MCPcopy Create free account
hub / github.com/tile-ai/tilelang / inf_init_kernel

Function inf_init_kernel

testing/python/amd/test_tilelang_hip_codegen.py:436–444  ·  view source on GitHub ↗
()

Source from the content-addressed store, hash-verified

434
435 @tilelang.jit
436 def inf_init_kernel():
437 @T.prim_func
438 def kernel(out: T.Tensor((N,), T.float32)) -> None:
439 with T.Kernel(1, threads=N) as _:
440 v = T.alloc_var(T.float32, init=-T.infinity(T.float32))
441 for i in T.Parallel(N):
442 out[i] = v
443
444 return kernel
445
446 out = torch.zeros(N, dtype=torch.float32, device="cuda")
447 inf_init_kernel()(out)

Callers 1

test_alloc_var_inf_initFunction · 0.85

Calls

no outgoing calls

Tested by

no test coverage detected

Used in the wild real call sites across dependent graphs

searching dependent graphs…