Skip to content

Commit

Permalink
[Docathon][Fix COPY-FROM No.6-10] (PaddlePaddle#6311)
Browse files Browse the repository at this point in the history
* [Docathon][Fix COPY-FROM No.6-10]

* [Docathon][Fix COPY-FROM No.6-10]
  • Loading branch information
Kaedeharai authored Nov 22, 2023
1 parent c9d92cd commit d8dca18
Show file tree
Hide file tree
Showing 5 changed files with 8 additions and 36 deletions.
28 changes: 0 additions & 28 deletions docs/api/paddle/distributed/InMemoryDataset_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -47,34 +47,6 @@ None。

COPY-FROM: paddle.distributed.InMemoryDataset.init

_init_distributed_settings(**kwargs)
'''''''''
**注意:**

**1. 该 API 只在非** `Dygraph <../../user_guides/howto/dygraph/DyGraph.html>`_ **模式下生效**
**2. 本 api 需要在机大规模参数服务器训练下生效,敬请期待详细使用文档**

对 InMemoryDataset 的实例进行分布式训练相关配置的初始化。

**参数**

- **kwargs** - 可选的关键字参数,由调用者提供,目前支持以下关键字配置。
- **merge_size** (int) - 通过样本 id 来设置合并,相同 id 的样本将会在 shuffle 之后进行合并,你应该在一个 data 生成器里面解析样本 id。merge_size 表示合并的最小数量,默认值为-1,表示不做合并。
- **parse_ins_id** (bool) - 是否需要解析每条样的 id,默认值为 False。
- **parse_content** (bool) - 是否需要解析每条样本的 content,默认值为 False。
- **fleet_send_batch_size** (int) - 设置发送 batch 的大小,默认值为 1024。
- **fleet_send_sleep_seconds** (int) - 设置发送 batch 后的睡眠时间,默认值为 0。
- **fea_eval** (bool) - 设置特征打乱特征验证模式,来修正特征级别的重要性,特征打乱需要 ``fea_eval`` 被设置为 True。默认值为 False。
- **candidate_size** (int) - 特征打乱特征验证模式下,用于随机化特征的候选池大小。默认值为 10000。

**返回**
None。


**代码示例**

COPY-FROM: paddle.distributed.InMemoryDataset._init_distributed_settings

update_settings(**kwargs)
'''''''''
Expand Down
4 changes: 2 additions & 2 deletions docs/api/paddle/text/Movielens_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@
Movielens
-------------------------------

.. py:class:: paddle.text.datasets.Movielens()
.. py:class:: paddle.text.Movielens(data_file = None, mode = 'train', test_ratio = 0.1, rand_seed = 0, download = True)
该类是对 `Movielens 1-M <https://grouplens.org/datasets/movielens/1m/>`_
Expand All @@ -25,4 +25,4 @@ Movielens
代码示例
:::::::::

COPY-FROM: paddle.text.datasets.Movielens
COPY-FROM: paddle.text.Movielens
4 changes: 2 additions & 2 deletions docs/api/paddle/text/UCIHousing_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@
UCIHousing
-------------------------------

.. py:class:: paddle.text.datasets.UCIHousing()
.. py:class:: paddle.text.UCIHousing(data_file = None, mode = 'train', download = True)
该类是对 `UCI housing <https://archive.ics.uci.edu/ml/datasets/Housing>`_
Expand All @@ -23,4 +23,4 @@ UCIHousing
代码示例
:::::::::

COPY-FROM: paddle.text.datasets.UCIHousing
COPY-FROM: paddle.text.UCIHousing
4 changes: 2 additions & 2 deletions docs/api/paddle/text/WMT14_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@
WMT14
-------------------------------

.. py:class:: paddle.text.datasets.WMT14()
.. py:class:: paddle.text.WMT14(data_file = None, mode = 'train', dict_size = -1, download = True)
该类是对 `WMT14 <http://www.statmt.org/wmt14/>`_ 测试数据集实现。
Expand Down Expand Up @@ -32,4 +32,4 @@ http://paddlemodels.bj.bcebos.com/wmt/wmt14.tgz
代码示例
:::::::::

COPY-FROM: paddle.text.datasets.WMT14
COPY-FROM: paddle.text.WMT14
4 changes: 2 additions & 2 deletions docs/api/paddle/text/WMT16_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@
WMT16
-------------------------------

.. py:class:: paddle.text.datasets.WMT16()
.. py:class:: paddle.text.WMT16(data_file = None, mode = 'train', src_dict_size = -1, trg_dict_size = -1, lang = 'ch', download = True)
该类是对 `WMT16 <http://www.statmt.org/wmt16/>`_ 测试数据集实现。
Expand Down Expand Up @@ -32,4 +32,4 @@ http://www.statmt.org/wmt16/multimodal-task.html#task1
代码示例
:::::::::

COPY-FROM: paddle.text.datasets.WMT16
COPY-FROM: paddle.text.WMT16

0 comments on commit d8dca18

Please sign in to comment.