Search public pages, research tools, and SAVRN solutions.
Open-weight model · Robotics
by Jeremy jermo1120/pi05-b1k-task38-spraying-for-bugs
pi05-b1k-task38-spraying-for-bugs is an open-weight model for robotics from Jeremy, released under Apache License 2.0. Its published files total 12.6 GB.
This is a single-task vision-language-action policy for task 38 sprayingforbugs of the BEHAVIOR-1K benchmark. It is Pi0.5-based, has 3.4B parameters, and controls the R1Pro robot from three 224×224 RGB cameras plus proprioception.
By Jeremy, published under apache-2.0, revision 5ab0bd2f2a9a.
This is a single-task vision-language-action policy for task 38 sprayingforbugs of the BEHAVIOR-1K benchmark. It is Pi0.5-based, has 3.4B parameters, and controls the R1Pro robot from three 224×224 RGB cameras plus proprioception. It predicts 30-step action chunks. It was fine-tuned for 25,000 steps on task 38 demonstrations only. The starting point was checkpoint3 of the 1st-place 2025 BEHAVIOR Challenge solution (Larchenko, Zarin, Karnatak; paper). checkpoint3 had been trained on 13 tasks (4, 27, 31–33, 35–39, 41, 46, 49), task 38 among them. Status: training only. This model has not been evaluated in the BEHAVIOR simulator yet. Lower training loss does not guarantee a higher task success…
spraying_for_bugs"Pick up the pesticide atomizer in the garden and spray insectifuge to fully cover both potted plants in the garden."
This is a single-task vision-language-action policy for task 38 spraying_for_bugs of the BEHAVIOR-1K benchmark. It is Pi0.5-based, has 3.4B parameters, and controls the R1Pro robot from three 224×224 RGB cameras plus proprioception. It predicts 30-step action chunks.
It was fine-tuned for 25,000 steps on task 38 demonstrations only. The starting point was checkpoint_3 of the 1st-place 2025 BEHAVIOR Challenge solution (Larchenko, Zarin, Karnatak; paper). checkpoint_3 had been trained on 13 tasks (4, 27, 31–33, 35–39, 41, 46, 49), task 38 among them.
Status: training only. This model has not been evaluated in the BEHAVIOR simulator yet. Lower training loss does not guarantee a higher task success rate.
| Path | What it is |
|---|---|
params/ |
Model weights (orbax, ~12.6 GB), from the final training step (24,999) |
assets/IliaLarchenko/behavior_224_rgb/ |
Norm stats + FAST tokenizer, copied unchanged from checkpoint_3 |
changes.patch, behavior1k_decode.patch |
Code changes used for training (see below) |
metrics.json |
Logged training metrics every 100 steps |
Optimizer state is not included. Intermediate checkpoints (every 2,500 steps) are kept separately in jermo1120/b1k-task38-spraying-for-bugs.
| start (steps 0–500 avg) | end (last 1k steps avg) | |
|---|---|---|
| action_loss | 0.051 | 0.045 |
The action loss rose to ~0.077 during the first ~3k steps, plateaued, then fell steadily as the learning rate decayed (2.5k-step averages: 0.067, 0.076, 0.075, 0.071, 0.073, 0.061, 0.057, 0.052, 0.049, 0.046). Two single batches (around steps 2,400 and 11,200) were outliers.
Final logged step (24,900): total loss 0.078, action loss 0.045, subtask accuracy 0.996, FAST token accuracy 0.793, grad norm 1.20. These are single batches of 16 with no held-out validation split.
