pytorch/CachingDeviceAllocator.cpp at 9fff8155c362da777e7ce31b85fb2dc7cfced2d5 - pytorch - Gitea: Git for Me

frozenleaves/pytorch

mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00

Files

Yu, Guangye d7114f05b1 Add DeviceAllocator as the base device allocator (#138222 )

# Motivation
In line with [RFC] [A device-agnostic Python device memory related API design for stream-based accelerators](https://github.com/pytorch/pytorch/issues/134978), some memory-related APIs are widely used in popular repositories, such as HuggingFace [so many if-else conditional code](https://github.com/search?q=repo%3Ahuggingface%2Faccelerate%20torch.cuda.empty_cache&type=code). We would like to introduce a generic API set under torch.accelerator namespace to generalize these user cases.

<div align="center">
<table>
<tr>
<td> Device-specific memory APIs torch.xxx.foo</td> <td> Device-agnostic memory APIs torch.accelerator.foo</td>
</tr>
<tr>
<td>

```python
torch.xxx.empty_cache
```

</td>
<td>

```python
torch.accelerator.empty_cache
```

</td>
</tr>

<tr>
<td>

```python
torch.xxx.reset_peak_memory_stats
```

</td>
<td>

```python
torch.accelerator.reset_peak_memory_stats
```

</td>
</tr>

<tr>
<td>

```python
torch.xxx.reset_accumulated_memory_stats
```

</td>
<td>

```python
torch.accelerator.reset_accumulated_memory_stats
```

</td>
</tr>

<tr>
<td>

```python
torch.xxx.memory_stats
```

</td>
<td>

```python
torch.accelerator.memory_stats
```

</td>
</tr>

<tr>
<td>

```python
torch.xxx.memory_allocated
```

</td>
<td>

```python
torch.accelerator.memory_allocated
```

</td>
</tr>

<tr>
<td>

```python
torch.xxx.max_memory_allocated
```

</td>
<td>

```python
torch.accelerator.max_memory_allocated
```

</td>
</tr>

<tr>
<td>

```python
torch.xxx.memory_reserved
```

</td>
<td>

```python
torch.accelerator.memory_reserved
```

</td>
</tr>

<tr>
<td>

```python
torch.xxx.max_memory_reserved
```

</td>
<td>

```python
torch.accelerator.max_memory_reserved
```

</td>
</tr>

</table>
</div>

# Solution
This design follows a similar pattern to `HostAllocator`. We're introducing a base class `DeviceAllocator`, from which `CUDAAllocator` and `XPUAllocator` will inherit. This allows us to provide a unified call path like: `torch.accelerator.empty_cache()` -> `GetDeviceAllocator(allocator)->empty_cache()`.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/138222
Approved by: https://github.com/albanD, https://github.com/Camyll

2025-08-08 17:41:10 +00:00

11 lines

306 B

C++

Raw Blame History

 #include <c10/core/CachingDeviceAllocator.h>
 namespace c10 {
 // Ensures proper DLL export of this pure virtual base class on Windows,
 // since it's mainly used in other DLLs outside c10.dll.
 DeviceAllocator::DeviceAllocator() = default;
 DeviceAllocator::~DeviceAllocator() = default;
 } // namespace c10