2024-02-05 19:02:08

totrans/vis017_03.yaml

 - en: TVTensors
+  id: totrans-0
   prefs:
   - PREF_H1
   type: TYPE_NORMAL
+  zh: TVTensors
 - en: 原文：[https://pytorch.org/vision/stable/tv_tensors.html](https://pytorch.org/vision/stable/tv_tensors.html)
+  id: totrans-1
   prefs:
   - PREF_BQ
   type: TYPE_NORMAL
+  zh: 原文：[https://pytorch.org/vision/stable/tv_tensors.html](https://pytorch.org/vision/stable/tv_tensors.html)
 - en: TVTensors are [`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
     "(in PyTorch v2.2)") subclasses which the v2 [transforms](transforms.html#transforms)
     use under the hood to dispatch their inputs to the appropriate lower-level kernels.
     Most users do not need to manipulate TVTensors directly.
+  id: totrans-2
   prefs: []
   type: TYPE_NORMAL
+  zh: TVTensors是[`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
+    "(在PyTorch v2.2中)")的子类，v2 [transforms](transforms.html#transforms)在内部使用它们来将输入分派到适当的底层内核。大多数用户不需要直接操作TVTensors。
 - en: Refer to [Getting started with transforms v2](auto_examples/transforms/plot_transforms_getting_started.html#sphx-glr-auto-examples-transforms-plot-transforms-getting-started-py)
     for an introduction to TVTensors, or [TVTensors FAQ](auto_examples/transforms/plot_tv_tensors.html#sphx-glr-auto-examples-transforms-plot-tv-tensors-py)
     for more advanced info.
+  id: totrans-3
   prefs: []
   type: TYPE_NORMAL
+  zh: 有关TVTensors的介绍，请参阅[开始使用transforms v2](auto_examples/transforms/plot_transforms_getting_started.html#sphx-glr-auto-examples-transforms-plot-transforms-getting-started-py)，或者查看[TVTensors
+    FAQ](auto_examples/transforms/plot_tv_tensors.html#sphx-glr-auto-examples-transforms-plot-tv-tensors-py)以获取更多高级信息。
 - en: '| [`Image`](generated/torchvision.tv_tensors.Image.html#torchvision.tv_tensors.Image
     "torchvision.tv_tensors.Image")(data, *[, dtype, device, requires_grad]) | [`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
     "(in PyTorch v2.2)") subclass for images. |'
+  id: totrans-4
   prefs: []
   type: TYPE_TB
+  zh: '| [`Image`](generated/torchvision.tv_tensors.Image.html#torchvision.tv_tensors.Image
+    "torchvision.tv_tensors.Image")(data, *[, dtype, device, requires_grad]) | 用于图像的[`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
+    "(在PyTorch v2.2中)")子类。 |'
 - en: '| [`Video`](generated/torchvision.tv_tensors.Video.html#torchvision.tv_tensors.Video
     "torchvision.tv_tensors.Video")(data, *[, dtype, device, requires_grad]) | [`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
     "(in PyTorch v2.2)") subclass for videos. |'
+  id: totrans-5
   prefs: []
   type: TYPE_TB
+  zh: '| [`Video`](generated/torchvision.tv_tensors.Video.html#torchvision.tv_tensors.Video
+    "torchvision.tv_tensors.Video")(data, *[, dtype, device, requires_grad]) | 用于视频的[`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
+    "(在PyTorch v2.2中)")子类。 |'
 - en: '| [`BoundingBoxFormat`](generated/torchvision.tv_tensors.BoundingBoxFormat.html#torchvision.tv_tensors.BoundingBoxFormat
     "torchvision.tv_tensors.BoundingBoxFormat")(value) | Coordinate format of a bounding
     box. |'
+  id: totrans-6
   prefs: []
   type: TYPE_TB
+  zh: '| [`BoundingBoxFormat`](generated/torchvision.tv_tensors.BoundingBoxFormat.html#torchvision.tv_tensors.BoundingBoxFormat
+    "torchvision.tv_tensors.BoundingBoxFormat")(value) | 边界框的坐标格式。 |'
 - en: '| [`BoundingBoxes`](generated/torchvision.tv_tensors.BoundingBoxes.html#torchvision.tv_tensors.BoundingBoxes
     "torchvision.tv_tensors.BoundingBoxes")(data, *, format, canvas_size) | [`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
