Bug: `orig_height` and `orig_width` variable undeifined in llava processing

[chenweize1998]

System Info

• transformers version: 4.47.0.dev0
• Platform: Linux-5.10.112-005.ali5000.alios7.x86_64-x86_64-with-glibc2.32
• Python version: 3.10.13
• Huggingface_hub version: 0.26.2
• Safetensors version: 0.4.5
• Accelerate version: 1.1.1
• Accelerate config: not found
• PyTorch version (GPU?): 2.4.0 (True)
• Tensorflow version (GPU?): not installed (NA)
• Flax version (CPU?/GPU?/TPU?): not installed (NA)
• Jax version: not installed
• JaxLib version: not installed
• Using distributed or parallel set-up in script?: True
• Using GPU in script?: True
• GPU type: NVIDIA H800

Who can help?

@zucchini-nlp

Information

• The official example scripts
• My own modified scripts

Tasks

• An officially supported task in the examples folder (such as GLUE/SQuAD, ...)
• My own task or dataset (give details below)

Reproduction

Here's a minimum script to reproduce the bug. When return_tensors is not specified, it raises UnboundLocalError: local variable 'orig_height' referenced before assignment.

from transformers import AutoProcessor
from datasets import load_dataset

processor = AutoProcessor.from_pretrained("models/llava-hf/llama3-llava-next-8b-hf/")
dataset = load_dataset("HuggingFaceM4/LLaVA_Wild_Modif")

text = processor.apply_chat_template(
    [
        {
            "role": "user",
            "content": [{"type": "text", "text": dataset["test"][0]["question"]}, {"type": "image", "text": None}],
        }
    ],
    add_generation_prompt=True,
)

# Raise the error "UnboundLocalError: local variable 'orig_height' referenced before assignment"
inputs = processor(images=[dataset["test"][0]["image"]], text=text)

# No error when specifying return_tensors:
# inputs = processor(images=[dataset["test"][0]["image"]], text=text, return_tensors='pt')

Expected behavior

The code should not raise error without specifying return_tensors. The bug is possibly introduced from #34779

[chenweize1998]

Just opened a PR. @zucchini-nlp Could you review the code and confirm whether the changes in code can fix the bug? Thanks