fix fake functions

Author: Varun Sundar Rabindranath

vllm/lora/ops/v1/lora_expand.py

@@ -219,6 +219,7 @@ def lora_expand_fake(
     inputs: torch.Tensor,
     lora_b_weights: torch.Tensor,
     output_tensor: torch.Tensor,
+    token_lora_mapping: torch.Tensor,
     token_indices_sorted_by_lora_ids: torch.Tensor,
     num_tokens_per_lora: torch.Tensor,
     lora_token_start_loc: torch.Tensor,

vllm/lora/ops/v1/lora_expand_slice.py

@@ -222,11 +222,11 @@ def _lora_expand_slice(
     )
     return
 
-
 def lora_expand_slice_fake(
     inputs: torch.Tensor,
     lora_b_weights: torch.Tensor,
     output_tensor: torch.Tensor,
+    token_lora_mapping: torch.Tensor,
     token_indices_sorted_by_lora_ids: torch.Tensor,
     num_tokens_per_lora: torch.Tensor,
     lora_token_start_loc: torch.Tensor,

vllm/lora/ops/v1/lora_shrink.py

@@ -227,11 +227,11 @@ def _lora_shrink(
     )
     return
 
-
 def lora_shrink_fake(
     inputs: torch.Tensor,
     lora_a_weights: torch.Tensor,
     output_tensor: torch.Tensor,
+    token_lora_mapping: torch.Tensor,
     token_indices_sorted_by_lora_ids: torch.Tensor,
     num_tokens_per_lora: torch.Tensor,
     lora_token_start_loc: torch.Tensor,