fix lightllm/server/router/model_infer/mode_backend/generic_pre_process.py

Author: flyinglandlord

lightllm/server/router/model_infer/mode_backend/generic_pre_process.py

@@ -7,7 +7,6 @@
 from lightllm.utils.envs_utils import (
     enable_diverse_mode_gqa_decode_fast_kernel,
     enable_triton_mtp_kernel,
-    enable_triton_mtp_kernel,
     get_diverse_max_batch_shared_group_size,
     enable_dynamic_mtp_verify,
     get_env_start_args
@@ -240,30 +239,3 @@ def build_mtp_shared_group_infos(
     if current_group:
         b_mark_shared_group.extend([0] * (len(current_group) - 1) + [len(current_group)])
     return torch.tensor(b_mark_shared_group, dtype=torch.int32, device="cpu")
-
-
-def build_mtp_shared_group_infos(
-    b_mtp_index: torch.Tensor,
-) -> torch.Tensor:
-    # Similar to build_diverse_shared_group_infos, 
-    # but the grouping logic is based on b_mtp_index, which indicates the MTP step of each request
-    max_batch_shared_group_size = get_diverse_max_batch_shared_group_size()
-    b_mark_shared_group = []
-    num_reqs = b_mtp_index.shape[0]
-    if num_reqs == 0:
-        return torch.zeros_like(b_mtp_index, dtype=torch.int32, device="cpu")
-    current_group = []
-    for i in range(num_reqs):
-        step = b_mtp_index[i].item()
-        if len(current_group) == 0:
-            current_group.append(i)
-        else:
-            prev_step = b_mtp_index[i - 1].item()
-            if step == prev_step + 1 and len(current_group) < max_batch_shared_group_size:
-                current_group.append(i)
-            else:
-                b_mark_shared_group.extend([0] * (len(current_group) - 1) + [len(current_group)])
-                current_group = [i]
-    if current_group:
-        b_mark_shared_group.extend([0] * (len(current_group) - 1) + [len(current_group)])
-    return torch.tensor(b_mark_shared_group, dtype=torch.int32, device="cpu")