[CI] Upgrade vllm to 0.11.1

.github/workflows/format_pr_body.yaml

@@ -36,7 +36,7 @@ jobs:
 
       - name: Get vLLM version
         run: |
-          VLLM_COMMIT=v0.11.0
+          VLLM_COMMIT=releases/v0.11.1
           echo "VLLM_COMMIT=https://github.com/vllm-project/vllm/commit/$VLLM_COMMIT" >> $GITHUB_ENV
 
       - name: Checkout repository

.github/workflows/vllm_ascend_test.yaml

@@ -42,7 +42,7 @@ jobs:
   lint:
     uses: ./.github/workflows/pre-commit.yml
     with:
-      vllm: v0.11.0
+      vllm: releases/v0.11.1
 
   changes:
     runs-on: ubuntu-latest
@@ -83,7 +83,7 @@ jobs:
         VLLM_USE_MODELSCOPE: True
     strategy:
       matrix:
-        vllm_version: [v0.11.0]
+        vllm_version: [releases/v0.11.1, v0.11.0]
     steps:
       - name: Install packages
         run: |
@@ -128,6 +128,7 @@ jobs:
           --ignore=tests/ut/kv_connector/test_remote_decode_lifecycle.py \
           --ignore=tests/ut/kv_connector/test_remote_prefill_lifecycle.py \
           --ignore=tests/ut/torchair/models/test_torchair_deepseek_v2.py \
+          --ignore=tests/ut/models/test_deepseek_v2.py \
 
       - name: Upload coverage to Codecov
         # only upload coverage when commits merged
@@ -144,7 +145,7 @@ jobs:
     name: e2e-light
     strategy:
       matrix:
-        vllm_version: [v0.11.0]
+        vllm_version: [releases/v0.11.1, v0.11.0]
     # Note (yikun): If CI resource are limited we can split job into two chain jobs
     needs: [lint, changes]
     # only trigger e2e test after lint passed and the change is e2e related with pull request.

.github/workflows/vllm_ascend_test_full.yaml

@@ -69,7 +69,7 @@ jobs:
     name: e2e-full
     strategy:
       matrix:
-        vllm_version: [v0.11.0]
+        vllm_version: [releases/v0.11.1, v0.11.0]
     needs: [changes]
     if: ${{ needs.changes.outputs.e2e_tracker == 'true' }}
     uses: ./.github/workflows/_e2e_test.yaml

vllm_ascend/core/schedule_config.py

@@ -59,7 +59,7 @@ def initialize_from_config(
                 scheduler_config[k] = getattr(ascend_scheduler_config, k)
         return cls(**scheduler_config)
 
-    def __post_init__(self) -> None:
+    def __post_init__(self, *args) -> None:
         self.max_num_encoder_input_tokens = self.max_num_batched_tokens
         self.encoder_cache_size = self.max_num_batched_tokens
         self.chunked_prefill_enabled = self.enable_chunked_prefill

vllm_ascend/core/scheduler.py

@@ -16,7 +16,7 @@
 #
 import time
 from collections import deque
-from typing import Iterable, Union
+from typing import Iterable, Optional, Union
 
 from vllm.config import VllmConfig
 from vllm.distributed.kv_events import KVEventBatch
@@ -32,27 +32,19 @@
 from vllm.v1.request import Request, RequestStatus
 from vllm.v1.structured_output import StructuredOutputManager
 
+from vllm_ascend.utils import vllm_version_is
+
 
 class AscendScheduler(Scheduler):
     """This Scheduler extends vllm's original v1 scheduler
     with prefill-first scheduling strategy."""
 
-    def __init__(
-        self,
-        vllm_config: VllmConfig,
-        kv_cache_config: KVCacheConfig,
-        structured_output_manager: StructuredOutputManager,
-        mm_registry: MultiModalRegistry = MULTIMODAL_REGISTRY,
-        include_finished_set: bool = False,
-        log_stats: bool = False,
-    ) -> None:
-        super().__init__(vllm_config, kv_cache_config,
-                         structured_output_manager, mm_registry,
-                         include_finished_set, log_stats)
+    def _initialize_common(self) -> None:
+        """Initialize common attributes shared across all versions."""
         self.scheduled_req_ids: set[str] = set()
         self.running: list[Request] = []
-
         self.finished_prefill_reqs: deque[Request] = deque()
+
         enable_pd_transfer = getattr(self.scheduler_config,
                                      'enable_pd_transfer', False)
         decode_max_num_seqs = getattr(self.scheduler_config,
@@ -61,6 +53,29 @@ def __init__(
         self.decode_max_num_running_reqs = max(self.max_num_running_reqs,
                                                decode_max_num_seqs)
 
