[Misc] Bump up transformers to v4.39.0 & Remove StarCoder2Config (#3551)

Co-authored-by: Roy <jasonailu87@gmail.com>
Co-authored-by: Roger Meier <r.meier@siemens.com>

vllm/model_executor/models/starcoder2.py

@@ -22,6 +22,7 @@ from typing import List, Optional, Tuple
 
 import torch
 from torch import nn
+from transformers import Starcoder2Config
 
 from vllm.model_executor.input_metadata import InputMetadata
 from vllm.model_executor.sampling_metadata import SamplingMetadata
@@ -42,13 +43,6 @@ from vllm.model_executor.weight_utils import (default_weight_loader,
                                               hf_model_weights_iterator)
 from vllm.sequence import SamplerOutput
 
-try:
-    from transformers import Starcoder2Config
-except ImportError:
-    # fallback to PretrainedConfig
-    # NOTE: Please install transformers from source or use transformers>=4.39.0
-    from transformers import PretrainedConfig as Starcoder2Config
-
 KVCache = Tuple[torch.Tensor, torch.Tensor]