[Bugfix][Model] Skip loading lm_head weights if using tie_word_embeddings (#6758)

Signed-off-by: Travis Johnson <tsjohnso@us.ibm.com>

vllm/model_executor/models/olmo.py

@@ -343,6 +343,11 @@ class OlmoForCausalLM(nn.Module):
                 # Models trained using ColossalAI may include these tensors in
                 # the checkpoint. Skip them.
                 continue
+            # With tie_word_embeddings, we can skip lm_head.weight
+            # The weight might appear unnecessarily in the files if the model is
+            # processed with quantization, LoRA, fine-tuning, etc.
+            if self.config.tie_word_embeddings and "lm_head.weight" in name:
+                continue
             for (param_name, weight_name, shard_id) in stacked_params_mapping:
                 if weight_name not in name:
                     continue