SO-101 VLA-JEPA Stack

so101_vla_jepa_stack is a PEFT/LoRA fine-tune of lerobot/VLA-JEPA-Pretrain for an SO-101 robot arm. It was trained on Samisaliveagain/so101_wm, which contains teleoperated stacking and unstacking demonstrations with two synchronized RGB camera views.

The dataset was recorded and the model was trained jointly by Samyak Jain (Samisaliveagain) and Shubham (shubham4413).

VLA-JEPA combines a Qwen3-VL vision-language backbone, a frozen V-JEPA2 encoder, an action-conditioned JEPA video predictor, and a flow-matching DiT action head.

Status

Training completed successfully: 30,000 / 30,000 steps
Final logged training loss: 0.134
Held-out offline evaluation: not yet reported
Real-robot success rate: not yet reported

The loss result demonstrates stable optimization of the training objective. It must not be interpreted as a physical task-success percentage.

Intended use

This checkpoint is intended for research and controlled evaluation of stacking/unstacking policies on the same 6-DoF SO-101 setup, camera arrangement, objects, and workspace represented in the training dataset.

The model is not intended for unsupervised operation around people, fragile objects, or safety-critical equipment. A human operator should remain at the emergency stop during every initial rollout.

Model inputs and outputs

Inputs

Feature	Type	Shape	Deployment source
`observation.images.exterior_1_left`	RGB image	`(3, 224, 224)`	Dataset/robot camera `left`
`observation.images.exterior_2_left`	RGB image	`(3, 224, 224)`	Dataset/robot camera `fpv`
Task instruction	Text	—	Stack or unstack instruction

The deployment pipeline must apply this exact mapping:

{
  "observation.images.left": "observation.images.exterior_1_left",
  "observation.images.fpv": "observation.images.exterior_2_left"
}

Output

Feature	Type	Shape
`action`	SO-101 joint-position action	`(6,)`

Joint order:

shoulder_pan.pos, shoulder_lift.pos, elbow_flex.pos,
wrist_flex.pos, wrist_roll.pos, gripper.pos

The action chunk size is 7 and the gripper index is 5.

What this repository contains

This is a PEFT checkpoint, not a standalone copy of all 3.1B base-model parameters.

adapter_model.safetensors contains:

rank-16 LoRA adapters for Qwen attention projections: q_proj, k_proj, v_proj, and o_proj;
rank-16 LoRA adapters for Qwen MLP projections: gate_proj, up_proj, and down_proj;
the complete fine-tuned model.action_model;
the complete fine-tuned model.video_predictor.

The original Qwen weights and V-JEPA2 encoder remain in the base model and are not duplicated here. LeRobot/PEFT loads lerobot/VLA-JEPA-Pretrain and applies the contents of this repository.

The small preprocessor and postprocessor safetensor files contain the normalization and unnormalization statistics required for correct robot actions.

Fine-tuning details

Training data

Property	Value
Dataset	`Samisaliveagain/so101_wm`
Total episodes	177
Training episodes	151
Held-out episodes	26
Training frames	302,957
Training samples consumed	240,000
Approximate passes over training frames	0.79
Cameras	`left`, `fpv`
Source resolution/rate	640×480 at 30 FPS
Tasks	Stack and unstack large 3D-printed nuts

The episode split used seed 1000. Episodes were split at episode level, not frame level.

Trainable components

Component	Training mode
Qwen3-VL backbone	Frozen base weights with LoRA adapters
DiT action model	Fully trained
Action/state projections	Reinitialized for 6-DoF and fully trained
JEPA video predictor	Fully trained
V-JEPA2 encoder	Frozen

Four tensors from the 7-DoF pretrained action/state interface were intentionally reinitialized for the 6-DoF SO-101:

model.action_model.action_encoder.layer1.weight
model.action_model.action_decoder.layer2.weight
model.action_model.action_decoder.layer2.bias
model.action_model.state_encoder.layer1.weight

