Add LingBot-VA Robotwin integration for V2

docs/source/community/faq.rst

@@ -51,13 +51,16 @@ integrations are:
 - :doc:`/models/causal_forcing` — Streaming Wan 2.1 T2V / I2V (1.3B).
 - :doc:`/models/causal_wan22` — FastVideo Causal Wan 2.2 14B MoE T2V.
 - :doc:`/models/lingbot_world` — Camera-controllable I2V world model.
+- :doc:`/models/lingbot_va` — Offline autoregressive RoboTwin
+  video-and-action rollout.
 - :doc:`/models/flashvsr` — Streaming video super-resolution.
 - :doc:`/models/wan21` — Bidirectional Wan 2.1 T2V / I2V reference.
 - :doc:`/models/cosmos_predict2` — Bidirectional Cosmos-Predict2.5
   T2V / I2V reference.
 
 Each model page has the canonical CLI invocation, checkpoint source,
-multi-GPU command, and per-method knobs.
+requirements, and per-method knobs. Pages include multi-GPU commands where the
+integration supports them.
 
 Installation and packaging
 --------------------------

docs/source/models/index.rst

@@ -28,6 +28,7 @@ Models
    flashvsr
    hy_worldplay
    lingbot_world
+   lingbot_va
    sana_wm_streaming
    sana_wm_bidirectional
    wan21
@@ -41,14 +42,13 @@ method.
 Available models
 ----------------
 
-The models come in three flavors. Streaming and autoregressive generation
-methods build a video step by step and stay fast once warmed up, aiming for
-sub-second latency per step; bidirectional methods produce a clip in a single
-pass and serve as the quality reference for their streaming counterparts; and
-super-resolution methods upscale existing frames in chunks, so their latency
-scales with output resolution rather than step count. Each card links to that
-method's page, where you'll find the exact command to run it, the checkpoint it
-uses, and the settings you can tune.
+The models cover streaming and autoregressive generation, bidirectional
+generation, and super-resolution. Streaming methods can present output
+incrementally; other autoregressive methods may generate step by step but
+return one complete rollout. Bidirectional methods produce a clip in a single
+pass, while super-resolution methods upscale existing frames in chunks. Each
+card links to that method's page, where you'll find the exact command to run,
+the checkpoint it uses, and the settings you can tune.
 
 .. container:: fd-eyebrow
 
@@ -127,6 +127,14 @@ uses, and the settings you can tune.
 
       Camera-controllable image-to-video world model.
 
+   .. grid-item-card:: LingBot-VA
+      :class-card: fd-feature
+      :link: /models/lingbot_va
+      :link-type: doc
+
+      Offline autoregressive RoboTwin image-and-instruction to joint
+      video/action rollout.
+
    .. grid-item-card:: HY-WorldPlay
       :class-card: fd-feature
       :link: /models/hy_worldplay