Improve weight loading for encoder models in Transformers backend (vllm-project#25289)

Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Signed-off-by: charlifu <charlifu@amd.com>

Author: hmellor

vllm/model_executor/models/transformers.py

@@ -702,21 +702,45 @@ def load_weights(self, weights: Iterable[tuple[str,
 class TransformersModel(TransformersBase):
     hf_to_vllm_mapper = WeightsMapper(
         orig_to_new_prefix={
+            # Handle BERT-like models
+            "bert": "model",
             # Add `model.` prefix for base model checkpoints
             "": "model.",
-            # Remove `model.` from places it should not be
+            # Remove `model.` prefix if it was already there
             "model.model.": "model.",
+            # Pooling adapters will be adjacent to `model`
+            "model.pooler": "pooler",
             "model.score": "score",
+            # Classifier adapter's classifier layer is renamed to score
+            "model.classifier": "score",
+        },
+        orig_to_new_suffix={
+            # Replace legacy suffixes used for norms
+            ".gamma": ".weight",
+            ".beta": ".bias",
         })
 
     def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""):
         super().__init__(vllm_config=vllm_config, prefix=prefix)
 
-        # Some encoder models have the position_ids buffer in the checkpoint
+        # After creating a pooling model, `pooler` will be duplicated.
+        # The one inside `model` comes from the Transformers modelling code.
+        # The one after `model` is an adapter from vLLM.
+        # We want to use the adapter so we nullify the original pooler.
+        if getattr(self.model, "pooler", None) is not None:
+            self.skip_prefixes.append("pooler.")
+            self.model.pooler = torch.nn.Identity()
+
+        # Some encoder models have the position_ids buffer in the checkpoint.
         # vLLM will always pass position_ids as an argument, so we skip loading
         # the buffer if it exists
         self.skip_substrs.append("position_ids")
 
+        # Some encoder models have the bias of the final classifier layer
+        # in the checkpoint. vLLM does not use this bias, so we skip loading
+        # it if it exists
+        self.skip_substrs.append("score.bias")
+
     def create_attention_instances(
             self, attn_type: AttentionType = AttentionType.DECODER):
         # TODO(hmellor): Better way to detect encoder models