JA: add images.ipynb, tf-records.ipynb translated in site/ja/tutorials/load_data #427
Conversation
|
Check out this pull request on ReviewNB: https://app.reviewnb.com/tensorflow/docs/pull/427 Visit www.reviewnb.com to know how we simplify your Jupyter Notebook workflows. |
lamberta
added
the
translation
|
Thank you, Masatoshi! @ohtaman, @taquo, @sfujiwara If you get a chance, please take a look :) For future work, I would suggest looking at the TF 2.0 Alpha docs: https://www.tensorflow.org/alpha They are still a work-in-progress, but it is the way forward when TF 2.0 stable is released. And there is overlap with the TF 1.x tutorials. Thank you! |
|
Thank you, Masatoshi. @lamberta I can review these *.ipynb next week (4/1~). |
|
Thanks. |
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
There was a problem hiding this comment.
I checked that this is fixed.
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
- This cell does not exists in original notebook.
- Should remove outputs. (This contains your local path)
Reply via ReviewNB
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
There was a problem hiding this comment.
Well. Adopt changes.
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
all_image_labelsやall_image_pathsの配列がある場合
Maybe: all_image_labelsやall_image_pathsのような配列がある場合
Reply via ReviewNB
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
- Very weak suggestion: "同じサイズに設定しています。" is better than "同じサイズに設定します。" Because, to set buffer_size the same size as the dataset is good in this case but not in other cases with larger datasets.
- In last paragraph, "を" are repeated.
- Suggestion: "tf.data.Dataset.apply メソッドを、...と組み合わせることで対処することができます。"
Reply via ReviewNB
There was a problem hiding this comment.
Good suggestions. Thanks.
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
| @@ -0,0 +1,3254 @@ | |||
| { | |||
There was a problem hiding this comment.
|
I reviewed images.ipynb.
|
|
Additionally, I changed the word "イメージ" to ”画像" and "ロード" to "読み込み". |
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
Suggestions:
- "キャッシュする" may be frequently used than "キャッシングする"
- "ファイルで表されます" should be "ファイルで定義されます" because the translation of "define" is "定義"
- Should use same word for the ranslation of "message type". "メッセージの型" and "メッセージタイプ" are appeared.
- I prefer "メッセージ型"
Reply via ReviewNB
There was a problem hiding this comment.
Thank you. Adopted your suggestion.
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
Weak suggestion:
- You can use "パース" as a translation of "parse". It is used in general enough.
Reply via ReviewNB
There was a problem hiding this comment.
Yes. Use "パース" for "parse".
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
The second paragraph is a little bit confusing.
この他の一般的なデータ型のほとんどは、強制的にこれらのうちの1つにすること可能です。
How about: (use the word "変換する")
その他の一般的なデータ型の多くは、これらの型のいずれかに強制的に変換することができます。
or, maybe it is better to remove "その他の" (translation of "other") for ease of understanding.
Reply via ReviewNB
There was a problem hiding this comment.
Thanks. Good suggestion.
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
I prefer "メッセージはすべて" or "すべてのメッセージ" to "主要なメッセージ" since the translation of "All" is "すべて" in general.
Reply via ReviewNB
There was a problem hiding this comment.
You are right. Changed.
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
各値は上記の関数を使って3種類の互換性のある型をからなる
Typo?: 型をからなる
Maybe: 型のうちの1つだけを含む
#2で作成したマップを特徴量メッセージに変換します。
"Features メッセージ" seems better than "特徴量メッセージ" because
Features message is defined in "https://github.com/tensorflow/tensorflow/blob/master/tensorflow/core/example/feature.proto#L85".
Reply via ReviewNB
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
This paragraph seems to be changed. In original text,
a boolean feature, False or True with equal probability
an integer feature uniformly randomly chosen from [0, 5)
a string feature generated from a string table by using the integer feature as an index
a float feature from a standard normal distribution
Translation example:
FalseまたはTrueを表す論理値。出現確率は等しいものとします。[0, 5)の範囲から一様にサンプリングした整数値。- 整数特徴量をインデックスとした文字列テーブルを使って生成した文字列特徴量
- 標準正規分布からサンプリングした浮動小数点数。
Reply via ReviewNB
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
# 整数特徴量 -10000 から 10000 の間の乱数
feature1 = np.random.randint(0, 5, n_observations)
This comment does not match to the code.
# バイト特徴量
strings = np.array([b'cat', b'dog', b'chicken', b'horse', b'goat'])
"文字列特徴量" or "バイト文字列特徴量" may be better than "バイト特徴量" since the original text is "string feature".
Reply via ReviewNB
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
The original document seems be changed:
- The "## TFRecords Format Details" section was added before this section.
Reply via ReviewNB
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
FYI: This section is modified in r2 version (and so I compared only with original text)
Reply via ReviewNB
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
Very weak suggestion:
The translation of "image string" , "画像文字列", may be a bit confusing.
May be "生画像のデータ列" or just "生画像データ" is better than it.
Reply via ReviewNB
| @@ -0,0 +1,1687 @@ | |||
| { | |||
There was a problem hiding this comment.
|
@masa-ita I finished adding comments to tf-records.ipynb now. It seems that you updated "tf_records.ipynb" instead of "tf_recordes.ipynb" in response to my comments I gave before. |
lamberta
force-pushed
the
site_ja_tutorials_load_data
branch
from
3f4e900 to
1f01b90
Compare
PiperOrigin-RevId: 244702159
…d_data PiperOrigin-RevId: 244702159
images.ipynb and tf-records.ipynb in site/ja/tutorials/load_data are translated in Japanese.