Skip to content

[Softmax] Update Online Softmax bindings #155

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Nov 29, 2024
Merged

[Softmax] Update Online Softmax bindings #155

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
f93b323
Update softmax.cu
DefTruth Nov 29, 2024
52db58e
Update softmax.py
DefTruth Nov 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 7 additions & 1 deletion kernels/softmax/softmax.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -600,9 +600,15 @@ online_safe_softmax_f32x4_pack_per_token_kernel<(H/4)><<<grid, block>>>( \
case 1024: \
LANUCH_ONLINE_SOFTMAX_F32X4_PACK_PER_TOKEN_KERNEL(1024) \
break; \
case 2048: \
LANUCH_ONLINE_SOFTMAX_F32X4_PACK_PER_TOKEN_KERNEL(2048) \
break; \
case 4096: \
LANUCH_ONLINE_SOFTMAX_F32X4_PACK_PER_TOKEN_KERNEL(4096) \
break; \
default: \
throw std::runtime_error( \
"only support H: 128/256/512/1024; raise error if warp_num*4 > H"); \
"only support H: 128/256/.../4096;"); \
break; \
}

Expand Down
20 changes: 11 additions & 9 deletions kernels/softmax/softmax.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,7 +24,7 @@

def run_benchmark(perf_func: callable, x: torch.Tensor,
tag: str, out: Optional[torch.Tensor] = None,
warmup: int = 10, iters: int = 1000,
warmup: int = 10, iters: int = 100,
show_all: bool = False):
if out is not None:
out.fill_(0)
Expand Down Expand Up @@ -60,7 +60,7 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
N = 128 * 128
print(" " * 45 + f"N={N}")
print("-" * 100)
x = torch.randn((N)).cuda().float()
x = torch.randn((N), device="cuda").cuda().float()
out = torch.zeros_like(x).cuda().float().contiguous()
run_benchmark(lib.softmax_f32, x, "f32(fence)", out)
run_benchmark(lib.softmax_f32x4, x, "f32x4(fence)", out)
Expand All @@ -71,7 +71,7 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
S, H = 4096, 256
print(" " * 45 + f"S={S}, H={H}")
print("-" * 100)
x = torch.randn((S, H)).cuda().float().contiguous()
x = torch.randn((S, H), device="cuda").cuda().float().contiguous()
out = torch.zeros_like(x).cuda().float().contiguous()
run_benchmark(lib.softmax_f32_per_token, x, "f32(per)", out)
run_benchmark(lib.softmax_f32x4_per_token, x, "f32x4(per)", out)
Expand All @@ -95,7 +95,7 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
S, H = 4096, 512
print(" " * 45 + f"S={S}, H={H}")
print("-" * 100)
x = torch.randn((S, H)).cuda().float().contiguous()
x = torch.randn((S, H), device="cuda").cuda().float().contiguous()
out = torch.zeros_like(x).cuda().float().contiguous()
run_benchmark(lib.softmax_f32_per_token, x, "f32(per)", out)
run_benchmark(lib.softmax_f32x4_per_token, x, "f32x4(per)", out)
Expand All @@ -119,7 +119,7 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
S, H = 4096, 1024
print(" " * 45 + f"S={S}, H={H}")
print("-" * 100)
x = torch.randn((S, H)).cuda().float().contiguous()
x = torch.randn((S, H), device="cuda").cuda().float().contiguous()
out = torch.zeros_like(x).cuda().float().contiguous()
run_benchmark(lib.softmax_f32_per_token, x, "f32(per)", out)
run_benchmark(lib.softmax_f32x4_per_token, x, "f32x4(per)", out)
Expand All @@ -143,10 +143,11 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
S, H = 4096, 2048
print(" " * 45 + f"S={S}, H={H}")
print("-" * 100)
x = torch.randn((S, H)).cuda().float().contiguous()
x = torch.randn((S, H), device="cuda").cuda().float().contiguous()
out = torch.zeros_like(x).cuda().float().contiguous()
run_benchmark(lib.softmax_f32x4_per_token, x, "f32x4(per)", out)
run_benchmark(lib.safe_softmax_f32x4_per_token, x, "f32x4(safe)", out)
run_benchmark(lib.online_safe_softmax_f32x4_pack_per_token, x, "f32x4(safe+online)", out)
run_benchmark(partial(torch.softmax, dim=1, out=out), x, "f32_th(per)")

print("-" * 100)
Expand All @@ -162,10 +163,11 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
S, H = 4096, 4096
print(" " * 45 + f"S={S}, H={H}")
print("-" * 100)
x = torch.randn((S, H)).cuda().float().contiguous()
x = torch.randn((S, H), device="cuda").cuda().float().contiguous()
out = torch.zeros_like(x).cuda().float().contiguous()
run_benchmark(lib.softmax_f32x4_per_token, x, "f32x4(per)", out)
run_benchmark(lib.safe_softmax_f32x4_per_token, x, "f32x4(safe)", out)
run_benchmark(lib.online_safe_softmax_f32x4_pack_per_token, x, "f32x4(safe+online)", out)
run_benchmark(partial(torch.softmax, dim=1, out=out), x, "f32_th(per)")

print("-" * 100)
Expand All @@ -180,7 +182,7 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
S, H = 4096, 8192
print(" " * 45 + f"S={S}, H={H}")
print("-" * 100)
x = torch.randn((S, H)).cuda().float().contiguous()
x = torch.randn((S, H), device="cuda").cuda().float().contiguous()
out = torch.zeros_like(x).cuda().float().contiguous()
x_f16 = x.half().contiguous()
out_f16 = out.half().contiguous()
Expand All @@ -192,7 +194,7 @@ def run_benchmark(perf_func: callable, x: torch.Tensor,
S, H = 8192, 8192
print(" " * 45 + f"S={S}, H={H}")
print("-" * 100)
x = torch.randn((S, H)).cuda().float().contiguous()
x = torch.randn((S, H), device="cuda").cuda().float().contiguous()
out = torch.zeros_like(x).cuda().float().contiguous()
x_f16 = x.half().contiguous()
out_f16 = out.half().contiguous()
Expand Down