VQModel issue in training VAR

[YannX1e]

Thanks for your great work!
I just followed the training code for XQGAN using "configs/MSVR10P2-4096.yaml". After training VQModel, I try to training VAR following given command in README.md.
But I got the Error as:

[rank2]: Traceback (most recent call last):
[rank2]:   File "***/ImageFolder/train.py", line 362, in <module>
[rank2]:     try: main_training()
[rank2]:   File "***/ImageFolder/train.py", line 209, in main_training
[rank2]:     stats, (sec, remain_time, finish_time) = train_one_ep(
[rank2]:   File "***/ImageFolder/train.py", line 331, in train_one_ep
[rank2]:     grad_norm, scale_log2 = trainer.train_step(
[rank2]:   File "***/ImageFolder/trainer.py", line 122, in train_step
[rank2]:     gt_idx_Bl_list = self.vae_local.img_to_idxBl(inp_B3HW)
[rank2]:   File "/home/ubuntu/anaconda3/envs/var/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1729, in __getattr__
[rank2]:     raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'")
[rank2]: AttributeError: 'VQModel' object has no attribute 'img_to_idxBl'

I find self.vae_local is built from function build_everything, which return a class VQModel as vae_local, code in "models/__init__.py" as bellow:

# build models
# vae_local = VQVAE(vocab_size=V, z_channels=Cvae, ch=ch, test_mode=True, share_quant_resi=share_quant_resi, v_patch_nums=patch_nums).to(device)
vae_local = VQ_models['VQ-16'](...)

Attribute img_to_idxBl can be found in class VQVAE, but xqtrain.py is the training script for the class VQModel, why?

Environment:
Ubuntu 22.04 LTS
conda environment as described in “environment.yml”