Remove unused kwargs from model definitions (#13555)

vllm/model_executor/models/chameleon.py

@@ -1,7 +1,7 @@
 # SPDX-License-Identifier: Apache-2.0
 
 from functools import cached_property
-from typing import (Any, Dict, Iterable, List, Literal, Mapping, Optional, Set,
+from typing import (Any, Dict, Iterable, Literal, Mapping, Optional, Set,
                     Tuple, TypedDict, Union)
 
 import torch
@@ -10,7 +10,7 @@ import torch.nn.functional as F
 from transformers import (BatchFeature, ChameleonConfig, ChameleonProcessor,
                           ChameleonVQVAEConfig)
 
-from vllm.attention import Attention, AttentionMetadata
+from vllm.attention import Attention
 from vllm.config import CacheConfig, VllmConfig
 from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size
 from vllm.logger import init_logger
@@ -310,15 +310,13 @@ class ChameleonAttention(nn.Module):
         self,
         positions: torch.Tensor,
         hidden_states: torch.Tensor,
-        kv_cache: torch.Tensor,
-        attn_metadata: AttentionMetadata,
     ) -> torch.Tensor:
         qkv, _ = self.qkv_proj(hidden_states)
         q, k, v = qkv.split([self.q_size, self.kv_size, self.kv_size], dim=-1)
         q, k = self._apply_qk_norm(q, k)
 
         q, k = self.rotary_emb(positions, q, k)
-        attn_output = self.attn(q, k, v, kv_cache, attn_metadata)
+        attn_output = self.attn(q, k, v)
         output, _ = self.o_proj(attn_output)
         return output
 
@@ -372,8 +370,6 @@ class ChameleonDecoderLayer(nn.Module):
         self,
         positions: torch.Tensor,
         hidden_states: torch.Tensor,
-        kv_cache: torch.Tensor,
-        attn_metadata: AttentionMetadata,
         residual: Optional[torch.Tensor],
     ) -> Tuple[torch.Tensor, torch.Tensor]:
 
@@ -386,8 +382,6 @@ class ChameleonDecoderLayer(nn.Module):
         hidden_states = self.self_attn(
             positions=positions,
             hidden_states=hidden_states,
-            kv_cache=kv_cache,
-            attn_metadata=attn_metadata,
         )
 
         # Fully Connected
@@ -447,8 +441,6 @@ class ChameleonSwinDecoderLayer(nn.Module):
         self,
         positions: torch.Tensor,
         hidden_states: torch.Tensor,
-        kv_cache: torch.Tensor,
-        attn_metadata: AttentionMetadata,
         residual: Optional[torch.Tensor],
     ) -> Tuple[torch.Tensor, torch.Tensor]:
 
@@ -456,8 +448,6 @@ class ChameleonSwinDecoderLayer(nn.Module):
         hidden_states = self.self_attn(
             positions=positions,
             hidden_states=hidden_states,
-            kv_cache=kv_cache,
-            attn_metadata=attn_metadata,
         )
 
         hidden_states = self.input_layernorm(hidden_states)
@@ -906,8 +896,6 @@ class ChameleonModel(nn.Module):
         self,
         input_ids: Optional[torch.Tensor],
         positions: torch.Tensor,
-        kv_caches: List[torch.Tensor],
-        attn_metadata: AttentionMetadata,
         intermediate_tensors: Optional[IntermediateTensors],
         inputs_embeds: Optional[torch.Tensor] = None,
     ) -> Union[torch.Tensor, IntermediateTensors]:
@@ -921,13 +909,10 @@ class ChameleonModel(nn.Module):
             assert intermediate_tensors is not None
             hidden_states = intermediate_tensors["hidden_states"]
             residual = intermediate_tensors["residual"]
-        for i in range(self.start_layer, self.end_layer):
-            layer = self.layers[i]
+        for layer in self.layers[self.start_layer:self.end_layer]:
             hidden_states, residual = layer(
                 positions,
                 hidden_states,
-                kv_caches[i - self.start_layer],
-                attn_metadata,
                 residual,
             )
         if not get_pp_group().is_last_rank:
@@ -1028,8 +1013,6 @@ class ChameleonForConditionalGeneration(nn.Module, SupportsMultiModal,
         self,
         input_ids: torch.Tensor,
         positions: torch.Tensor,
-        kv_caches: List[torch.Tensor],
-        attn_metadata: AttentionMetadata,
         intermediate_tensors: Optional[IntermediateTensors] = None,
         inputs_embeds: Optional[torch.Tensor] = None,
         **kwargs,
@@ -1048,8 +1031,6 @@ class ChameleonForConditionalGeneration(nn.Module, SupportsMultiModal,
 
         hidden_states = self.model(input_ids,
                                    positions,
-                                   kv_caches,
-                                   attn_metadata,
                                    intermediate_tensors,
                                    inputs_embeds=inputs_embeds)
         return hidden_states