[hmellor]

For some models, SomeVLTextConfig may also have a tie_word_embeddings field. This is only the case if SomeVLTextConfig is also used for a text only version of the same model. For example:

SomeVLModelForCausalLM:
    self.model = SomeLanguageModel(SomeVLTextConfig)
    self.lm_head = nn.Linear()

Therefore, the presence of tie_word_embeddings in SomeVLTextConfig cannot be used as a signal for whether tie_word_embeddings should be copied from hf_config to the language_model config.

[gemini-code-assist]

Code Review

This pull request refines the logic within the with_hf_config method in vllm/config/vllm.py by simplifying the conditional check for tie_word_embeddings in multimodal models. The change removes a redundant check for the absence of tie_word_embeddings in the text configuration. Additionally, the accompanying comments have been significantly expanded to provide a more detailed and accurate explanation of how tie_word_embeddings should be handled across different multimodal model architectures, clarifying its relationship between the main model configuration and the language model's specific configuration. There is no feedback to provide from the review comments.