| Init | IliaLarchenko/behavior_submission/checkpoint_3/params (all weights, incl. task embeddings) |
| Data | Task 38 only: 200 episodes, ~1.30M frames (avg demo 3.6 min), from IliaLarchenko/behavior_224_rgb (224×224 RGB, 3 cameras) |
| Steps / batch | 25,000 steps × batch 16 (~31% of one epoch). Batch 32 runs out of memory on one 80 GB A100. |
| LR | cosine: warmup 200 → peak 5e-6 → 5e-7 at 25k |
| Model config | Same as the winners' pi_behavior_b1k_fast (FAST aux loss, correlated noise, 15 flow samples, frozen vision backbone) |
| Norm stats / tokenizer | Reused from checkpoint_3 (not recomputed on task 38) |
| Hardware / time | 1× A100-SXM4-80GB, 3.0 s/step, ~21 h of training steps |
| Code | behavior-1k-solution ca556f7 + changes.patch; BEHAVIOR-1K 684a83050 + behavior1k_decode.patch |
-g 4, with frame counts verified identical, because random-frame decoding was the bottleneck. The re-encode is lossy but visually near-identical at 224px.behavior1k_decode.patch). This is safe only with the short-GOP videos above.tasks filter to DataConfig. A weight-loader fix was needed for orbax 0.11.13 (changes.patch).uv sync today needed chex==0.1.89 (no longer pulled in by optax) and pyarrow==20.0.0 (newer versions require NumPy 2). Run training with uv run --no-sync so uv keeps them.TORCHINDUCTOR_COMPILE_THREADS=1 (stops each data worker from spawning a compile pool) and num_workers=12 fixed it in most cases. The run was resumed from checkpoints 3 times after such crashes (from steps 2,500, 22,500 and 22,500); resuming restores the weights, optimizer state and step count, so the LR schedule continued as planned. Steps 22,500–25,000 were therefore trained on three times, and only the final attempt was kept.Use with the winners' repo and config pi_behavior_b1k_fast (same architecture):
huggingface-cli download jermo1120/pi05-b1k-task38-spraying-for-bugs --local-dir ./ckpt
uv run scripts/serve_b1k.py policy:checkpoint --policy.config pi_behavior_b1k_fast --policy.dir ./ckpt
Then run the BEHAVIOR eval with task.name=spraying_for_bugs. The model's task embedding uses the 2025 numbering (task id 38). For the 2026 evaluator, override the incoming task id with 38.
Receding horizon control is done at inference time by the serving wrapper (B1KPolicyWrapper); training is unchanged. Defaults: execute 26 of the 30 predicted actions (--actions_to_execute 26), keep the next 4 to inpaint the start of the next chunk (actions_to_keep=4), and play the 26 actions back in 20 control steps (execute_in_n_steps=20).
Base model and training code: Ilia Larchenko, Gleb Zarin, Akash Karnatak (repo, checkpoints), built on Physical Intelligence's Pi0.5 / openpi. Benchmark: BEHAVIOR-1K, Stanford Vision and Learning Lab.
44 files, 12.6 GB in total.
| File | Type | Size | SHA-256 |
|---|---|---|---|
| assets/IliaLarchenko/behavior_224_rgb/fast_tokenizer/metadata.json | Configuration | 447 B | — |
| assets/IliaLarchenko/behavior_224_rgb/fast_tokenizer/processor_config.json | Configuration | 250 B | — |
| assets/IliaLarchenko/behavior_224_rgb/fast_tokenizer/special_tokens_map.json | Configuration | 3 B | — |
| assets/IliaLarchenko/behavior_224_rgb/norm_stats.json | Configuration | 18.0 MB | ccd14a0210fc |
| metrics.json | Configuration | 53.4 KB | — |
