Model: HuggingFaceVLA/smolvla_libero (revision 6721902bc4d61e50a3bfdb11dfb4cb626f05d102)
Route: manipulation-vla (SmolVLA) — 2x RGB (256×256) + state(8) + language(48 tokens) → continuous action chunk [50,7]
VLM: HuggingFaceTB/SmolVLM2-500M-Instruct (SigLIP + SmolLM2 500M) + small expert (cross-attn, 10 flow-matching steps)
Task: LIBERO spatial (pick-up black bowl between plate and ramekin and place on plate)
Hardware: Ascend 910 (Ascend910_9362) — single npu:0 — CPU fallback is failure
observation.images.image [3,256,256] and observation.images.image2 [3,256,256] (RGB 0-1, resized with padding to 512×512, normalized to [-1,1] for SigLIP), observation.state [8] padded to 32, observation.language.tokens [48] + mask (SmolVLM2 tokenizer, task = natural language instruction, max_length 48).action_chunk [50,7] (7-DoF: XYZ + quat/gripper, normalized MEAN_STD, executed horizon n_action_steps=1). Raw chunk kept as raw_output; normalized values compared for consistency.[-0.156, 0.984, 0.176, -0.172, -0.579, 0.172, 1.039][-0.156, 0.997, 0.197, -0.172, -0.573, 0.169, 1.038] — shape [1,50,7], dtype float32, device npu:0validate_policy_outputs.pynpu-smi info shows Ascend910_9362, torch.npu.is_available()==True, model params/image/state/noise/output all on npu:0, timing wrapped with torch.npu.synchronize().pip install -r requirements.txt # torch 2.9.0 + torch_npu 2.9.0.post1 + lerobot 0.4.4 + transformers 4.57.1
export HF_HUB_OFFLINE=1 # if local /tmp/smolvla_libero exists, otherwise downloads from HuggingFaceVLA/smolvla_libero
python inference.pyinference.py loads from /tmp/smolvla_libero if present else from HuggingFaceVLA/smolvla_libero (fixed revision), moves to npu:0, runs identical batch on CPU/NPU with fixed noise seed, prints model/revision/route/backend/device/dtype, input summary, action summaries, synchronized latency and PASS/FAIL.
transformers SigLIP bucketize cast to float32 (BF16 unsupported on NPU searchsorted), lerobot eager attention where condition cast to bool (Long mask not accepted on strict torch 2.9). No model weights changed, no dummy policy.HuggingFaceTB/SmolVLM2-500M-Instruct shares blob b9bfd... with SmolVLM2-500M-Video-Instruct (same commit 7b375e), local cache symlinked — no remote code added.trust_remote_code=False.assets/agent_workflow.png — full workflow, consistency and performanceassets/npu_device_call.png — npu-smi, torch.npu availability, device name and tensor devicesassets/model_result.png — default python inference.py result
