hist3/README.md @ step 6000

hist3/README.md

@@ -0,0 +1,49 @@
+---
+license: cc-by-nc-4.0
+tags: [robotics, vla, pi05, openpi, manipulation]
+---
+
+# hist3
+
+Three frames at 0.5 s spacing plus a relative-motion state (recent pose deltas rather than an absolute pose).
+
+| | |
+|---|---|
+| training config | `pi05_frame_hist3` |
+| checkpoint step | 6,000 of 7,280  **(partial — run did not finish)** |
+| samples seen | 192,000 |
+| batch size | 32 |
+| peak LR | 0.0005, cosine over 7,280 steps |
+| conditioning | 3 frame(s), state `relative_history` |
+| action space | relative EE, `world_rotvec`, horizon 50 |
+| base weights | `pi05_base` (PyTorch), LoRA r16 `all_linear` |
+| frame yaw aug | none |
+| normalizer key | `mix/tuned_h50_relative_history3` |
+
+## Contents
+
+- `model.safetensors` — the weights
+- `norm_stats.json` — the quantile normalizer this checkpoint was trained against. Not
+  interchangeable between variants: the absolute-pose and relative-motion state
+  representations have different scales, and using the wrong one silently rescales output.
+- `train_config.json` — the resolved training config
+- `mixture.json` — dataset sampling weights
+
+## Loading
+
+```python
+from experiments.infer import VariantPolicy
+policy = VariantPolicy.load("hist3")
+actions = policy.infer(images, poses, prompt)   # absolute EE targets, [horizon, 10]
+```
+
+## Provenance
+
+Trained from `pi05_base` (Physical Intelligence openpi) on cong-lab manipulation data plus a
+DROID subset. Inherits the licence terms of both.
+
+## Caveats
+
+Trained on a 10-hour single-GPU budget, which is roughly 0.05 of one pass
+over the 4,245,385-sample mixture. These are ablation checkpoints for a specific
+comparison, not a converged production policy.