| README.md | Documentation | 6.0 KB | — |
| behavior1k_decode.patch | Other | 886 B | — |
| changes.patch | Other | 6.5 KB | — |
| loss_curve.png | Other | 78.1 KB | — |
| params/_METADATA | Other | 31.1 KB | — |
| params/_sharding | Other | 18.1 KB | — |
| params/array_metadatas/process_0 | Other | 12.7 KB | — |
| params/d/e2878c5aa4884f25d5e550bfadeae7a3 | Other | 2.9 KB | — |
| params/manifest.ocdbt | Other | 118 B | — |
| params/ocdbt.process_0/d/04665ea37d53d1bef7bf6c4284f5ced0 | Other | 112.5 MB | d7cb66b420bf |
| params/ocdbt.process_0/d/126ed20a7a597012abbfa4ba5818aea4 | Other | 1.2 KB | — |
| params/ocdbt.process_0/d/131e833087e509ba40aa310e8cfe5bfe | Other | 3.1 GB | 002acb6bcd2d |
| params/ocdbt.process_0/d/1a1eaf5db69204c1f9e773645ae061fa | Other | 1.3 KB | — |
| params/ocdbt.process_0/d/2118c88b9519cbe3ace037cfda539770 | Other | 559.1 MB | 4a5ad172a227 |
| params/ocdbt.process_0/d/319a9ca4b303ea0d35322df8f278c130 | Other | 950 B | — |
| params/ocdbt.process_0/d/41cef683021953b72be802976392aef4 | Other | 701.9 MB | 1d00886978dd |
| params/ocdbt.process_0/d/467f26af6ea638dfc591d4ff1d7e4e62 | Other | 496 B | — |
| params/ocdbt.process_0/d/5bb654973f8fed6d7d20fc25e697973c | Other | 2.9 KB | — |
| params/ocdbt.process_0/d/6ccb150678cdb9513b96e0da91ed59f6 | Other | 994.1 MB | 4b16d8dd83bb |
| params/ocdbt.process_0/d/6e317b1882dc94a54ecbe929742add4a | Other | 70.0 MB | 9748d5a5f508 |
| params/ocdbt.process_0/d/6e68a0f6e29c86d8ec1750c18a1a628e | Other | 1.4 KB | — |
| params/ocdbt.process_0/d/6f9268b39defa0524957c7991014937b | Other | 79.9 MB | 06738f9adcc2 |
| params/ocdbt.process_0/d/75f68c2e597ed9322144f41c0446d7bf | Other | 697.2 KB | 849a822da122 |
| params/ocdbt.process_0/d/7c268f34909b763542eb0d2ec02851a8 | Other | 2.2 GB | 18c6d883e7f3 |
| params/ocdbt.process_0/d/7e664ae1b07c38ef2c0a71fc7640bffd | Other | 840.2 MB | 2c4900fa09ef |
| params/ocdbt.process_0/d/7f6c5570be29306f3073fe7582e71f60 | Other | 1.1 GB | 0cb8f707ed1a |
| params/ocdbt.process_0/d/9fef35fd80dadcccf85a05097bc6179c | Other | 2.2 GB | bdca4eb016ea |
| params/ocdbt.process_0/d/a100f9fe63e066a1afaee0e3a2e5ae03 | Other | 398.9 MB | 4dd65f9990ee |
| params/ocdbt.process_0/d/cc48c8e10c163d7f31681003f8b5732a | Other | 676 B | — |
| params/ocdbt.process_0/d/d5fa837b2396cb0cad2a4c6dce3dd0a1 | Other | 213 B | — |
| params/ocdbt.process_0/d/e9bbc14072ded746c0eb382a32c74a64 | Other | 1.1 KB | — |
| params/ocdbt.process_0/d/fab62a4e3082e9268d2b27efb26ac63c | Other | 133.0 MB | 197a12b37fa7 |
| params/ocdbt.process_0/d/fb109a317435c6359c1ef07331f6467b | Other | 61.9 MB | bef8ccb72a98 |
| params/ocdbt.process_0/manifest.ocdbt | Other | 417 B | — |
| .gitattributes | Repository | 3.0 KB | — |
| assets/IliaLarchenko/behavior_224_rgb/fast_tokenizer/__pycache__/processing_action_tokenizer.cpython-311.pyc | Tokenizer | 8.7 KB | — |
| assets/IliaLarchenko/behavior_224_rgb/fast_tokenizer/processing_action_tokenizer.py | Tokenizer | 6.1 KB | — |
| assets/IliaLarchenko/behavior_224_rgb/fast_tokenizer/tokenizer.json | Tokenizer | 386.7 KB | — |
| assets/IliaLarchenko/behavior_224_rgb/fast_tokenizer/tokenizer_config.json | Tokenizer | 353 B | — |
Released by Jeremy through its official repository on Hugging Face. Read the license.
