CUDAGraph — PyTorch 2.0 documentation (original) (raw)

class torch.cuda.CUDAGraph[source]¶

Wrapper around a CUDA graph.

Warning

This API is in beta and may change in future releases.

capture_begin(pool=None)[source]¶

Begins capturing CUDA work on the current stream.

Typically, you shouldn’t call capture_begin yourself. Use graph or make_graphed_callables(), which call capture_begin internally.

Parameters:

pool (optional) – Token (returned by graph_pool_handle() orother_Graph_instance.pool()) that hints this graph may share memory with the indicated pool. See Graph memory management.

capture_end()[source]¶

Ends CUDA graph capture on the current stream. After capture_end, replay may be called on this instance.

Typically, you shouldn’t call capture_end yourself. Use graph or make_graphed_callables(), which call capture_end internally.

debug_dump(debug_path)[source]¶

Parameters:

debug_path (required) – Path to dump the graph to.

Calls a debugging function to dump the graph if the debugging is enabled via CUDAGraph.enable_debug_mode()

enable_debug_mode()[source]¶

Enables debugging mode for CUDAGraph.debug_dump.

pool()[source]¶

Returns an opaque token representing the id of this graph’s memory pool. This id can optionally be passed to another graph’s capture_begin, which hints the other graph may share the same memory pool.

replay()[source]¶

Replays the CUDA work captured by this graph.

reset()[source]¶

Deletes the graph currently held by this instance.