[Model] Replace embedding models with pooling adapter (#10769)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

vllm/model_executor/models/qwen2.py

@@ -31,6 +31,7 @@ from vllm.attention import Attention, AttentionMetadata, AttentionType
 from vllm.compilation.decorators import support_torch_compile
 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
 from vllm.model_executor.layers.activation import SiluAndMul
 from vllm.model_executor.layers.layernorm import RMSNorm
 from vllm.model_executor.layers.linear import (MergedColumnParallelLinear,
@@ -55,6 +56,8 @@ from .utils import (AutoWeightsLoader, PPMissingLayer, WeightsMapper,
                     make_empty_intermediate_tensors_factory, make_layers,
                     maybe_prefix)
 
+logger = init_logger(__name__)
+
 
 class Qwen2MLP(nn.Module):
 
@@ -433,7 +436,6 @@ class Qwen2ForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
         config = vllm_config.model_config.hf_config
         quant_config = vllm_config.quant_config
         lora_config = vllm_config.lora_config
-        pooler_config = vllm_config.model_config.pooler_config
 
         self.config = config
         self.lora_config = lora_config
@@ -454,14 +456,6 @@ class Qwen2ForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
         self.logits_processor = LogitsProcessor(config.vocab_size)
         self.sampler = get_sampler()
 
-        # The same model class supports both language generation and embedding
-        # because the architecture name is the same
-        self._pooler = Pooler.from_config_with_defaults(
-            pooler_config,
-            pooling_type=PoolingType.LAST,
-            normalize=True,
-            softmax=False)
-
         self.make_empty_intermediate_tensors = (
             self.model.make_empty_intermediate_tensors)
 
@@ -499,13 +493,6 @@ class Qwen2ForCausalLM(nn.Module, SupportsLoRA, SupportsPP):
         next_tokens = self.sampler(logits, sampling_metadata)
         return next_tokens
 
-    def pooler(
-        self,
-        hidden_states: torch.Tensor,
-        pooling_metadata: PoolingMetadata,
-    ) -> Optional[PoolerOutput]:
-        return self._pooler(hidden_states, pooling_metadata)
-
     def load_weights(self, weights: Iterable[Tuple[str,
                                                    torch.Tensor]]) -> Set[str]:
         loader = AutoWeightsLoader(
@@ -553,6 +540,15 @@ class Qwen2EmbeddingModel(nn.Module, SupportsLoRA, SupportsPP):
         self.model = Qwen2Model(vllm_config=vllm_config,
                                 prefix=maybe_prefix(prefix, "model"))
 
+        # TODO: Replace this model class with for_embedding(Qwen2ForCausalLM),
+        # after changing the default pooling method
+        if pooler_config.pooling_type is None:
+            logger.warning(
+                "This embedding model will default to last-token pooling in "
+                "an upcoming version. To avoid breaking changes, you should "
+                "pass `--override-pooler-config '{\"pooling_type\": \"MEAN\"}'`"
+                " explicitly.")
+
         self._pooler = Pooler.from_config_with_defaults(
             pooler_config,
             pooling_type=PoolingType.MEAN,