Skip to content

[None][chore] Clean up unused and confusing code in moe test #9019

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 1 commit into from
Nov 11, 2025
Merged

[None][chore] Clean up unused and confusing code in moe test #9019

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
27 changes: 8 additions & 19 deletions tests/unittest/_torch/thop/parallel/test_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -263,7 +263,7 @@ def routing_reference_no_aux(expert_logits,


# TopK -> Softmax
def routing_reference_renormalize(expert_logits, top_k, num_experts, padding):
def routing_reference_renormalize(expert_logits, top_k, padding):
topk_values, topk_idx = torch.topk(expert_logits, k=top_k, dim=-1)
topk_values = torch.nn.functional.softmax(topk_values.float(), dim=-1)

Expand All @@ -279,8 +279,7 @@ def routing_reference_renormalize(expert_logits, top_k, num_experts, padding):


# Softmax->TopK -> Normalize
def routing_reference_renormalize_naive(expert_logits, top_k, num_experts,
padding):
def routing_reference_renormalize_naive(expert_logits, top_k, padding):
norm_topk_prob = True
scores = torch.nn.functional.softmax(expert_logits.float(), dim=-1)
topk_values, topk_idx = torch.topk(scores, k=top_k, dim=-1)
Expand Down Expand Up @@ -1002,7 +1001,6 @@ class TestMoeFp4:
{
"num_experts": 256,
"top_k": 8,
"padding": 8,
"n_groups": 8,
"top_k_groups": 4,
"routed_scaling": 2.5,
Expand All @@ -1014,7 +1012,6 @@ class TestMoeFp4:
{
"num_experts": 72,
"top_k": 6,
"padding": 8,
"n_groups": 1,
"top_k_groups": 1,
"routed_scaling": 2.5,
Expand All @@ -1026,7 +1023,6 @@ class TestMoeFp4:
{
"num_experts": 128,
"top_k": 8,
"padding": 8,
"n_groups": None,
"top_k_groups": None,
"routed_scaling": None,
Expand All @@ -1038,7 +1034,6 @@ class TestMoeFp4:
{
"num_experts": 128,
"top_k": 4,
"padding": 8,
"n_groups": None,
"top_k_groups": None,
"routed_scaling": None,
Expand All @@ -1050,7 +1045,6 @@ class TestMoeFp4:
{
"num_experts": 512,
"top_k": 10,
"padding": 8,
"n_groups": None,
"top_k_groups": None,
"routed_scaling": None,
Expand Down Expand Up @@ -1080,7 +1074,6 @@ def test_autotune(self, num_tokens, hidden_size, intermediate_size,
{
"num_experts": 72,
"top_k": 6,
"padding": 8,
"n_groups": 1,
"top_k_groups": 1,
"routed_scaling": 2.5,
Expand Down Expand Up @@ -1110,7 +1103,6 @@ def test_autotune_fp8_fp4(self, num_tokens, hidden_size, intermediate_size,
{
"num_experts": 256,
"top_k": 8,
"padding": 8,
"n_groups": 8,
"top_k_groups": 4,
"routed_scaling": 2.5,
Expand All @@ -1122,7 +1114,6 @@ def test_autotune_fp8_fp4(self, num_tokens, hidden_size, intermediate_size,
{
"num_experts": 128,
"top_k": 4,
"padding": 8,
"n_groups": None,
"top_k_groups": None,
"routed_scaling": None,
Expand All @@ -1134,7 +1125,6 @@ def test_autotune_fp8_fp4(self, num_tokens, hidden_size, intermediate_size,
{
"num_experts": 512,
"top_k": 10,
"padding": 8,
"n_groups": None,
"top_k_groups": None,
"routed_scaling": None,
Expand Down Expand Up @@ -1166,7 +1156,6 @@ def test_no_autotune(self, num_tokens, hidden_size, intermediate_size,
{
"num_experts": 128,
"top_k": 4,
"padding": 8,
"n_groups": None,
"top_k_groups": None,
"routed_scaling": None,
Expand Down Expand Up @@ -1305,10 +1294,10 @@ def run_moe_fp4_test(self, num_tokens: int, hidden_size: int,
routed_scaling, padding)
elif routing_method_type == RoutingMethodType.Renormalize:
permute_info, scores = routing_reference_renormalize(
expert_logits, top_k, num_experts, padding)
expert_logits, top_k, padding)
elif routing_method_type == RoutingMethodType.RenormalizeNaive:
permute_info, scores = routing_reference_renormalize_naive(
expert_logits, top_k, num_experts, padding)
expert_logits, top_k, padding)

args = moe_args(num_tokens, num_experts, hidden_size, intermediate_size,
top_k, padding, hidden_states_fp4_bytes,
Expand Down Expand Up @@ -1552,10 +1541,10 @@ def run_moe_fp8_fp4_test(self, num_tokens: int, hidden_size: int,
routed_scaling, padding)
elif routing_method_type == RoutingMethodType.Renormalize:
permute_info, scores = routing_reference_renormalize(
expert_logits, top_k, num_experts, padding)
expert_logits, top_k, padding)
elif routing_method_type == RoutingMethodType.RenormalizeNaive:
permute_info, scores = routing_reference_renormalize_naive(
expert_logits, top_k, num_experts, padding)
expert_logits, top_k, padding)

args = moe_args(num_tokens, num_experts, hidden_size, intermediate_size,
top_k, padding, hidden_states_fp8, None,
Expand Down Expand Up @@ -2028,10 +2017,10 @@ def test_moe_mxe2m1_weights(num_tokens, hidden_size, intermediate_size,
sf_block_size) # ue8m0 scaling factors
if routing_method_type == RoutingMethodType.Renormalize:
permute_info, scores = routing_reference_renormalize(
expert_logits, top_k, num_experts, padding)
expert_logits, top_k, padding)
elif routing_method_type == RoutingMethodType.RenormalizeNaive:
permute_info, scores = routing_reference_renormalize_naive(
expert_logits, top_k, num_experts, padding)
expert_logits, top_k, padding)
else:
raise ValueError("Invalid routing method type")

Expand Down
Loading