Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

docsV0.8.0 删除接口check list #8695

Closed
laoliu97 opened this issue Jul 19, 2022 · 5 comments
Closed

docsV0.8.0 删除接口check list #8695

laoliu97 opened this issue Jul 19, 2022 · 5 comments

Comments

@laoliu97
Copy link
Contributor

laoliu97 commented Jul 19, 2022

请对照下方PR检查各自负责的文档模块,是否存在删除、弃用或改名的接口,并将有无改动和改动内容在本issue中进行显式说明

PR:#8392

task 负责人
oneflow Chengguoliang
oneflow.nn Chengguoliang
oneflow.Tensor Chengguoliang
oneflow.linalg Chengguoliang
oneflow.nn.init Chengguoliang
oneflow.utils.data Chengguoliang
oneflow.optim Chengguoliang
oneflow.autograd Chengguoliang
tensor attribute Liuxuan 无
oneflow.cuda Liuxuan
oneflow.nn.functional Liuxuan
oneflow.nn.graph ChenQiaoLing #8695 (comment)
oneflow.distributed ChenQiaoLing 无
oneflow.oneEmbedding ChenQiaoLing 无
@laoliu97 laoliu97 added bug community events from community labels Jul 19, 2022
@lmyybh
Copy link
Contributor

lmyybh commented Jul 19, 2022

以下模块缺少的接口基本都是因为重构时只重点关注了与 pytorch 文档排布结构一致,比较 pytorch 文档中有,而 oneflow 未实现的接口,缺少与 最初 rst 文件的比较

oneflow

oneflow 接口 情况说明
BoolTensor, ByteTensor, CharTensor, DoubleTensor, FloatTensor, HalfTensor, IntTensor, LongTensor pytorch 模块有,但未列入文档
adaptive_avg_pool1d, adaptive_avg_pool2d, adaptive_avg_pool3d pytorch 模块有,但未列入文档
masked_fill, randint_like, selu, softmax pytorch 模块有,但未列入文档
batch_gather, cast, decode_onerec, expand, gather_nd, gelu, in_top_k, mish, new_ones, nms, repeat, roc_auc_score, silu, slice, slice_update, softplus, softsign, tensor_scatter_nd_update pytorch 没有,oneflow 文档遗漏

oneflow.nn

oneflow.nn 接口 情况说明
nn.COCOReader, nn.CoinFlip, nn.CropMirrorNormalize, nn.FakeQuantization, nn.MinMaxObserver, nn.MovingAverageMinMaxObserver, nn.OFRecordBytesDecoder, nn.OFRecordImageDecoder, nn.OFRecordImageDecoderRandomCrop, nn.OFRecordRawDecoder, nn.OFRecordReader, nn.Quantization pytorch 没有,oneflow 文档遗漏
nn.Parameter 没有 docstring,遗漏

oneflow.optim

与之前版本相比,缺少了 oneflow.optim.Optimizer 接口 (无 docstring,pytorch 中有,但未列入文档)

oneflow.Tensor

与之前版本相比,缺少了 Tensor.log1p 接口 (pytorch 文档中有,记录时遗漏)


以下四个模块相比之前版本,没有接口缺失

oneflow.linalg

oneflow.utils.data

oneflow.autograd

oneflow.nn.init

@laoliu97
Copy link
Contributor Author

oneflow.cuda 删除的接口

oneflow.cuda 接口 接口删除情况说明 处理办法
cuda.empty_cache 文档重构时以docsv0.8分支作为参考,而此接口于3周前添加文档,重构时并未发现该接口。也许在重构后有同事添加,或者是docsv0.8分支未merge master获取更新。https://github.com/Oneflow-Inc/oneflow/pull/8514/files#diff-fa1840d691e1cc1768c6cd2b2d296cd68e7052861a5b6cac85e949cb347815ee 补充增加

oneflow.nn.functional删除的接口

oneflow.nn.functional接口 接口删除情况说明 处理办法
functional.ctc_greedy_decoder 文档重构时以docsv0.8分支作为参考,文档重构于2周前完成,重构时并未发现该接口。也许在重构后有同事添加,或者是docsv0.8分支未merge master获取更新。PR: https://github.com/Oneflow-Inc/oneflow/pull/8514/files#diff-2c9d00219c966a687953069d2e4e18e113c3068d4cd0be023bb4fcdd248e82c7 重构时的接口检查:https://github.com/Oneflow-Inc/OneTeam/issues/1452#issuecomment-1168297596 补充增加
functional.upsample 文档重构时以docsv0.8分支作为参考,而此接口于2周前添加,文档重构已完成,重构过程中该接口并未实现。也许在重构后有同事添加,或者是docsv0.8分支未merge master获取更新。PR:https://github.com/Oneflow-Inc/oneflow/pull/8514/files#diff-2c9d00219c966a687953069d2e4e18e113c3068d4cd0be023bb4fcdd248e82c7 重构时的接口检查:https://github.com/Oneflow-Inc/OneTeam/issues/1452#issuecomment-1168297596 补充增加

Tensor Attributes 删除的接口

oneflow.cuda 接口 接口删除情况说明 处理办法

@QiaolingChen00
Copy link
Contributor

QiaolingChen00 commented Jul 21, 2022

nn.graph v0.8.0 重构时漏下的 oneflow 特有接口

        set_zero_redundancy_optimizer_mode,
        set_zero_redundancy_optimizer_min_size_after_split,
        enable_xla_jit,
        enable_tensorrt,
        enable_openvino,

疑问??

上述接口在 0.7.0 版本中被引入,但在重构前的 master 版本又被删除,是否是已经被弃用的接口?
验证如下:

>>> import oneflow
>>> oneflow.nn.graph.graph_config.GraphConfig.enable_xla_jit
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: type object 'GraphConfig' has no attribute 'enable_xla_jit'
>>> oneflow.nn.graph.graph_config.GraphConfig.enable_xla_jit()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: type object 'GraphConfig' has no attribute 'enable_xla_jit'
>>> oneflow.nn.graph.graph_config.GraphConfig.enable_tensorrt
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: type object 'GraphConfig' has no attribute 'enable_tensorrt'

@strint
Copy link
Contributor

strint commented Jul 21, 2022

nn.graph v0.8.0 重构时漏下的 oneflow 特有接口

        set_zero_redundancy_optimizer_mode,
        set_zero_redundancy_optimizer_min_size_after_split,
        enable_xla_jit,
        enable_tensorrt,
        enable_openvino,

疑问??

上述接口在 0.7.0 版本中被引入,但在重构前的 master 版本又被删除,是否是已经被弃用的接口? 验证如下:

没错,上面的5个接口已经废弃。

@laoliu97 laoliu97 added api and removed community events from community labels Jul 21, 2022
@laoliu97
Copy link
Contributor Author

检查oneflow、oneflow.nn、oneflow.optim、oneflow.Tensor遗漏接口的docstring情况

image

不含docstring的接口有:
oneflow.roc_auc_score
nn.OFRecordImageDecoder
nn.OFRecordImageDecoderRandomCrop
nn.OFRecordRawDecoder
nn.OFRecordReader
nn.Parameter
oneflow.optim.Optimizer

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

4 participants