+    def __init__(
+        self,
+        vllm_config: VllmConfig,
+        kv_cache_config: KVCacheConfig,
+        structured_output_manager: StructuredOutputManager,
+        block_size: Optional[int] = None,
+        mm_registry: MultiModalRegistry = MULTIMODAL_REGISTRY,
+        include_finished_set: bool = False,
+        log_stats: bool = False,
+    ) -> None:
+        # Call the parent class's __init__ method
+        if vllm_version_is("0.11.0"):
+            super().__init__(vllm_config, kv_cache_config,
+                             structured_output_manager, mm_registry,
+                             include_finished_set, log_stats)
+        else:
+            super().__init__(vllm_config, kv_cache_config,
+                             structured_output_manager, block_size,
+                             mm_registry, include_finished_set, log_stats)
+
+        # Initialize common attributes
+        self._initialize_common()
+
     def schedule(self) -> SchedulerOutput:
         if self.scheduler_config.chunked_prefill_enabled:
             return super().schedule()
@@ -440,9 +455,14 @@ def skip_cur_request():
             self.kv_cache_config.kv_cache_groups)
         if self.running:
             any_request = self.running[0]
-            num_common_prefix_blocks = (
-                self.kv_cache_manager.get_num_common_prefix_blocks(
-                    any_request, len(self.running)))
+            if vllm_version_is("0.11.0"):
+                num_common_prefix_blocks = (
+                    self.kv_cache_manager.get_num_common_prefix_blocks(
+                        any_request, len(self.running)))
+            else:
+                num_common_prefix_blocks = (
+                    self.kv_cache_manager.get_num_common_prefix_blocks(
+                        any_request.request_id))
 
         # Construct the scheduler output.
         new_reqs_data = [

vllm_ascend/models/deepseek_v2.py

@@ -45,7 +45,6 @@
                                                ReplicatedLinear,
                                                RowParallelLinear)
 from vllm.model_executor.layers.logits_processor import LogitsProcessor
-from vllm.model_executor.layers.mla import MultiHeadLatentAttention
 from vllm.model_executor.layers.quantization import QuantizationConfig
 from vllm.model_executor.layers.rotary_embedding import get_rope
 from vllm.model_executor.layers.vocab_parallel_embedding import (
@@ -70,6 +69,13 @@
                                            AscendSparseFlashAttention, Indexer)
 from vllm_ascend.ops.common_fused_moe import AscendFusedMoE
 from vllm_ascend.ops.linear import AscendLinearBase
+from vllm_ascend.utils import vllm_version_is
+
+if vllm_version_is("0.11.0"):
+    from vllm.model_executor.layers.mla import MultiHeadLatentAttention
+else:
+    from vllm.model_executor.layers.mla import \
+        MultiHeadLatentAttentionWrapper as MultiHeadLatentAttention
 
 
 @support_torch_compile

vllm_ascend/models/layers/mla.py

@@ -27,10 +27,17 @@
 from vllm.attention import Attention, AttentionMetadata
 from vllm.config import CacheConfig, get_current_vllm_config
 from vllm.forward_context import ForwardContext, get_forward_context
-from vllm.model_executor.layers.mla import MultiHeadLatentAttention
 from vllm.model_executor.layers.quantization import QuantizationConfig
 from vllm.utils import direct_register_custom_op
 
+from vllm_ascend.utils import vllm_version_is
+
+if vllm_version_is("0.11.0"):
+    from vllm.model_executor.layers.mla import MultiHeadLatentAttention
+else:
+    from vllm.model_executor.layers.mla import \
+        MultiHeadLatentAttentionWrapper as MultiHeadLatentAttention
+
 
 @dataclass
 class AscendMLAModules:

vllm_ascend/models/layers/sfa.py

@@ -28,10 +28,17 @@
 from vllm.config import CacheConfig, get_current_vllm_config
 from vllm.forward_context import ForwardContext, get_forward_context
 from vllm.model_executor.layers.linear import ReplicatedLinear
-from vllm.model_executor.layers.mla import MultiHeadLatentAttention
 from vllm.model_executor.layers.quantization import QuantizationConfig
 from vllm.utils import direct_register_custom_op
 
+from vllm_ascend.utils import vllm_version_is
+
+if vllm_version_is("0.11.0"):
+    from vllm.model_executor.layers.mla import MultiHeadLatentAttention
+else:
+    from vllm.model_executor.layers.mla import \
+        MultiHeadLatentAttentionWrapper as MultiHeadLatentAttention
+
 
 @dataclass
 class AscendSFAModules:

vllm_ascend/ops/common_fused_moe.py

@@ -28,7 +28,6 @@
 from vllm.model_executor.layers.fused_moe.layer import (
     FusedMoE, UnquantizedFusedMoEMethod, determine_expert_map,
     get_compressed_expert_map)
-from vllm.model_executor.layers.shared_fused_moe import SharedFusedMoE
 
 from vllm_ascend.ascend_config import get_ascend_config
 from vllm_ascend.ascend_forward_context import MoECommType
@@ -40,7 +39,13 @@
 from vllm_ascend.ops.moe.moe_comm_method import setup_moe_comm_method
 from vllm_ascend.utils import (ACL_FORMAT_FRACTAL_NZ, enable_sp, is_310p,
                                is_enable_nz, npu_stream_switch,
-                               shared_expert_dp_enabled)
+                               shared_expert_dp_enabled, vllm_version_is)
+
+if vllm_version_is("0.11.0"):
+    from vllm.model_executor.layers.shared_fused_moe import SharedFusedMoE  # type: ignore # isort:skip
+else:
+    from vllm.model_executor.layers.fused_moe.shared_fused_moe import \
+        SharedFusedMoE
 
 
 class AscendUnquantizedFusedMoEMethod(UnquantizedFusedMoEMethod):