40f1ec4/hi_sam/modeling/prompt_encoder.py#L189 2. in the paper (section 3.5) "After the final token-to-image attention, we slice the last...
edit vgg16_bn.py. forward() *because torch jit does not supply "named tuple" ```python def forward(self, X): h = self.slice1(X) h_relu2_2 = h h = self.slice2(h) h_relu3_2 = h h = self.slice3(h)...