README: document 4 checkpoint subfolders (30k/40k/50k/60k)

README.md

@@ -13,6 +13,17 @@ base_model: 2toINF/X-VLA-Pt
 
 Fine-tune of [`2toINF/X-VLA-Pt`](https://huggingface.co/2toINF/X-VLA-Pt) on a single BEHAVIOR-1K task (turning on the radio receiver), using the **v20 architecture** from [markli1hoshipu/behavior1k-xvla @ v20](https://github.com/markli1hoshipu/behavior1k-xvla/tree/v20).
 
+## Available checkpoints
+
+| Subfolder | Steps | total loss | joints | skill_cls | progress |
+|---|---|---|---|---|---|
+| [`ckpt-30000/`](./ckpt-30000) | 30,000 | 0.0232 | 0.0204 | 0.0000 | 0.0027 |
+| [`ckpt-40000/`](./ckpt-40000) | 40,000 | 0.0165 | 0.0158 | 0.0000 | 0.0007 |
+| [`ckpt-50000/`](./ckpt-50000) | 50,000 | 0.0131 | 0.0117 | 0.0000 | 0.0014 |
+| [`ckpt-60000/`](./ckpt-60000) | 60,000 (final) | 0.0135 | 0.0107 | 0.0000 | 0.0028 |
+
+Each subfolder is fully self-contained — load any of them with `subfolder="ckpt-XXXXX"` (see Usage).
+
 ## Architecture (v20)
 
 - **Additive per-task + per-skill soft prompts** (`task_prompt_hub[task_id] + skill_prompt_hub[skill_id]`), 32 tokens × 1024 dim each, zero-initialized
@@ -60,8 +71,11 @@ Fine-tune of [`2toINF/X-VLA-Pt`](https://huggingface.co/2toINF/X-VLA-Pt) on a si
 ```python
 from transformers import AutoModel, AutoConfig
 
-config = AutoConfig.from_pretrained("Hoshipu/xvla-v20-task0-mp-radio", trust_remote_code=True)
-model  = AutoModel.from_pretrained("Hoshipu/xvla-v20-task0-mp-radio", trust_remote_code=True)
+REPO = "Hoshipu/xvla-v20-task0-mp-radio"
+CKPT = "ckpt-60000"   # or ckpt-30000 / ckpt-40000 / ckpt-50000
+
+config = AutoConfig.from_pretrained(REPO, subfolder=CKPT, trust_remote_code=True)
+model  = AutoModel.from_pretrained(REPO, subfolder=CKPT, trust_remote_code=True)
 ```
 
 Or deploy as an inference WebSocket server (handles all pre/post-processing for OmniGibson observations):
@@ -70,13 +84,19 @@ Or deploy as an inference WebSocket server (handles all pre/post-processing for
 git clone -b v20 https://github.com/markli1hoshipu/behavior1k-xvla.git
 cd behavior1k-xvla
 bash setup.sh
+
+# Download a single checkpoint to a local dir
+huggingface-cli download Hoshipu/xvla-v20-task0-mp-radio \
+    --include "ckpt-60000/*" \
+    --local-dir ./xvla-v20-task0-mp-radio
+
 cd behavior1k_training
-python deploy_b1k.py --model_path <local-snapshot-of-this-repo> --port 8000
+python deploy_b1k.py --model_path ../xvla-v20-task0-mp-radio/ckpt-60000 --port 8000
 ```
 
 See [`INFERENCE_README.md`](https://github.com/markli1hoshipu/behavior1k-xvla/blob/v20/behavior1k_training/INFERENCE_README.md) for the protocol.
 
-## Files
+## Files in each `ckpt-XXXXX/` subfolder
 
 - `model.safetensors` — model weights (3.5 GB, bf16)
 - `config.json`, `preprocessor_config.json`, `tokenizer*`, `vocab.json`, `merges.txt` — config / tokenizer