image_resolution 参数文档是不是写错了？

[rover5056]

Reminder

• I have read the above rules and searched the existing issues.

System Info

• llamafactory version: 0.9.1.dev0
• Platform: Linux-5.4.119-1-tlinux4-0010.3-x86_64-with-glibc2.35
• Python version: 3.10.12
• PyTorch version: 2.4.1+cu124 (GPU)
• Transformers version: 4.46.1
• Datasets version: 3.0.1
• Accelerate version: 1.0.1
• PEFT version: 0.11.1
• TRL version: 0.9.6
• GPU type: NVIDIA A800-SXM4-80GB
• DeepSpeed version: 0.15.4
• vLLM version: 0.6.3.post1

Reproduction

文档里关于图片分辨率的参数 是 ： image_resolution ，默认值是 512*512 。看 代码也是

LLaMA-Factory/src/llamafactory/hparams/model_args.py

Line 61 in 0279427

image_resolution: int = field(

LLaMA-Factory/src/llamafactory/hparams/model_args.py

Line 62 in 0279427

default=512 * 512,

所以这里应该是个 int， 512*512 = 262144

但是实际处理的时候，又是

LLaMA-Factory/src/llamafactory/data/mm_plugin.py

Line 73 in 74a79cc

def _preprocess_image(self, image: "ImageObject", **kwargs) -> "ImageObject":

LLaMA-Factory/src/llamafactory/data/mm_plugin.py

Line 78 in 74a79cc

if max(image.width, image.height) > image_resolution:

直接取得值和长短边比较了？ 应该是直接和 512 比？
容易造成 confuse

Others

No response

[rover5056]

看错分支了，最新的代码已经正常。。