     "(in PyTorch v2.2)") subclass for bounding boxes. |'
+  id: totrans-7
   prefs: []
   type: TYPE_TB
+  zh: '| [`BoundingBoxes`](generated/torchvision.tv_tensors.BoundingBoxes.html#torchvision.tv_tensors.BoundingBoxes
+    "torchvision.tv_tensors.BoundingBoxes")(data, *, format, canvas_size) | 用于边界框的[`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
+    "(在PyTorch v2.2中)")子类。 |'
 - en: '| [`Mask`](generated/torchvision.tv_tensors.Mask.html#torchvision.tv_tensors.Mask
     "torchvision.tv_tensors.Mask")(data, *[, dtype, device, requires_grad]) | [`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
     "(in PyTorch v2.2)") subclass for segmentation and detection masks. |'
+  id: totrans-8
   prefs: []
   type: TYPE_TB
+  zh: '| [`Mask`](generated/torchvision.tv_tensors.Mask.html#torchvision.tv_tensors.Mask
+    "torchvision.tv_tensors.Mask")(data, *[, dtype, device, requires_grad]) | 用于分割和检测掩码的[`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
+    "(在PyTorch v2.2中)")子类。 |'
 - en: '| [`TVTensor`](generated/torchvision.tv_tensors.TVTensor.html#torchvision.tv_tensors.TVTensor
     "torchvision.tv_tensors.TVTensor") | Base class for all TVTensors. |'
+  id: totrans-9
   prefs: []
   type: TYPE_TB
+  zh: '| [`TVTensor`](generated/torchvision.tv_tensors.TVTensor.html#torchvision.tv_tensors.TVTensor
+    "torchvision.tv_tensors.TVTensor") | 所有TVTensors的基类。 |'
 - en: '| [`set_return_type`](generated/torchvision.tv_tensors.set_return_type.html#torchvision.tv_tensors.set_return_type
     "torchvision.tv_tensors.set_return_type")(return_type) | Set the return type of
     torch operations on [`TVTensor`](generated/torchvision.tv_tensors.TVTensor.html#torchvision.tv_tensors.TVTensor
     "torchvision.tv_tensors.TVTensor"). |'
+  id: totrans-10
   prefs: []
   type: TYPE_TB
+  zh: '| [`set_return_type`](generated/torchvision.tv_tensors.set_return_type.html#torchvision.tv_tensors.set_return_type
+    "torchvision.tv_tensors.set_return_type")(return_type) | 设置[`TVTensor`](generated/torchvision.tv_tensors.TVTensor.html#torchvision.tv_tensors.TVTensor
+    "torchvision.tv_tensors.TVTensor")上torch操作的返回类型。 |'
 - en: '| [`wrap`](generated/torchvision.tv_tensors.wrap.html#torchvision.tv_tensors.wrap
     "torchvision.tv_tensors.wrap")(wrappee, *, like, **kwargs) | Convert a [`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
     "(in PyTorch v2.2)") (`wrappee`) into the same [`TVTensor`](generated/torchvision.tv_tensors.TVTensor.html#torchvision.tv_tensors.TVTensor
     "torchvision.tv_tensors.TVTensor") subclass as `like`. |'
+  id: totrans-11
   prefs: []
   type: TYPE_TB
+  zh: '| [`wrap`](generated/torchvision.tv_tensors.wrap.html#torchvision.tv_tensors.wrap
+    "torchvision.tv_tensors.wrap")(wrappee, *, like, **kwargs) | 将[`torch.Tensor`](https://pytorch.org/docs/stable/tensors.html#torch.Tensor
+    "(在PyTorch v2.2中)") (`wrappee`)转换为与`like`相同的[`TVTensor`](generated/torchvision.tv_tensors.TVTensor.html#torchvision.tv_tensors.TVTensor
+    "torchvision.tv_tensors.TVTensor")子类。 |'

totrans/vis017_06.yaml

 - en: Utils
+  id: totrans-0
   prefs:
   - PREF_H1
   type: TYPE_NORMAL
+  zh: Utils
 - en: 原文：[https://pytorch.org/vision/stable/utils.html](https://pytorch.org/vision/stable/utils.html)
+  id: totrans-1
   prefs:
   - PREF_BQ
   type: TYPE_NORMAL
+  zh: 原文：[https://pytorch.org/vision/stable/utils.html](https://pytorch.org/vision/stable/utils.html)
 - en: The `torchvision.utils` module contains various utilities, mostly [for visualization](auto_examples/others/plot_visualization_utils.html#sphx-glr-auto-examples-others-plot-visualization-utils-py).
