Video to Video — FLUX 3 Action turns what the robot sees into what it does next. Give it the scene camera image, the wrist camera image, the current SO-101 joint state and a plain-language instruction.

≈ 26.887 kredi/üretim

Ayarlar

Scene Image Urlzorunlu

Current frame from the fixed scene camera (tiled left). Resized to 256x256.

Wrist Image Urlzorunlu

Current frame from the wrist camera (tiled right). Resized to 256x256.

Statezorunlu

Measured state: shoulder pan, shoulder lift, elbow flex, wrist flex, wrist roll (degrees), gripper (percent).

State Historyisteğe bağlı

Measured states at the 7 control ticks (30 Hz) before `state`, oldest first. Requires `command_history`. Omit for a fresh episode.

Command Historyisteğe bağlı

Absolute commands in effect at each of the 8 observation ticks, oldest first; the last row is the most recent command sent. Requires `state_history`.

History Scene Image Urlisteğe bağlı

Scene frame from the oldest history tick (7 ticks before now). Requires `state_history`. Defaults to the current frame.

History Wrist Image Urlisteğe bağlı

Wrist frame from the oldest history tick. Requires `state_history`.

Num Inference Stepsisteğe bağlı

Euler denoising steps. The checkpoint was validated at 4.

Guidance Scaleisteğe bağlı

Classifier-free guidance on video and action; 1.0 disables CFG.

Seedisteğe bağlı

Sampling seed. Omit for a random seed; the reference evaluation seed is 0 for DROID and 42 for SO-101.

Üretmek için giriş yap

Modeli inceleyebilirsin; üretim için hesabınla giriş yapman gerekiyor.

Giriş yap