Downloads · 30 days
10
50% of all-time downloads
bklassen3434/smolvla_pick_pen_v2_lr1e4
smolvla_pick_pen_v2_lr1e4 is a robotics model from bklassen3434. Use it for the robotics task on the model card, and read the license before you ship it in a product. It is set up for lerobot. The card lists the license as apache-2.0.
SmolVLA (450M) fine-tuned from lerobot/smolvlabase on an SO-101 teleop dataset for language-conditioned pen selection: "Pick up the blue pen" / "Pick up the pink pen" / "Pick up the grey pen" (all three pens present i…
Downloads · 30 days
10
50% of all-time downloads
All-time downloads
20
Public
Parameters
450M
907 MB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors907 MB · 100%
How the weights are stored.
BF16447M · 99%
From the Hugging Face model README
SmolVLA (450M) fine-tuned from lerobot/smolvla_base on an SO-101 teleop dataset for
language-conditioned pen selection: "Pick up the blue pen" / "Pick up the pink pen" /
"Pick up the grey pen" (all three pens present in every episode).
Part of a 3-way sweep — see also smolvla_pick_pen_v2_lr5e5 (lower LR) and
smolvla_pick_pen_v2_frozen (VLM frozen).
| Base model | lerobot/smolvla_base |
| Dataset | bklassen3434/pick_pen_v2_20260920_124400 |
| Steps | 10,000 (batch size 64, ~24 epochs) |
| Learning rate | 1e-4 (default), cosine decay over 10,000 steps |
| Vision encoder | unfrozen (freeze_vision_encoder=false, train_expert_only=false) |
| Hardware | 1x A100-80GB (Modal), 2h38m, 58.7 GB peak |
| Final loss | 0.029 |
Cameras are recorded as observation.images.top / observation.images.wrist and remapped
onto SmolVLA's baked-in camera keys at train time:
--rename_map='{"observation.images.top": "observation.images.camera1",
"observation.images.wrist": "observation.images.camera2"}'
The same rename map must be passed at eval time (lerobot-rollout / lerobot-record).