+  id: totrans-2
   prefs: []
   type: TYPE_NORMAL
+  zh: '`torchvision.utils` 模块包含各种实用工具，主要用于[可视化](auto_examples/others/plot_visualization_utils.html#sphx-glr-auto-examples-others-plot-visualization-utils-py)。'
 - en: '| [`draw_bounding_boxes`](generated/torchvision.utils.draw_bounding_boxes.html#torchvision.utils.draw_bounding_boxes
     "torchvision.utils.draw_bounding_boxes")(image, boxes[, labels, ...]) | Draws
     bounding boxes on given image. |'
+  id: totrans-3
   prefs: []
   type: TYPE_TB
+  zh: '| [`draw_bounding_boxes`](generated/torchvision.utils.draw_bounding_boxes.html#torchvision.utils.draw_bounding_boxes
+    "torchvision.utils.draw_bounding_boxes")(image, boxes[, labels, ...]) | 在给定的图像上绘制边界框。
+    |'
 - en: '| [`draw_segmentation_masks`](generated/torchvision.utils.draw_segmentation_masks.html#torchvision.utils.draw_segmentation_masks
     "torchvision.utils.draw_segmentation_masks")(image, masks[, ...]) | Draws segmentation
     masks on given RGB image. |'
+  id: totrans-4
   prefs: []
   type: TYPE_TB
+  zh: '| [`draw_segmentation_masks`](generated/torchvision.utils.draw_segmentation_masks.html#torchvision.utils.draw_segmentation_masks
+    "torchvision.utils.draw_segmentation_masks")(image, masks[, ...]) | 在给定的 RGB 图像上绘制分割蒙版。
+    |'
 - en: '| [`draw_keypoints`](generated/torchvision.utils.draw_keypoints.html#torchvision.utils.draw_keypoints
     "torchvision.utils.draw_keypoints")(image, keypoints[, ...]) | Draws Keypoints
     on given RGB image. |'
+  id: totrans-5
   prefs: []
   type: TYPE_TB
+  zh: '| [`draw_keypoints`](generated/torchvision.utils.draw_keypoints.html#torchvision.utils.draw_keypoints
+    "torchvision.utils.draw_keypoints")(image, keypoints[, ...]) | 在给定的 RGB 图像上绘制关键点。
+    |'
 - en: '| [`flow_to_image`](generated/torchvision.utils.flow_to_image.html#torchvision.utils.flow_to_image
     "torchvision.utils.flow_to_image")(flow) | Converts a flow to an RGB image. |'
+  id: totrans-6
   prefs: []
   type: TYPE_TB
+  zh: '| [`flow_to_image`](generated/torchvision.utils.flow_to_image.html#torchvision.utils.flow_to_image
+    "torchvision.utils.flow_to_image")(flow) | 将光流转换为 RGB 图像。 |'
 - en: '| [`make_grid`](generated/torchvision.utils.make_grid.html#torchvision.utils.make_grid
     "torchvision.utils.make_grid")(tensor[, nrow, padding, ...]) | Make a grid of
     images. |'
+  id: totrans-7
   prefs: []
   type: TYPE_TB
+  zh: '| [`make_grid`](generated/torchvision.utils.make_grid.html#torchvision.utils.make_grid
+    "torchvision.utils.make_grid")(tensor[, nrow, padding, ...]) | 制作图像网格。 |'
 - en: '| [`save_image`](generated/torchvision.utils.save_image.html#torchvision.utils.save_image
     "torchvision.utils.save_image")(tensor, fp[, format]) | Save a given Tensor into
     an image file. |'
+  id: totrans-8
   prefs: []
   type: TYPE_TB
+  zh: '| [`save_image`](generated/torchvision.utils.save_image.html#torchvision.utils.save_image
+    "torchvision.utils.save_image")(tensor, fp[, format]) | 将给定的张量保存为图像文件。 |'

totrans/vis017_08.yaml

 - en: Decoding / Encoding images and videos
+  id: totrans-0
   prefs:
   - PREF_H1
   type: TYPE_NORMAL
+  zh: 解码/编码图像和视频
 - en: 原文：[https://pytorch.org/vision/stable/io.html](https://pytorch.org/vision/stable/io.html)
+  id: totrans-1
   prefs:
   - PREF_BQ
   type: TYPE_NORMAL
+  zh: 原文：[https://pytorch.org/vision/stable/io.html](https://pytorch.org/vision/stable/io.html)
 - en: The `torchvision.io` package provides functions for performing IO operations.
