deep-learning-with-python-notebooks
deep-learning-with-python-notebooks copied to clipboard
fchollet
Jupyter notebooks for the code samples of the book "Deep Learning with Python"
```cpp I don't think I can do it. [start] no creo que pueda hacerlo [end] - We dance together. [start] asumí que [UNK] [end] - Tom has a 13-year-old daughter....
Should the action of relu on final image of the transformed letter k be only in the top right quarantine? I'd expect the component wise relu to kill the part...
the two "source" paramters should be "key" and "value" repectively, in function of multi_head_attention in the code snippet below ,in chapter 15 ``` query_dense = [layers.Dense(head_dim) for i in range(num_heads)]...
Several typo references to "TensorFlow" style of relu when discussing PyTorch ReLU; ie: torch.nn.relu(... ... rather than: torch.nn.ReLU(... ... eg in Chapter 3: p77, p81
The text says: ``` Tensor product The tensor product, also called dot product or matmul (short for “matrix multiplication”) is one of the most common, most useful tensor operations. In...
Hello, I don't know if it's an error, or if i missunderstand, but in training_step method ```py loss = mean_squared_error(predictions, targets) ``` And loss method take params in wrong (i...
The calculation should use the projected vectors? ```py def multi_head_attention(query, key, value): head_outputs = [] for i in range(num_heads): query = query_dense[i](query) key = key_dense[i](key) value = value_dense[i](value) scores =...
In the about the book section, in the second to last sentence of the second paragraph, there is a typo with "We are make the entire book available to read...
