biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Hardware] Replace torch.cuda.empty_cache with torch.accelerator.empty_cache (#30681)

Browse Source 
Signed-off-by: Kunshang Ji <kunshang.ji@intel.com>
Signed-off-by: Kunshang Ji <jikunshang95@gmail.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
This commit is contained in:
Kunshang Ji
2026-03-04 17:49:47 +08:00
committed by GitHub
parent 5dc3538736
commit 16d2ad1d38
35 changed files with 110 additions and 59 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
tests/v1/sample/test_logprobs.py
View File

@@ -530,7 +530,7 @@ def test_logprobs_mode(logprobs_mode: LogprobsMode):
assert positive_values > 0
finally:
del llm
torch.cuda.empty_cache()
torch.accelerator.empty_cache()
cleanup_dist_env_and_memory()
@@ -1065,7 +1065,7 @@ def test_spec_decode_logprobs(
for logprobs in output.logprobs:
ref_logprobs.extend(logprobs.values())
del ref_llm
torch.cuda.empty_cache()
torch.accelerator.empty_cache()
cleanup_dist_env_and_memory()
# Run spec decode LLM.
@@ -1095,7 +1095,7 @@ def test_spec_decode_logprobs(
for logprobs in output.logprobs:
spec_logprobs.extend(logprobs.values())
del spec_llm
torch.cuda.empty_cache()
torch.accelerator.empty_cache()
cleanup_dist_env_and_memory()
# Per-token logprobs are expected to be the same.