Skip to content

Update attention template #32

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

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Update attention template #32

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
26 changes: 13 additions & 13 deletions attentionbench/attention_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -76,14 +76,15 @@ def get_lowering_config(self) -> str:
+ f">"
)

def get_mma_schedule(self) -> str:
return (
f"#iree_gpu.mma_schedule<"
+ f"intrinsic = #iree_gpu.mma_layout<{self.intrinsic}>"
+ f", subgroup_m_count = {self.M_warp}"
+ f", subgroup_n_count = {self.N_warp}"
+ f">"
)
def get_lowering_config_for_mmt(self, extra_args) -> str:
base_str = (f"#iree_gpu.lowering_config<{{"
+ f"mma_kind = #iree_gpu.mma_layout<{self.intrinsic}>"
+ f", subgroup_m_count = {self.M_warp}"
+ f", subgroup_n_count = {self.N_warp}")
for arg in extra_args:
base_str += f", {arg}"
base_str += "}>"
return base_str

def get_translation_info(self) -> str:
llvm_func_attrs = []
Expand All @@ -93,11 +94,10 @@ def get_translation_info(self) -> str:
llvm_func_attrs += [f'"denormal-fp-math-f32" = "preserve-sign"']
return (
f"#iree_codegen.translation_info<"
+ f"LLVMGPUVectorDistribute"
+ f"pipeline = LLVMGPUVectorDistribute"
+ f" workgroup_size = [{self.N_warp * self.M_warp * 64}]"
+ f" subgroup_size = 64"
+ f" ,{{mma_schedule = {self.get_mma_schedule()}"
+ f" , llvm_func_attrs = {{ {','.join(llvm_func_attrs)} }}"
+ f" , {{llvm_func_attrs = {{ {','.join(llvm_func_attrs)} }}"
+ f"}}"
+ f">"
)
Expand Down Expand Up @@ -139,8 +139,8 @@ def generate_mlir(config: AttentionConfig, tuning: Optional[TuningSpec] = None):
%O = iree_linalg_ext.attention
{{ indexing_maps = [#Q, #K, #V, #S, #O]
,decomposition_config = {{
qk_attrs = {{attention_qk_matmul, lowering_config = #iree_gpu.lowering_config<{{promote_operands = [0, 1]}}>}},
pv_attrs = {{attention_pv_matmul, lowering_config = #iree_gpu.lowering_config<{{promote_operands = [1]}}>}}
qk_attrs = {{attention_qk_matmul, lowering_config = {tuning.get_lowering_config_for_mmt(["promote_operands = [0, 1]"])}}},
pv_attrs = {{attention_pv_matmul, lowering_config = {tuning.get_lowering_config_for_mmt(["promote_operands = [1]"])}}}
}}
{",compilation_info = #tuning" if tuning and config.dtype == "f16" else ""}
}}
Expand Down
Loading