     They are currently specific to reading and writing images and videos.
+  id: totrans-2
   prefs: []
   type: TYPE_NORMAL
+  zh: '`torchvision.io`包提供了执行IO操作的函数。目前这些函数专门用于读取和写入图像和视频。'
 - en: Images[](#images "Permalink to this heading")
+  id: totrans-3
   prefs:
   - PREF_H2
   type: TYPE_NORMAL
+  zh: 图像[](#images "跳转到此标题")
 - en: '| [`read_image`](generated/torchvision.io.read_image.html#torchvision.io.read_image
     "torchvision.io.read_image")(path[, mode]) | Reads a JPEG or PNG image into a
     3 dimensional RGB or grayscale Tensor. |'
+  id: totrans-4
   prefs: []
   type: TYPE_TB
+  zh: '| [`read_image`](generated/torchvision.io.read_image.html#torchvision.io.read_image
+    "torchvision.io.read_image")(path[, mode]) | 将JPEG或PNG图像读入三维RGB或灰度张量。 |'
 - en: '| [`decode_image`](generated/torchvision.io.decode_image.html#torchvision.io.decode_image
     "torchvision.io.decode_image")(input[, mode]) | Detects whether an image is a
     JPEG or PNG and performs the appropriate operation to decode the image into a
     3 dimensional RGB or grayscale Tensor. |'
+  id: totrans-5
   prefs: []
   type: TYPE_TB
+  zh: '| [`decode_image`](generated/torchvision.io.decode_image.html#torchvision.io.decode_image
+    "torchvision.io.decode_image")(input[, mode]) | 检测图像是JPEG还是PNG，并执行适当的操作将图像解码为三维RGB或灰度张量。
+    |'
 - en: '| [`encode_jpeg`](generated/torchvision.io.encode_jpeg.html#torchvision.io.encode_jpeg
     "torchvision.io.encode_jpeg")(input[, quality]) | Takes an input tensor in CHW
     layout and returns a buffer with the contents of its corresponding JPEG file.
