Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Website
Tasks
HuggingChat
Collections
Languages
Organizations
Community
Blog
Posts
Daily Papers
Hardware
Learn
Discord
Forum
GitHub
Solutions
Team & Enterprise
Hugging Face PRO
Enterprise Support
Inference Providers
Inference Endpoints
Storage Buckets
Log In
Sign Up
SeonghoonYu
/
RACE2_vlabench
Like
0
Robotics
Safetensors
vlabench
pi0.5
race2
License:
mit
Model card
Files
Files and versions
xet
Community
Copy to bucket
new
main
RACE2_vlabench
340 GB
Ctrl+K
Ctrl+K
1 contributor
History:
70 commits
SeonghoonYu
plainft_fullft_h20_40k: README, eval (2460 episodes, 5-model comparison), provenance, training config, source
16491f6
verified
9 days ago
ablation
ablation/abl_randcnt_h20_40k: RACE2 random-label (random count + position) H20 40k EMA weights (full model, overlay, norm stats)
15 days ago
acot_h10_40k
Add ACoT-VLA H=10 40k EMA params (JAX/orbax)
22 days ago
acot_h20_40k
Add ACoT-VLA H=20 40k EMA params (JAX/orbax)
19 days ago
data
Add files using upload-large-folder tool
17 days ago
fullft_h20_40k
Upload folder using huggingface_hub
14 days ago
h10_40k
race2 h10 40k EMA: full serving checkpoint (model.safetensors + assets) and slim overlay
25 days ago
h15_40k
race2 h15 40k EMA: full serving checkpoint (model.safetensors + assets) and slim overlay
25 days ago
h20_40k
h20_40k: add full RACE2 H20 training log (40k steps, 2 GPU x micro 32, 2026-09-07/08)
19 days ago
h25_40k
Add files using upload-large-folder tool
22 days ago
h25_support
Add files using upload-large-folder tool
22 days ago
h30_40k
Add RACE2 h30_40k (40k EMA, seed 42)
21 days ago
h30_h35_support
Add H30/H35 loading support (config overlay)
21 days ago
h35_40k
Add RACE2 h35_40k (40k EMA, seed 42)
21 days ago
h40_40k
Add h40_40k (RACE2 VLABench, 40k EMA + full serving model, logs, provenance)
15 days ago
h50_40k
Add h50_40k (RACE2 VLABench, 40k EMA + full serving model, logs, provenance)
15 days ago
h5_40k
Add h5_40k (RACE2 VLABench, 40k EMA + full serving model, logs, provenance)
15 days ago
plainft_fullft_h20_40k
plainft_fullft_h20_40k: README, eval (2460 episodes, 5-model comparison), provenance, training config, source
9 days ago
plainft_h10_40k
Add naive-FT (expert-only, VLM+vision frozen) h10 step 40000 β pi0.5 VLABench teacher + 40k, batch 64, warmup 500 -> 5e-5 -> cosine -> 5e-6, EMA 0.999
25 days ago
plainft_h15_40k
Add naive-FT (expert-only, VLM+vision frozen) h15 step 40000 β pi0.5 VLABench teacher + 40k, batch 64, warmup 500 -> 5e-5 -> cosine -> 5e-6, EMA 0.999
25 days ago
plainft_h20_40k
Add naive-FT (expert-only, VLM+vision frozen) h20 step 40000 β pi0.5 VLABench teacher + 40k, batch 64, warmup 500 -> 5e-5 -> cosine -> 5e-6, EMA 0.999
25 days ago
plainft_h25_40k
Add files using upload-large-folder tool
22 days ago
plainft_h30_40k
Upload folder using huggingface_hub
21 days ago
teacher_official
Add the official VLABench pi0.5 teacher (VLABench/pi05-primitive-10task, JAX -> PyTorch via examples/convert_jax_model_to_pytorch.py) with its 32-dim norm stats β the init of every run here
25 days ago
.gitattributes
Safe
3.2 kB
Move abl_timing_only_h20_40k into ablation/ alongside A6-A9
18 days ago
README.md
Safe
1.93 kB
Add files using upload-large-folder tool
17 days ago