arrow-hf/xvla-robotwin-place-empty-cup-300ep

X-VLA RoboTwin place_empty_cup (300 ep, single instruction)

X-VLA policy fine-tuned on 300 demonstration episodes of the place_empty_cup task from RoboTwin 2.0 (demo_clean config), starting from the lerobot/xvla-base base checkpoint (Florence2 backbone, 879M params).

Training

Config	Value
Base checkpoint	lerobot/xvla-base
Training data	300 RoboTwin demonstrations, single instruction "place the empty cup"
Batch size	32
Steps	20000
dtype	bfloat16
Optimizer	AdamW, lr=1e-4
Chunk size	32

Evaluation

RoboTwin 2.0 sim (demo_clean), 10 episodes, max_steps=400, action_chunk_exec=32.

Success rate: 7/10 (70%)

Usage

from lerobot.policies.xvla import XVLAPolicy
policy = XVLAPolicy.from_pretrained("arrow-hf/xvla-robotwin-place-empty-cup-300ep")

Important: Use action_chunk_exec=32 (full chunk). Default action_chunk_exec=16 causes 0% success on these tasks (TOPP replanning interference).