     |'
+  id: totrans-6
   prefs: []
   type: TYPE_TB
+  zh: '| [`encode_jpeg`](generated/torchvision.io.encode_jpeg.html#torchvision.io.encode_jpeg
+    "torchvision.io.encode_jpeg")(input[, quality]) | 将输入张量按CHW布局编码为其对应JPEG文件内容的缓冲区。
+    |'
 - en: '| [`decode_jpeg`](generated/torchvision.io.decode_jpeg.html#torchvision.io.decode_jpeg
     "torchvision.io.decode_jpeg")(input[, mode, device]) | Decodes a JPEG image into
     a 3 dimensional RGB or grayscale Tensor. |'
+  id: totrans-7
   prefs: []
   type: TYPE_TB
+  zh: '| [`decode_jpeg`](generated/torchvision.io.decode_jpeg.html#torchvision.io.decode_jpeg
+    "torchvision.io.decode_jpeg")(input[, mode, device]) | 将JPEG图像解码为三维RGB或灰度张量。 |'
 - en: '| [`write_jpeg`](generated/torchvision.io.write_jpeg.html#torchvision.io.write_jpeg
     "torchvision.io.write_jpeg")(input, filename[, quality]) | Takes an input tensor
     in CHW layout and saves it in a JPEG file. |'
+  id: totrans-8
   prefs: []
   type: TYPE_TB
+  zh: '| [`write_jpeg`](generated/torchvision.io.write_jpeg.html#torchvision.io.write_jpeg
+    "torchvision.io.write_jpeg")(input, filename[, quality]) | 将输入张量按CHW布局保存为JPEG文件。
+    |'
 - en: '| [`encode_png`](generated/torchvision.io.encode_png.html#torchvision.io.encode_png
     "torchvision.io.encode_png")(input[, compression_level]) | Takes an input tensor
     in CHW layout and returns a buffer with the contents of its corresponding PNG
     file. |'
+  id: totrans-9
   prefs: []
   type: TYPE_TB
+  zh: '| [`encode_png`](generated/torchvision.io.encode_png.html#torchvision.io.encode_png
+    "torchvision.io.encode_png")(input[, compression_level]) | 将输入张量按CHW布局编码为其对应PNG文件内容的缓冲区。
+    |'
 - en: '| [`decode_png`](generated/torchvision.io.decode_png.html#torchvision.io.decode_png
     "torchvision.io.decode_png")(input[, mode]) | Decodes a PNG image into a 3 dimensional
     RGB or grayscale Tensor. |'
+  id: totrans-10
   prefs: []
   type: TYPE_TB
+  zh: '| [`decode_png`](generated/torchvision.io.decode_png.html#torchvision.io.decode_png
+    "torchvision.io.decode_png")(input[, mode]) | 将PNG图像解码为三维RGB或灰度张量。 |'
 - en: '| [`write_png`](generated/torchvision.io.write_png.html#torchvision.io.write_png
     "torchvision.io.write_png")(input, filename[, compression_level]) | Takes an input
     tensor in CHW layout (or HW in the case of grayscale images) and saves it in a
     PNG file. |'
+  id: totrans-11
   prefs: []
   type: TYPE_TB
+  zh: '| [`write_png`](generated/torchvision.io.write_png.html#torchvision.io.write_png
+    "torchvision.io.write_png")(input, filename[, compression_level]) | 将输入张量按CHW布局（或灰度图像的情况下按HW布局）保存为PNG文件。
+    |'
 - en: '| [`read_file`](generated/torchvision.io.read_file.html#torchvision.io.read_file
     "torchvision.io.read_file")(path) | Reads and outputs the bytes contents of a
     file as a uint8 Tensor with one dimension. |'
+  id: totrans-12
   prefs: []
   type: TYPE_TB
+  zh: '| [`read_file`](generated/torchvision.io.read_file.html#torchvision.io.read_file
+    "torchvision.io.read_file")(path) | 读取文件的字节内容，并输出为具有一维uint8张量。 |'
 - en: '| [`write_file`](generated/torchvision.io.write_file.html#torchvision.io.write_file
     "torchvision.io.write_file")(filename, data) | Writes the contents of an uint8
     tensor with one dimension to a file. |'
+  id: totrans-13
   prefs: []
   type: TYPE_TB
+  zh: '| [`write_file`](generated/torchvision.io.write_file.html#torchvision.io.write_file
+    "torchvision.io.write_file")(filename, data) | 将具有一维的uint8张量内容写入文件。 |'
 - en: '| [`ImageReadMode`](generated/torchvision.io.ImageReadMode.html#torchvision.io.ImageReadMode
     "torchvision.io.ImageReadMode")(value) | Support for various modes while reading
     images. |'
+  id: totrans-14
   prefs: []
   type: TYPE_TB
+  zh: '| [`ImageReadMode`](generated/torchvision.io.ImageReadMode.html#torchvision.io.ImageReadMode
+    "torchvision.io.ImageReadMode")(value) | 在读取图像时支持各种模式。 |'