Yes. pi05-b1k-task38-spraying-for-bugs is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
LIBERO 4in1（liberospatial / liberoobject / liberogoal / libero10）共 53.19 GB 的 Wan2.2-VAE 编码 latent 缓存： 训练 LIBERO policy（action head / VLA）时直接读取 latent 缓存，避免重复 VAE 编码。 - 窗口模式: windowed（--windowed） - 输出:.pt 文件，每 episode 一个 - dataset（推荐）: https://huggingface.co/datasets/MangoGoes/libero4in1wan2.2vaelatentdataset - model（本仓库）: https://huggingface.co/MangoGoes/libero4in1wan2.2vaelatentcosmosstyle
Wan2.2-TI2V-5B video DiT + 48-joint action head, trainingmode=joint. The base is suhyeok's finalized B-method recipe: a teacher-forced (sigma=0.25) self-EMA teacher plus an iBOT prototype loss at L18 L18, gamma=0.01, two-view. On top of it the 3 PAST cond latent frames are pooled into one motion frame before a chosen block. These are NOT the surrogate ctxpool runs. The surrogate line (older base, pd8 x GA1) lives in hmkang/wamctxpoolxattn and hmkang/wamctxpoolavg. Do not compare across the two sets. Geometry: 4-latin (numframesin=25, numframesout=41, fdf 2) = 4 cond + 2 future latent slots, 96 tokens per latent frame, 576 tokens per row. Effective batch 16 clips x GA 2 x 2 views = 64 rows…
weighted/imatrix quants of https://huggingface.co/DeepCybo/PhysBrain1.5-8B For a convenient overview and download list, visit our model page for this model. static quants are available at https://huggingface.co/mradermacher/PhysBrain1.5-8B-GGUF This is a vision model - mmproj files (if any) will be in the static repository. If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality quant And here are Artefact2's thoughts on the matter…
static quants of https://huggingface.co/MirilAI/Miril-DroneVLM-2B-2 For a convenient overview and download list, visit our model page for this model. weighted/imatrix quants seem not to be available (by me) at this time. If they do not show up a week or so after the static ones, I have probably not planned for them. Feel free to request them by opening a Community Discussion. If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files. (sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants) Here is a handy graph by ikawrakow comparing some lower-quality quant…
Own lineage: PPO from scratch (nb1, commit 46c48b1) continued as nb2 and nb4 (commit fac88cc), checkpoint 850 (1.328 s); evolution strategies on the weights against the Arena sprint scorer (0.952 s); then the network frozen and wrapped in a step counter carried in the headroll action that adds joint-target and forward-command offsets from about 0.5 m to the line, found by CMA-ES on the event's official seeds: 0.940 s. No other entry's weights used. Simulation only. A perpetual policy for the microduck (61-D observation, 14 actions, 50 Hz). Runs until told otherwise — a gait, loaded into a policy slot. The observation normalizer is baked into policy.onnx; feed raw observations. manifest.json…
EgoSteer-3B-RealMan is the generalist policy from EgoSteer: A Full-Stack System Towards Steerable Dexterous Manipulation from Egocentric Videos. It is a world-model-enhanced Vision-Language-Action (VLA) policy built on a Qwen3-VL backbone with a flow-matching action expert, post-trained from EgoSteer-3B-Base on real-world demonstrations collected on the RealMan robot. This is a robot-ready dual-camera (head + chest) checkpoint, corresponding to EgoSteer-9.6K-DG in the paper. For the base model to fine-tune on your own data, see EgoSteer-3B-Base. Our full-stack system integrates EgoSmith (data pipeline), Robot Stack (deployment), and EgoSteer (policy) to learn from 9.6k hours of large-scale…