Docs changes to BucketSentenceIter, Optimizer, ImageIter #6112
Conversation
python/mxnet/image.py
Outdated
| label_width : int | ||
| Label dimension. | ||
| label_width : int, optional | ||
| Label dimension. The default label width is 1. |
There was a problem hiding this comment.
Label dimension (number of labels per example)
There was a problem hiding this comment.
Does this mean the number of classes? Perhaps we can be clearer? Right now we say "Label dimension" and "label width" ostensibly to refer to the same quantity. Let's use consistent notation.
There was a problem hiding this comment.
its not number of classes, which is the range of label. Its the number of labels.
python/mxnet/optimizer.py
Outdated
|
|
||
| Parameters | ||
| ---------- | ||
| index : int | ||
| An unique index to identify the weight. | ||
| A unique index to identify the parameter. |
There was a problem hiding this comment.
explain relationship to set_lr_mult and set_wd_mult
python/mxnet/image.py
Outdated
| label_width : int | ||
| Label dimension. | ||
| label_width : int, optional | ||
| Label dimension. The default label width is 1. |
There was a problem hiding this comment.
Does this mean the number of classes? Perhaps we can be clearer? Right now we say "Label dimension" and "label width" ostensibly to refer to the same quantity. Let's use consistent notation.
python/mxnet/image.py
Outdated
| @@ -462,6 +462,7 @@ def __init__(self, batch_size, data_shape, label_width=1, | |||
| self.reset() | |||
|
|
|||
| def reset(self): | |||
| """Reset the iterator to the beginning of the data.""" | |||
python/mxnet/image.py
Outdated
| @@ -493,6 +494,7 @@ def next_sample(self): | |||
| return header.label, img | |||
|
|
|||
| def next(self): | |||
| """Return the next batch of data.""" | |||
python/mxnet/optimizer.py
Outdated
| @@ -154,18 +154,18 @@ def create_state(self, index, weight): | |||
| """ | |||
|
|
|||
| def update(self, index, weight, grad, state): | |||
| """Update the weight given the corresponding gradient and state. | |||
| """Update the parameters given the corresponding gradients and state. | |||
There was a problem hiding this comment.
Is this updating a single parameter tensor or multiple? If a single one let's just say: "Updates the given parameter using the corresponding gradient and state."
python/mxnet/rnn/io.py
Outdated
| @@ -135,6 +135,7 @@ def __init__(self, sentences, batch_size, buckets=None, invalid_label=-1, | |||
| self.reset() | |||
|
|
|||
| def reset(self): | |||
| """Reset the iterator to the beginning of the data.""" | |||
python/mxnet/rnn/io.py
Outdated
| @@ -150,6 +151,7 @@ def reset(self): | |||
| self.ndlabel.append(ndarray.array(label, dtype=self.dtype)) | |||
|
|
|||
| def next(self): | |||
| """Return the next batch of data.""" | |||
|
Hey can we get these changes made so we can merge in? |
|
@zackchase @piiswrong updated with suggested changes |
|
@piiswrong can we merge?
…On May 10, 2017 10:56 AM, "lxn2" ***@***.***> wrote:
@zackchase <https://github.com/zackchase> @piiswrong
<https://github.com/piiswrong> updated with suggested changes
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#6112 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/ACR4znwhdzAUVhpex--xbrYq4bjdMquiks5r4fpTgaJpZM4NRW06>
.
|
* Docs changes to BucketSentenceIter, Optimizer, ImageIter * Change verbiage and correct details
* Docs changes to BucketSentenceIter, Optimizer, ImageIter * Change verbiage and correct details
No description provided.