 - en: Video[](#video "Permalink to this heading")
+  id: totrans-15
   prefs:
   - PREF_H2
   type: TYPE_NORMAL
+  zh: 视频[](#video "跳转到此标题")
 - en: '| [`read_video`](generated/torchvision.io.read_video.html#torchvision.io.read_video
     "torchvision.io.read_video")(filename[, start_pts, end_pts, ...]) | Reads a video
     from a file, returning both the video frames and the audio frames |'
+  id: totrans-16
   prefs: []
   type: TYPE_TB
+  zh: '| [`read_video`](generated/torchvision.io.read_video.html#torchvision.io.read_video
+    "torchvision.io.read_video")(filename[, start_pts, end_pts, ...]) | 从文件中读取视频，返回视频帧和音频帧
+    |'
 - en: '| [`read_video_timestamps`](generated/torchvision.io.read_video_timestamps.html#torchvision.io.read_video_timestamps
     "torchvision.io.read_video_timestamps")(filename[, pts_unit]) | List the video
     frames timestamps. |'
+  id: totrans-17
   prefs: []
   type: TYPE_TB
+  zh: '| [`read_video_timestamps`](generated/torchvision.io.read_video_timestamps.html#torchvision.io.read_video_timestamps
+    "torchvision.io.read_video_timestamps")(filename[, pts_unit]) | 列出视频帧的时间戳。 |'
 - en: '| [`write_video`](generated/torchvision.io.write_video.html#torchvision.io.write_video
     "torchvision.io.write_video")(filename, video_array, fps[, ...]) | Writes a 4d
     tensor in [T, H, W, C] format in a video file |'
+  id: totrans-18
   prefs: []
   type: TYPE_TB
+  zh: '| [`write_video`](generated/torchvision.io.write_video.html#torchvision.io.write_video
+    "torchvision.io.write_video")(filename, video_array, fps[, ...]) | 将[T, H, W,
+    C]格式的4维张量写入视频文件 |'
 - en: Fine-grained video API[](#fine-grained-video-api "Permalink to this heading")
+  id: totrans-19
   prefs:
   - PREF_H3
   type: TYPE_NORMAL
+  zh: 细粒度视频API[](#fine-grained-video-api "跳转到此标题")
 - en: In addition to the `read_video` function, we provide a high-performance lower-level
     API for more fine-grained control compared to the `read_video` function. It does
     all this whilst fully supporting torchscript.
+  id: totrans-20
   prefs: []
   type: TYPE_NORMAL
+  zh: 除了`read_video`函数外，我们还提供了一个高性能的低级API，用于比`read_video`函数更精细的控制。它在完全支持torchscript的同时完成所有这些操作。
 - en: Warning
+  id: totrans-21
   prefs: []
   type: TYPE_NORMAL
+  zh: 警告
 - en: The fine-grained video API is in Beta stage, and backward compatibility is not
     guaranteed.
+  id: totrans-22
   prefs: []
   type: TYPE_NORMAL
+  zh: 细粒度视频API处于Beta阶段，不保证向后兼容性。
 - en: '| [`VideoReader`](generated/torchvision.io.VideoReader.html#torchvision.io.VideoReader
     "torchvision.io.VideoReader")(src[, stream, num_threads]) | Fine-grained video-reading
     API. |'
+  id: totrans-23
   prefs: []
   type: TYPE_TB
+  zh: '| [`VideoReader`](generated/torchvision.io.VideoReader.html#torchvision.io.VideoReader
+    "torchvision.io.VideoReader")(src[, stream, num_threads]) | 细粒度视频读取API。 |'
 - en: 'Example of inspecting a video:'
+  id: totrans-24
   prefs: []
   type: TYPE_NORMAL
+  zh: 检查视频的示例：
 - en: '[PRE0]'
+  id: totrans-25
   prefs: []
   type: TYPE_PRE
+  zh: '[PRE0]'

totrans/vis017_10.yaml

 - en: Examples and training references
+  id: totrans-0
   prefs:
   - PREF_H1
   type: TYPE_NORMAL
+  zh: 示例和培训参考资料

totrans/vis017_12.yaml

 - en: Training references
+  id: totrans-0
   prefs:
   - PREF_H1
   type: TYPE_NORMAL
+  zh: 训练参考
 - en: 原文：[https://pytorch.org/vision/stable/training_references.html](https://pytorch.org/vision/stable/training_references.html)
+  id: totrans-1
   prefs:
   - PREF_BQ
   type: TYPE_NORMAL
+  zh: 原文：[https://pytorch.org/vision/stable/training_references.html](https://pytorch.org/vision/stable/training_references.html)
 - en: On top of the many models, datasets, and image transforms, Torchvision also
     provides training reference scripts. These are the scripts that we use to train
     the [models](models.html#models) which are then available with pre-trained weights.
