Skip to content

GitLab

Commit 7102738f authored by zhuwenwen's avatar zhuwenwen
Browse files

update activation

parent e5fc1b17
Show whitespace changes
Inline Side-by-side
Showing with 6 additions and 12 deletions
vllm/model_executor/layers/fused_moe/fused_moe.py
View file @ 7102738f
...@@ -1593,8 +1593,7 @@ def inplace_fused_experts( ...@@ -1593,8 +1593,7 @@ def inplace_fused_experts(
w2: torch.Tensor, w2: torch.Tensor,
topk_weights: torch.Tensor, topk_weights: torch.Tensor,
topk_ids: torch.Tensor, topk_ids: torch.Tensor,
# activation: str = "silu", activation: str = "silu",
activation: Optional[str] = None,
apply_router_weight_on_input: bool = False, apply_router_weight_on_input: bool = False,
use_fp8_w8a8: bool = False, use_fp8_w8a8: bool = False,
use_int8_w8a8: bool = False, use_int8_w8a8: bool = False,
...@@ -1630,8 +1629,7 @@ def inplace_fused_experts_fake( ...@@ -1630,8 +1629,7 @@ def inplace_fused_experts_fake(
w2: torch.Tensor, w2: torch.Tensor,
topk_weights: torch.Tensor, topk_weights: torch.Tensor,
topk_ids: torch.Tensor, topk_ids: torch.Tensor,
# activation: str = "silu", activation: str = "silu",
activation: Optional[str] = None,
apply_router_weight_on_input: bool = False, apply_router_weight_on_input: bool = False,
use_fp8_w8a8: bool = False, use_fp8_w8a8: bool = False,
use_int8_w8a8: bool = False, use_int8_w8a8: bool = False,
...@@ -1672,8 +1670,7 @@ def outplace_fused_experts( ...@@ -1672,8 +1670,7 @@ def outplace_fused_experts(
w2: torch.Tensor, w2: torch.Tensor,
topk_weights: torch.Tensor, topk_weights: torch.Tensor,
topk_ids: torch.Tensor, topk_ids: torch.Tensor,
# activation: str = "silu", activation: str = "silu",
activation: Optional[str] = None,
apply_router_weight_on_input: bool = False, apply_router_weight_on_input: bool = False,
use_fp8_w8a8: bool = False, use_fp8_w8a8: bool = False,
use_int8_w8a8: bool = False, use_int8_w8a8: bool = False,
...@@ -1709,8 +1706,7 @@ def outplace_fused_experts_fake( ...@@ -1709,8 +1706,7 @@ def outplace_fused_experts_fake(
w2: torch.Tensor, w2: torch.Tensor,
topk_weights: torch.Tensor, topk_weights: torch.Tensor,
topk_ids: torch.Tensor, topk_ids: torch.Tensor,
# activation: str = "silu", activation: str = "silu",
activation: Optional[str] = None,
use_fp8_w8a8: bool = False, use_fp8_w8a8: bool = False,
use_int8_w8a8: bool = False, use_int8_w8a8: bool = False,
use_int8_w8a16: bool = False, use_int8_w8a16: bool = False,
...@@ -1768,8 +1764,7 @@ def fused_experts( ...@@ -1768,8 +1764,7 @@ def fused_experts(
topk_weights: torch.Tensor, topk_weights: torch.Tensor,
topk_ids: torch.Tensor, topk_ids: torch.Tensor,
inplace: bool = False, inplace: bool = False,
# activation: str = "silu", activation: str = "silu",
activation: Optional[str] = None,
apply_router_weight_on_input: bool = False, apply_router_weight_on_input: bool = False,
global_num_experts: int = -1, global_num_experts: int = -1,
expert_map: Optional[torch.Tensor] = None, expert_map: Optional[torch.Tensor] = None,
...@@ -1888,8 +1883,7 @@ def fused_experts_impl( ...@@ -1888,8 +1883,7 @@ def fused_experts_impl(
topk_weights: torch.Tensor, topk_weights: torch.Tensor,
topk_ids: torch.Tensor, topk_ids: torch.Tensor,
inplace: bool = False, inplace: bool = False,
# activation: str = "silu", activation: str = "silu",
activation: Optional[str] = None,
apply_router_weight_on_input: bool = False, apply_router_weight_on_input: bool = False,
use_fp8_w8a8: bool = False, use_fp8_w8a8: bool = False,
use_int8_w8a8: bool = False, use_int8_w8a8: bool = False,
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment