Skip to content

fused moe triton tuning script support qwen3 #5842

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 6 commits into from
Apr 28, 2025
Merged

fused moe triton tuning script support qwen3 #5842

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
f70400c
fused moe triton tuning script support qwen23
BBuf Apr 28, 2025
4d1d63b
ipd
BBuf Apr 28, 2025
deae116
ipd
BBuf Apr 28, 2025
9665cd6
upd
BBuf Apr 28, 2025
a0a91b5
upd
BBuf Apr 28, 2025
b8fd895
Merge branch 'main' into fused_moe_triton_tuning_support_qwen3
BBuf Apr 28, 2025
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
7 changes: 7 additions & 0 deletions benchmark/kernels/fused_moe_triton/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,6 +20,13 @@ python benchmark/kernels/fused_moe_triton/tuning_fused_moe_triton.py \
--dtype fp8_w8a8 \
--tune

# Tune Qwen3-235B-A22B-FP8 and TP=4
python benchmark/kernels/fused_moe_triton/tuning_fused_moe_triton.py \
--model Qwen/Qwen3-235B-A22B-FP8 \
--tp-size 4 \
--dtype fp8_w8a8 \
--tune

# Tune DeepSeek-V3 with FP8, TP=8 and n_share_experts_fusion=8
python benchmark/kernels/fused_moe_triton/tuning_fused_moe_triton.py \
--model deepseek-ai/DeepSeek-V3-0324 \
Expand Down
7 changes: 6 additions & 1 deletion benchmark/kernels/fused_moe_triton/benchmark_torch_compile_fused_moe.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,10 +30,15 @@ def get_model_config(model_name: str, tp_size: int):
topk = config.num_experts_per_tok
intermediate_size = config.moe_intermediate_size
shard_intermediate_size = 2 * intermediate_size // tp_size
elif config.architectures[0] == "Qwen3MoeForCausalLM":
E = config.num_experts
topk = config.num_experts_per_tok
intermediate_size = config.moe_intermediate_size
shard_intermediate_size = 2 * intermediate_size // tp_size
elif config.architectures[0] in ["DeepseekV2ForCausalLM", "DeepseekV3ForCausalLM"]:
E = config.n_routed_experts
topk = config.num_experts_per_tok
intermediate_size = config.intermediate_size
intermediate_size = config.moe_intermediate_size
shard_intermediate_size = 2 * intermediate_size // tp_size
elif config.architectures[0] in [
"Grok1ForCausalLM",
Expand Down
5 changes: 5 additions & 0 deletions benchmark/kernels/fused_moe_triton/benchmark_vllm_vs_sglang_fused_moe_triton.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,11 @@ def get_model_config(model_name: str, tp_size: int):
topk = config.num_experts_per_tok
intermediate_size = config.moe_intermediate_size
shard_intermediate_size = 2 * intermediate_size // tp_size
elif config.architectures[0] == "Qwen3MoeForCausalLM":
E = config.num_experts
topk = config.num_experts_per_tok
intermediate_size = config.moe_intermediate_size
shard_intermediate_size = 2 * intermediate_size // tp_size
elif config.architectures[0] in ["DeepseekV2ForCausalLM", "DeepseekV3ForCausalLM"]:
E = config.n_routed_experts
topk = config.num_experts_per_tok
Expand Down
Loading