Unverified Commit 4a229ddb authored by Yuqi Dong's avatar Yuqi Dong Committed by GitHub
Browse files

[Bugfix] Fallback `torch.accelerator.synchronize()` to `torch.cuda.synchronize()` (#987)

* [Refactor]:Add support for torch version lower than 2.6.0

* update
parent fc41463c
...@@ -97,7 +97,7 @@ def do_bench( ...@@ -97,7 +97,7 @@ def do_bench(
# Initial function call and synchronization # Initial function call and synchronization
fn() fn()
torch.accelerator.synchronize() torch.cuda.synchronize()
# Create L2 cache flush buffer (256 MB) # Create L2 cache flush buffer (256 MB)
# Fast flush uses int32 (4 bytes), regular uses int8 (1 byte) # Fast flush uses int32 (4 bytes), regular uses int8 (1 byte)
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment