### 实时文档

在本地开发时，可以使用该脚本构建站点并检查所做的任何更改，并实时重载：

<div class="termy">

```console
$ python ./scripts/docs.py live

<span style="color: green;">[INFO]</span> Serving on http://127.0.0.1:8008
<span style="color: green;">[INFO]</span> Start watching changes
<span style="color: green;">[INFO]</span> Start detecting changes
```

</div>

文档服务将运行在 `http://127.0.0.1:8008`。

这样，你可以编辑文档 / 源文件并实时查看更改。

!!! tip

    @Override
    public void render(String text, SimpleHtmlWriter htmlWriter) throws IOException {
        // Wrap in a <span>, to work around CSS problem in IE
        htmlWriter.startElement("span").attribute("class", "code")
            .startElement("pre").characters(text).endElement()
        .endElement();
    }

    params->emplace_back(handle);
  }
  return absl::OkStatus();
}

// Runs `model` maybe wrapped in a function.
Status RunModel(Model model, AbstractContext* ctx,
                absl::Span<AbstractTensorHandle* const> inputs,
                absl::Span<AbstractTensorHandle*> outputs, bool use_function) {
  if (use_function) {
    const char* fn_name = "test_fn";
    core::RefCountPtr<AbstractFunction> scoped_func;

    <a class="badge" href="https://dpeuniversity.gradle.com/app/courses/ec69d0b8-9171-4969-ac3e-82dea16f87b0/" target="_blank">
        <span class="badge-type button--blue">LEARN</span>
        <span class="badge-text">Incremental Builds and Build Caching with Gradle&nbsp;&nbsp;&nbsp;&gt;</span>
    </a>
</div>
++++

Gradle uses two main features to reduce build time: *incremental builds* and *build caching*.

    <a class="badge" href="https://dpeuniversity.gradle.com/app/courses/ec69d0b8-9171-4969-ac3e-82dea16f87b0/" target="_blank">
        <span class="badge-type button--blue">LEARN</span>
        <span class="badge-text">Incremental Builds and Build Caching with Gradle&nbsp;&nbsp;&nbsp;&gt;</span>
    </a>
</div>
++++

Learn the basics of Gradle's incremental builds.

****
**In this section you will:**

  // XlaCompilationResult pointed to by `result`.
  virtual Status Compile(const XlaCompiler::CompileOptions& options,
                         const NameAttrList& function,
                         absl::Span<const XlaArgument> args,
                         XlaCompilationResult* result) = 0;

  // Compiles a Tensorflow single op to an HloModuleProto stored in the
  // XlaCompilationResult pointed to by `result`.

==============================================================================*/
#ifndef TENSORFLOW_C_EAGER_GRADIENT_CHECKER_H_
#define TENSORFLOW_C_EAGER_GRADIENT_CHECKER_H_

#include <memory>

#include "absl/types/span.h"
#include "tensorflow/c/eager/abstract_tensor_handle.h"
#include "tensorflow/c/eager/unified_api_testutil.h"

namespace tensorflow {
namespace gradients {

			continue
		}
		obj, span, objIndex := findObject(ptr, 0, 0)
		if obj == 0 {
			continue
		}
		// TODO: Consider making two passes where the first
		// just prefetches the mark bits.
		mbits := span.markBitsForIndex(objIndex)
		if mbits.isMarked() {
			continue
		}
		mbits.setMarked()

		// Mark span.
		arena, pageIdx, pageMask := pageIndexOf(span.base())

  const string& DeviceName() const override;
  Status SetDeviceName(const char* name) override;
  Status AddInput(AbstractTensorHandle* input) override;
  Status AddInputList(absl::Span<AbstractTensorHandle* const> inputs) override;
  Status Execute(absl::Span<AbstractTensorHandle*> retvals,
                 int* num_retvals) override;
  Status SetAttrString(const char* attr_name, const char* data,
                       size_t length) override;

                 AbstractTensorHandle* const tensor_names,
                 AbstractTensorHandle* const shape_and_slices,
                 absl::Span<AbstractTensorHandle*> tensors,
                 absl::Span<DataType> dtypes, const char* name,
                 const char* raw_device_name) {
  AbstractOperationPtr op_ptr(ctx->CreateOperation());