Merge remote-tracking branch 'origin/v0.15.1-dev' into v0.15.1-dev

# Conflicts:
#	vllm/model_executor/layers/fused_moe/config.py
#	vllm/model_executor/layers/fused_moe/layer.py
#	vllm/model_executor/layers/quantization/compressed_tensors/compressed_tensors_marlin.py

vllm/model_executor/models/interfaces.py

@@ -44,10 +44,6 @@ else:
     _ProcessorFactories = object
     IntermediateTensors = object
 
-if TYPE_CHECKING:
-    from vllm.config import LoRAConfig, MultiModalConfig, SchedulerConfig
-    from vllm.sequence import IntermediateTensors
-
 logger = init_logger(__name__)
 
 MultiModalEmbeddings: TypeAlias = list[Tensor] | Tensor | tuple[Tensor, ...]
@@ -607,8 +603,6 @@ class SupportsPP(Protocol):
 
     def forward(
         self,
-        input_ids: Tensor | None,
-        positions: Tensor,
         *,
         intermediate_tensors: IntermediateTensors | None,
     ) -> IntermediateTensors | None:
@@ -637,8 +631,6 @@ class _SupportsPPType(Protocol):
 
     def forward(
         self,
-        input_ids: Tensor | None,
-        positions: Tensor,
         *,
         intermediate_tensors: IntermediateTensors | None,
     ) -> Tensor | IntermediateTensors: ...

vllm/model_executor/models/intern_vit.py

@@ -33,8 +33,6 @@ from vllm.model_executor.layers.linear import (
 )
 from vllm.model_executor.layers.quantization import QuantizationConfig
 from vllm.model_executor.model_loader.weight_utils import default_weight_loader
-import vllm.envs as envs
-
 
 from .vision import run_dp_sharded_vision_model
 

vllm/model_executor/models/internlm2.py

@@ -284,7 +284,7 @@ class InternLM2Model(nn.Module):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,
@@ -350,7 +350,7 @@ class InternLM2ForCausalLM(nn.Module, SupportsPP, SupportsLoRA):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None,
         inputs_embeds: torch.Tensor | None = None,
@@ -446,7 +446,7 @@ class InternLM2ForRewardModel(InternLM2ForCausalLM):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/internlm2_ve.py

@@ -101,7 +101,7 @@ class InternLM2VEModel(InternLM2Model):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/interns1.py

@@ -782,7 +782,7 @@ class InternS1ForConditionalGeneration(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/internvl.py

@@ -1371,7 +1371,7 @@ class InternVLChatModel(nn.Module, SupportsMultiModal, SupportsPP, SupportsLoRA)
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/iquest_loopcoder.py

@@ -438,7 +438,7 @@ class IQuestLoopCoderModel(nn.Module):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,
@@ -570,7 +570,7 @@ class IQuestLoopCoderForCausalLM(nn.Module):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/isaac.py

@@ -1450,7 +1450,7 @@ class IsaacForConditionalGeneration(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/jais.py

@@ -280,7 +280,7 @@ class JAISModel(nn.Module):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         position_ids: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,
@@ -344,7 +344,7 @@ class JAISLMHeadModel(nn.Module, SupportsPP):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/jais2.py

@@ -483,7 +483,7 @@ class Jais2ForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/jamba.py

@@ -348,7 +348,7 @@ class JambaModel(nn.Module):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,
@@ -516,7 +516,7 @@ class JambaForCausalLM(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/jina_vl.py

@@ -125,7 +125,7 @@ class JinaVLForSequenceClassification(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/kanana_v.py

@@ -732,7 +732,7 @@ class KananaVForConditionalGeneration(nn.Module, SupportsMultiModal, SupportsPP)
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/keye.py

@@ -1438,7 +1438,7 @@ class BaseKeyeModule(nn.Module, SupportsMultiModal):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/kimi_linear.py

@@ -506,7 +506,7 @@ class KimiLinearForCausalLM(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/kimi_vl.py

@@ -389,7 +389,7 @@ class KimiVLForConditionalGeneration(nn.Module, SupportsMultiModal, SupportsPP):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/lfm2.py

@@ -342,7 +342,7 @@ class Lfm2Model(nn.Module):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,
@@ -503,7 +503,7 @@ class Lfm2ForCausalLM(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/lfm2_moe.py

@@ -457,7 +457,7 @@ class Lfm2MoeModel(nn.Module):
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,
@@ -730,7 +730,7 @@ class Lfm2MoeForCausalLM(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/lfm2_vl.py

@@ -769,7 +769,7 @@ class Lfm2VLForConditionalGeneration(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,

vllm/model_executor/models/llama.py

@@ -651,7 +651,7 @@ class LlamaForCausalLM(
 
     def forward(
         self,
-        input_ids: torch.Tensor | None,
+        input_ids: torch.Tensor,
         positions: torch.Tensor,
         intermediate_tensors: IntermediateTensors | None = None,
         inputs_embeds: torch.Tensor | None = None,