Parameter counts

Parameters	Count
Total	3,104,588,172
Learnable	334,258,694
Learnable fraction	10.77%

Hyperparameters

Hyperparameter	Value
Steps	30,000
Batch size	8
Optimizer	AdamW
Peak learning rate	`1e-4`
Warm-up	5,000 steps
Schedule	Cosine decay
Final learning rate	`1e-6`
LoRA rank	16
LoRA alpha	32
LoRA dropout	0.0
World-model loss weight	0.1
Gradient clipping	1.0
Training dtype	bfloat16

Compute

One NVIDIA H100-class Hopper GPU on RWTH HPC
Slurm job 2274795
2026-07-25 10:51:08 to 2026-07-26 03:24:21
Runtime: 16 h 33 min 13 s
Typical GPU memory usage: approximately 35.6 GB
Typical throughput: approximately 4 samples/s

Training results

Step	Samples	Training loss	Gradient norm
100	800	1.410	0.894
1,000	8,000	0.195	2.303
5,000	40,000	0.158	0.519
10,000	80,000	0.146	0.316
15,000	120,000	0.142	0.226
20,000	160,000	0.139	0.204
25,000	200,000	0.135	0.149
30,000	240,000	0.134	0.139

The logged loss decreased by approximately 90.5% from step 100 to step 30,000 and plateaued around 0.133–0.135. The successful run contained no NaNs, CUDA out-of-memory events, or fatal CUDA errors.

No validation loss, held-out action-error metric, or physical success rate is claimed because those measurements have not yet been completed.

Installation

Use the LeRobot revision that produced this checkpoint:

git clone https://github.com/huggingface/lerobot.git
cd lerobot
git checkout 3dd19d04

python -m venv .venv
source .venv/bin/activate
pip install --upgrade pip
pip install -e ".[vla_jepa,core_scripts,feetech,peft]"

Download the checkpoint before powering the robot:

hf download Samisaliveagain/so101_vla_jepa_stack \
  --local-dir "$HOME/models/so101_vla_jepa_stack"

Real-robot rollout

Replace every <...> placeholder with values verified on the deployment computer. Do not guess the arm port, calibration ID, camera identities, or safe positional-change limit.

lerobot-rollout \
  --strategy.type=base \
  --policy.path="$HOME/models/so101_vla_jepa_stack" \
  --device=cuda \
  --robot.type=so101_follower \
  --robot.port=<FOLLOWER_PORT> \
  --robot.id=<FOLLOWER_CALIBRATION_ID> \
  --robot.max_relative_target=<VALIDATED_CONSERVATIVE_POSITION_LIMIT> \
  --robot.cameras='{left: {type: opencv, index_or_path: <LEFT_CAMERA>, width: 640, height: 480, fps: 30}, fpv: {type: opencv, index_or_path: <FPV_CAMERA>, width: 640, height: 480, fps: 30}}' \
  --rename_map='{"observation.images.left":"observation.images.exterior_1_left","observation.images.fpv":"observation.images.exterior_2_left"}' \
  --task="<use the exact stack or unstack instruction represented in the dataset>" \
  --fps=30 \
  --duration=10 \
  --return_to_initial_position=true \
  --display_data=true

For initial deployment:

Run held-out offline action evaluation first.
Confirm that left and fpv are correctly assigned and reproduce the training views.
Start with an empty, padded workspace and a central arm pose.
Keep a human operator at the emergency stop.
Abort on jerky motion, joint-limit seeking, incorrect gripper direction, or increasing latency.
Measure success over 20–30 controlled trials before routine use.

Limitations

No held-out action MSE/MAE has been reported yet.
No real-robot success rate has been reported yet.
Data comes from one robot, workspace, lighting setup, and operator.
The dataset contains successful demonstrations but no recovery/failure episodes.
There are no force, torque, or depth observations.
Camera mounting or key mismatches can cause immediate distribution shift.
A 2B VLM may not sustain the desired control rate on an 8 GB mobile GPU.
The checkpoint requires its base model; the 1.3 GB adapter is not standalone.