[Feature][LingBot World]stepwise execution (#6844)

Signed-off-by: BruceLoveDecimal <liuqihao970610@gmail.com>
Signed-off-by: liuqihao <liuqihao970610@gmail.com>

examples/offline_inference/diffusion/lingbot_world_v2.py

@@ -30,7 +30,6 @@ from typing import Any
 import numpy as np
 
 _MODEL = "robbyant/lingbot-world-v2-14b-causal-fast-diffusers"
-_NUM_INFERENCE_STEPS = 4
 _MAX_SEQUENCE_LENGTH = 512
 _MAX_PIXEL_AREA = 480 * 832
 _MAX_RAW_FRAMES = 117
@@ -209,7 +208,6 @@ def build_request(
         "height": args.height,
         "width": args.width,
         "num_frames": args.num_frames,
-        "num_inference_steps": _NUM_INFERENCE_STEPS,
         "max_sequence_length": _MAX_SEQUENCE_LENGTH,
         "seed": args.seed,
         "fps": args.fps,

examples/offline_inference/diffusion/lingbot_world_v2_realtime.py

@@ -155,7 +155,6 @@ async def run(argv: Sequence[str] | None = None) -> Path:
         height=args.height,
         width=args.width,
         num_frames=9,
-        num_inference_steps=4,
         max_sequence_length=512,
         seed=args.seed,
         output_type="latent",