wip

Amir-19 · Amir-19 · commit 97d8cb56e339 · 2025-09-29T21:08:53.000-07:00
diff --git a/tests/test_cute_dsl_blockscaled_gemm_allreduce_two_shot.py b/tests/test_cute_dsl_blockscaled_gemm_allreduce_two_shot.py
@@ -38,7 +38,7 @@ def create_mc_tensor(torch_tensor_cpu, dtype, is_dynamic_layout=True):
         dtype=torch_tensor_cpu.dtype
     )
 
-    torch_symm_tensor_lmn.copy_(torch_tensor_cpu)
+    torch_symm_tensor_lmn.copy_(torch_tensor_cpu_lmn)
     torch_symm_tensor = torch_symm_tensor_lmn.permute(1, 2, 0)
     symm = symm_mem.rendezvous(torch_symm_tensor, group=dist.group.WORLD.group_name)
     mc_ptr = symm.multicast_ptr

Original file line number	Diff line number	Diff line change
`@@ -38,7 +38,7 @@ def create_mc_tensor(torch_tensor_cpu, dtype, is_dynamic_layout=True):`
`38`	`38`	`dtype=torch_tensor_cpu.dtype`
`39`	`39`	`)`
`40`	`40`
`41`		`- torch_symm_tensor_lmn.copy_(torch_tensor_cpu)`
	`41`	`+ torch_symm_tensor_lmn.copy_(torch_tensor_cpu_lmn)`
`42`	`42`	`torch_symm_tensor = torch_symm_tensor_lmn.permute(1, 2, 0)`
`43`	`43`	`symm = symm_mem.rendezvous(torch_symm_tensor, group=dist.group.WORLD.group_name)`
`44`	`44`	`mc_ptr = symm.multicast_ptr`