Update card: trained 2026-07-08 (40k, results)

README.md

@@ -35,7 +35,7 @@ Single-handed pick task ("grab the bottle", RIGHT hand). Trained at the **CloudW
 
 **Predecessors:** [v2 (`…-210ep-v2-finetune`)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-210ep-v2-finetune) — the current **validated production champion** (`checkpoint-20000`, 11/12) — [v4 (radius-5)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-417ep-v4-finetune), [v5 (radius-20)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-314ep-v5-finetune), [v6 (merged radius-20)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-502ep-v6-finetune), [v7 (speedup-3mm raw)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-220ep-v7-finetune), [v8 (speedup-3mm + cycle-removed raw)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-405ep-v8-finetune), [v9 (speedup-2mm on curated)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-193ep-v9-finetune), [v10 (Break-Down merge)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-371ep-v10-finetune) — all eval pending — and [v1 (105ep)](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-105ep-v1-finetune).
 
-This is a behavior-cloning fine-tune of the full 3B model. **Status: training pending** (recipe: **40,000 steps**, save every 10,000 → `{checkpoint-10000, checkpoint-20000, checkpoint-30000, checkpoint-40000}`, all four to be published here). Closed-loop comparison against v10 (speed-up isolation) and the v2 champion is pending (see [Evaluation](#evaluation)).
+This is a behavior-cloning fine-tune of the full 3B model. **Status: trained 2026-07-08** (**40,000 steps**, save every 10,000 → `{checkpoint-10000, checkpoint-20000, checkpoint-30000, checkpoint-40000}`, all four published here; wall-clock 2 h 12 min, W&B offline `redacted`). Closed-loop comparison against v10 (speed-up isolation) and the v2 champion is pending (see [Evaluation](#evaluation)).
 
 ## Quick facts
 
@@ -54,9 +54,9 @@ This is a behavior-cloning fine-tune of the full 3B model. **Status: training pe
 | Steps / batch | 40,000 / global batch 48 (8 per GPU × 6 GPUs); checkpoints saved every 10,000 → `{checkpoint-10000, checkpoint-20000, checkpoint-30000, checkpoint-40000}` |
 | Epochs | 5.51 @ 10k / 11.02 @ 20k / 16.52 @ 30k / 22.03 @ 40k (1.92M frame-views ÷ 87,148 frames) |
 | Augmentation | color jitter (brightness 0.3, contrast 0.4, saturation 0.5, hue 0.08) |
-| Wall-clock | ~2 h 05 min for 40k steps on 6× B200 (est.); steady-state ~5.3 it/s |
-| Final train loss | training pending |
-| W&B run | training pending (project `g1_grab_bottle`; offline → synced) |
+| Wall-clock | **2 h 12 min 21 s** for 40k steps on 6× B200; ~5.3 it/s |
+| Final train loss | per rung: **10k = 0.0698 · 20k = 0.0412 · 30k = 0.0240 · 40k = 0.0302** (per-step log; min 0.0144, mean 0.0649 over 4000 logged steps). No held-out split — a fit probe, not generalization. |
+| W&B run | offline `redacted` (project `g1_grab_bottle`) |
 
 ## Repository contents
 
@@ -138,7 +138,7 @@ Closed-loop control of the G1 (sim or real) is driven by the **SONIC whole-body
 | [v8](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-405ep-v8-finetune) | [speedup-3mm-cycle-removed-v1](https://huggingface.co/datasets/cloudwalk-research/gr00t-g1-grab-bottle-right-hand-speedup-3mm-cycle-removed-v1) | 405 | 47,944 | 20.0 @20k | ⏳ TBD | **DP speedup + segment removal** on the **raw merged** set (the v7 follow-up). |
 | [v9](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-193ep-v9-finetune) | [speedup-2mm-v3](https://huggingface.co/datasets/cloudwalk-research/gr00t-g1-grab-bottle-right-hand-speedup-2mm-v3) | 193 | 32,786 | 29.3 @20k | ⏳ TBD | **Gentle 2 mm DP speedup on the *curated* (v2-lineage) data**. Smallest set. Trained (loss 0.0308 @20k). |
 | [v10](https://huggingface.co/cloudwalk-research/gr00t-n17-g1-grab-bottle-rh-371ep-v10-finetune) | [right-hand-v10](https://huggingface.co/datasets/cloudwalk-research/gr00t-g1-grab-bottle-right-hand-v10) | 371 | 138,546 | 10.4 @30k | ⏳ TBD | **Largest set** — merged DS1+DS2, **Break-Down v1-wandering only** (no speedup). Trained (loss 0.0280 @30k). **Speed-up-isolation sibling of v11.** |
-| **v11 (this)** | [right-hand-v11](https://huggingface.co/datasets/cloudwalk-research/gr00t-g1-grab-bottle-right-hand-v11) | 355 | 87,148 | **22.0 @40k** | ⏳ TBD | **Break-Down + Speed-Up** (v10 merge, then DP 2 mm/frame dynamic resampling). 40k steps → `{20k,30k}` bracket v2's ~15.3 sweet spot, 40k probes the overfit knee. Compare vs **v10** (speedup isolation) and the **v2 champion**. |
+| **v11 (this)** | [right-hand-v11](https://huggingface.co/datasets/cloudwalk-research/gr00t-g1-grab-bottle-right-hand-v11) | 355 | 87,148 | **22.0 @40k** | ⏳ TBD | **Break-Down + Speed-Up** (v10 merge, then DP 2 mm/frame dynamic resampling). Trained 2026-07-08 (30k=0.0240, 40k=0.0302, min 0.0144). 40k steps → `{20k,30k}` bracket v2's ~15.3 sweet spot, 40k probes the overfit knee. Compare vs **v10** (speedup isolation) and the **v2 champion**. |
 
 ## References