Downloads · 30 days
0
0% of all-time downloads
learner1119/ffw_sh5_n17_260820_left_h50_rel_vis
ffw_sh5_n17_260820_left_h50_rel_vis is a robotics model from learner1119. 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 transformers. The card lists the license as other.
Fine-tune of nvidia/GR00T-N1.7-3B on learner1119/260820 (100 episodes / 44,136 frames, FFW-SH5, left arm only). One of four runs differing only in action representation and whether the vision tower is trained; this re…
Downloads · 30 days
0
0% of all-time downloads
All-time downloads
11
Public
Repo size
34.6 GB
Likes
0
Public
Click a slice to open those files.
.safetensors34.6 GB · 100%
From the Hugging Face model README
Fine-tune of nvidia/GR00T-N1.7-3B on
learner1119/260820 (100 episodes / 44,136 frames,
FFW-SH5, left arm only). One of four runs differing only in action representation and whether the
vision tower is trained; this repo is RELATIVE · tune_visual=True.
Checkpoints are subfolders, each a complete weights-only checkpoint as the trainer wrote it
(BF16, 2 shards, no optimizer state). Load one with model_path="learner1119/ffw_sh5_n17_260820_left_h50_rel_vis/checkpoint-50000"
or download the subfolder.
| folder | step | train loss (25-pt moving avg) |
|---|---|---|
checkpoint-10000 | 10,000 | 0.0358 |
checkpoint-20000 | 20,000 | 0.0202 |
checkpoint-30000 | 30,000 | 0.0120 |
checkpoint-40000 | 40,000 | 0.0072 |
checkpoint-50000 | 50,000 | 0.0059 |
Final run: 30h 59m on 4× A100 80GB at 2.23 s/step. Train loss only — no validation split
(eval_strategy='no'); tune_visual=True reaching ~half the loss on 100 episodes may be overfitting.
Compare on held-out rollouts before choosing.
The parquet stores 16-dim state/action (both arms) but only the left 8 move — over all 44,136 frames the
right-arm dims have std ≤ 0.0022 rad and gripper_r is constant — so the modality config declares only
left_arm (0:7) and left_gripper (7:8). The SH5 hand's open/close is interpolated by the teleop stack to a
single 0–1 scalar, so it is treated as 1-dof. Camera: observation.images.agentview 480×640 as cam_head.
Chunk: 50 steps (2.5 s at 20 fps).
left_arm targets are relative — deltas from the joint angles at the observation timestep, one reference for
all 50 steps; the gripper stays absolute. Normalisation statistics are per-horizon-step (relative_stats.json).
global batch 64 (16/GPU × 4, DeepSpeed ZeRO-2, bf16) · 50k steps · lr 1e-4 cosine, warmup 0.05, wd 1e-5 ·
state_dropout 0.2 · tune_llm False / tune_visual True / projector True / diffusion True · seed 42 ·
backbone nvidia/Cosmos-Reason2-2B (LLM layers ≤ 12). The modality config used is ffw_sh5_left8_h50_rel_config.py at the repo root —
import it to register NEW_EMBODIMENT before loading. experiment_cfg/ inside each checkpoint is verbatim.
Sibling runs: ffw_sh5_n17_260820_left_h50_abs, ffw_sh5_n17_260820_left_h50_abs_vis, ffw_sh5_n17_260820_left_h50_rel.
Code: KimDoYoung1997/Isaac-GR00T @ n1.7-doyoung-a100.