Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/pytorch/pytorch
/ current_stream
Method
current_stream
torch/_dynamo/device_interface.py:85–86 ·
view source on GitHub ↗
()
Source
from the content-addressed store, hash-verified
83
84
@staticmethod
85
def
current_stream():
86
raise
NotImplementedError()
87
88
@staticmethod
89
def
set_stream(stream: torch.Stream):
Callers
15
__dlpack__
Method · 0.80
from_dlpack
Function · 0.80
to_map
Function · 0.80
__init__
Method · 0.80
__init__
Method · 0.80
_init_streams
Function · 0.80
_pre_forward_unshard
Function · 0.80
_root_pre_forward
Function · 0.80
_pre_backward_hook
Function · 0.80
_post_backward_hook
Function · 0.80
_cast_grad_to_param_dtype
Function · 0.80
_post_backward_final_callback
Function · 0.80
Calls
no outgoing calls
Tested by
15
async_cuda_sleep_and_set_to_one
Function · 0.64
forward
Method · 0.64
_slow_add_on_user_stream
Method · 0.64
test_fused_attention_vs_math_ref_grads_cudagraph
Method · 0.64
test_current_stream
Method · 0.64
test_default_stream
Method · 0.64
test_stream_event_device
Method · 0.64
test_stream_context
Method · 0.64
test_streams_multi_gpu
Method · 0.64
test_streams_multi_gpu_query
Method · 0.64
test_streams_multi_gpu_eq
Method · 0.64
_stream_synchronize
Method · 0.64