MiniVLA — MyCobot 280 Pi Task2 (v6, Epoch 2)
Vision-Language-Action model fine-tuned for: "Put the yellow block on the shelf onto the black table on the right".
Performance
| Metric |
Value |
| Overall MAE |
1.31 |
| Overall RMSE |
2.52 |
| Gripper 3-Class Accuracy |
98.03% |
| Dim |
MAE |
| dx |
2.58 mm |
| dy |
1.61 mm |
| dz |
1.58 mm |
| droll |
1.19° |
| dpitch |
0.75° |
| dyaw |
1.43° |
| gripper |
0.018 |
Model
- Base: prism-qwen25-extra-dinosiglip-224px+0_5b (LIBERO-90 pretrained)
- LLM: Qwen2.5 0.5B, fully unfrozen
- Vision: DINOv2 + SigLIP 224px, fully unfrozen
- Action tokenizer: extra_action_tokenizer (vocab=151921)
- Total params: 1.25B
- Training: 90 trajectories, 2 epochs, lr=2e-5 cosine, batch=1
Quick Start
import os
os.environ["PRISMATIC_DATA_ROOT"] = "./"
os.environ["TOKENIZERS_PARALLELISM"] = "false"
from prismatic.models import load_vla
from deploy.mybot_deploy import MyBotDeploy
vla = load_vla("purplehihi/minivla-myCobot280pi-task2-v6", hf_token="dummy")
deploy = MyBotDeploy(stats_path="dataset_statistics.json")
deploy.set_current_encoders(current_encoders)
action = vla.predict_action(image, instruction, unnorm_key="mycobot_task2")
encoders, gripper_pwm, info = deploy.step(action)
robot.send_angles(encoders.tolist(), speed=50)
robot.set_gripper(gripper_pwm)
Deployment Pipeline (mybot_deploy.py)
VLA output [dx,dy,dz,droll,dpitch,dyaw,gripper]
→ unnormalize (BOUNDS: q01/q99)
→ current EEF + delta = target EEF
→ numerical IK (damped least squares)
→ 6 joint angles → motor encoders
→ gripper [0,1] → PWM [0,100]
The IK uses damped least squares (Levenberg-Marquardt) with numerical Jacobian.
Accuracy: ~0.3mm position, ~0.2° orientation (100 random tests).
Normalization Bounds
| Dim |
Min |
Max |
| dx |
-35 |
+35 mm |
| dy |
-30 |
+30 mm |
| dz |
-25 |
+25 mm |
| droll |
-25 |
+25° |
| dpitch |
-15 |
+15° |
| dyaw |
-20 |
+20° |
| gripper |
0 |
1 (absolute, not normalized) |
Files
model.pt — Fine-tuned weights (step 60920, loss=2.7e-5)dataset_statistics.json — Normalization stats with BOUNDS q01/q99config.json — Training config and metadatamybot_deploy.py — Edge deployment: VLA output → encoder commands + IK
