fix vllm qwen2 models (#11879)

python/llm/src/ipex_llm/vllm/xpu/model_convert.py

@@ -30,7 +30,7 @@ from vllm.config import DeviceConfig
 from vllm.model_executor.sampling_metadata import SamplingMetadata
 from vllm.model_executor.parallel_utils.communication_op import tensor_model_parallel_gather
 
-from typing import Tuple, Optional
+from typing import Tuple, Optional, Union
 from ipex_llm.utils.common import invalidInputError
 from vllm.sequence import SamplerOutput
 
@@ -51,8 +51,10 @@ def _Qwen2_sample(
     sampling_metadata: SamplingMetadata,
 ) -> Optional[SamplerOutput]:
     if self.config.tie_word_embeddings:
-        lm_head_weight = self.model.embed_tokens
+        # Embedding layer is not optimized to LowBitLinear
+        lm_head_weight = self.model.embed_tokens.weight
     else:
+        # This layer is optimized to LowBitLinear
         lm_head_weight = self.lm_head
     next_tokens = self.sampler(lm_head_weight, hidden_states,
                                sampling_metadata)
@@ -70,8 +72,14 @@ def _Chatglm_sample(
     return next_tokens
 
 
-def _sample_get_logits(self, hidden_states: torch.Tensor, embedding: torch.nn.Module,
+def _sample_get_logits(self, hidden_states: torch.Tensor,
+                       embedding: Union[torch.nn.Module, torch.Tensor],
                        embedding_bias: Optional[torch.Tensor]) -> torch.Tensor:
+    # For tie_word_embedding models, the embedding is not optimized as
+    # the low_bit_linear layer...
+    if isinstance(embedding, torch.Tensor):
+        logits = torch.matmul(hidden_states, embedding.t())
+    else:
         logits = embedding(hidden_states)
     if embedding_bias is not None:
         logits += embedding_bias