Mismatch of default VAE tiling config in training / inferencing

[cauphenuny]

The default tiling config in training (train_recammaster.py) is tiled=False, tile_size=(34,34), tile_stride=(18, 16)，while the default config in inferencing (inference_recammaster.py, diffsynth/pipelines/wan_video.py) is tiled=True, tile_size=(30, 52), tile_stride=(15, 26).

This mismatch can cause about 0.2 absolute error in encoded VAE features. Is this a significant issue?

# train_recammaster.py
    parser.add_argument(
        "--tiled",
        default=False,
        action="store_true",
        help="Whether enable tile encode in VAE. This option can reduce VRAM required.",
    )
    parser.add_argument(
        "--tile_size_height",
        type=int,
        default=34,
        help="Tile size (height) in VAE.",
    )
    parser.add_argument(
        "--tile_size_width",
        type=int,
        default=34,
        help="Tile size (width) in VAE.",
    )
    parser.add_argument(
        "--tile_stride_height",
        type=int,
        default=18,
        help="Tile stride (height) in VAE.",
    )
    parser.add_argument(
        "--tile_stride_width",
        type=int,
        default=16,
        help="Tile stride (width) in VAE.",
    )

# inference_recammaster.py
        video = pipe(
            prompt=target_text,
            negative_prompt="色调艳丽，过曝，静态，细节模糊不清，字幕，风格，作品，画作，画面，静止，整体发灰，最差质量，低质量，JPEG压缩残留，丑陋的，残缺的，多余的手指，画得不好的手部，画得不好的脸部，畸形的，毁容的，形态畸形的肢体，手指融合，静止不动的画面，杂乱的背景，三条腿，背景人很多，倒着走",
            source_video=source_video,
            target_camera=target_camera,
            cfg_scale=args.cfg_scale,
            num_inference_steps=50,
            seed=0, tiled=True
        )

# diffsynth/pipelines/wan_video.py
    @torch.no_grad()
    def __call__(
        self,
        prompt,
        negative_prompt="",
        source_video=None,
        target_camera=None,
        input_image=None,
        input_video=None,
        denoising_strength=1.0,
        seed=None,
        rand_device="cpu",
        height=480,
        width=832,
        num_frames=81,
        cfg_scale=5.0,
        num_inference_steps=50,
        sigma_shift=5.0,
        tiled=True,
        tile_size=(30, 52),
        tile_stride=(15, 26),
        tea_cache_l1_thresh=None,
        tea_cache_model_id="",
        progress_bar_cmd=tqdm,
        progress_bar_st=None,
    ):