+  id: totrans-2
   prefs: []
   type: TYPE_NORMAL
+  zh: 除了许多模型、数据集和图像转换之外，Torchvision还提供训练参考脚本。这些脚本是我们用来训练[模型](models.html#models)的，然后这些模型就可以使用预训练的权重。
 - en: These scripts are not part of the core package and are instead available [on
     GitHub](https://github.com/pytorch/vision/tree/main/references). We currently
     provide references for [classification](https://github.com/pytorch/vision/tree/main/references/classification),
@@ -18,22 +24,32 @@
     [segmentation](https://github.com/pytorch/vision/tree/main/references/segmentation),
     [similarity learning](https://github.com/pytorch/vision/tree/main/references/similarity),
     and [video classification](https://github.com/pytorch/vision/tree/main/references/video_classification).
+  id: totrans-3
   prefs: []
   type: TYPE_NORMAL
+  zh: 这些脚本不是核心包的一部分，而是在[GitHub](https://github.com/pytorch/vision/tree/main/references)上提供。我们目前为[分类](https://github.com/pytorch/vision/tree/main/references/classification)、[检测](https://github.com/pytorch/vision/tree/main/references/detection)、[分割](https://github.com/pytorch/vision/tree/main/references/segmentation)、[相似性学习](https://github.com/pytorch/vision/tree/main/references/similarity)和[视频分类](https://github.com/pytorch/vision/tree/main/references/video_classification)提供参考。
 - en: While these scripts are largely stable, they do not offer backward compatibility
     guarantees.
+  id: totrans-4
   prefs: []
   type: TYPE_NORMAL
+  zh: 尽管这些脚本在很大程度上是稳定的，但它们不提供向后兼容性保证。
 - en: 'In general, these scripts rely on the latest (not yet released) pytorch version
     or the latest torchvision version. This means that to use them, **you might need
     to install the latest pytorch and torchvision versions**, with e.g.:'
+  id: totrans-5
   prefs: []
   type: TYPE_NORMAL
+  zh: 一般来说，这些脚本依赖于最新（尚未发布）的pytorch版本或最新的torchvision版本。这意味着要使用它们，**您可能需要安装最新的pytorch和torchvision版本**，例如：
 - en: '[PRE0]'
+  id: totrans-6
   prefs: []
   type: TYPE_PRE
+  zh: '[PRE0]'
 - en: If you need to rely on an older stable version of pytorch or torchvision, e.g.
     torchvision 0.10, then it’s safer to use the scripts from that corresponding release
     on GitHub, namely [https://github.com/pytorch/vision/tree/v0.10.0/references](https://github.com/pytorch/vision/tree/v0.10.0/references).
+  id: totrans-7
   prefs: []
   type: TYPE_NORMAL
+  zh: 如果您需要依赖于较旧的稳定版本的pytorch或torchvision，例如torchvision 0.10，那么最好使用GitHub上对应发布的脚本，即[https://github.com/pytorch/vision/tree/v0.10.0/references](https://github.com/pytorch/vision/tree/v0.10.0/references)。

totrans/vis017_13.yaml

 - en: PyTorch Libraries
+  id: totrans-0
   prefs:
   - PREF_H1
   type: TYPE_NORMAL
+  zh: PyTorch库