A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 5, 14:45 UTC
[CI] Fix: pass string cache_dtype in test_register_kv_caches (#38836)
zhanqiuhu committed
Apr 2, 2026 at 15:42 UTC
7b743ba953b094d02cf72a5a8b80a56c546d6c4f
1 file changed
+1
-1
tests/v1/kv_connector/unit/test_nixl_connector.py
+1
-1
@@ -1613,7 +1613,7 @@ def test_register_kv_caches(
1613
)
1614
]
1615
],
1616
- cache_dtype=torch.bfloat16,
1616
+ cache_dtype="bfloat16",
1617
device=torch.accelerator.current_device_index(),
1618
kernel_block_sizes=[block_size],
1619
)