# tensorplay.cuda.graph_annotations API

Source: https://www.tensorplay.cn/docs/api/tensorplay.cuda.graph_annotations.html

## Functions 4

[#](#api-tensorplay.cuda.graph_annotations.clear_kernel_annotations)

### clear_kernel_annotations

function[Full reference ↗](/docs/generated/tensorplay.cuda.graph_annotations.clear_kernel_annotations.html)

```python
tensorplay.cuda.graph_annotations.clear_kernel_annotations()
```

Clear all recorded kernel annotations.

[#](#api-tensorplay.cuda.graph_annotations.get_kernel_annotations)

### get_kernel_annotations

function[Full reference ↗](/docs/generated/tensorplay.cuda.graph_annotations.get_kernel_annotations.html)

```python
tensorplay.cuda.graph_annotations.get_kernel_annotations()
```

Return the recorded kernel annotations mapping.

[#](#api-tensorplay.cuda.graph_annotations.is_available)

### is_available

function[Full reference ↗](/docs/generated/tensorplay.cuda.graph_annotations.is_available.html)

```python
tensorplay.cuda.graph_annotations.is_available() → bool
```

Whether graph-kernel annotation recording is supported (always False here).

[#](#api-tensorplay.cuda.graph_annotations.mark_kernels)

### mark_kernels

function[Full reference ↗](/docs/generated/tensorplay.cuda.graph_annotations.mark_kernels.html)

```python
tensorplay.cuda.graph_annotations.mark_kernels(name, metadata=None)
```

Tag the GPU work captured within its scope with user metadata.