Transformer L18 Depth4 CC-Gate LIBERO no-LoRA fullVLM

30k-step LIBERO finetuned checkpoint for the Transformer action-mixer ablation with pooled ctxD gated conditioning.

Checkpoint

  • OpenPI config: pi05_transformer_ctxd18_gated_fullvlm_libero_5090
  • Experiment: icra_nolora_fullvlm_transformer_gated_l18_depth4_cosine5e5_5e6_b16_acc1_seed42_resub
  • Step: 30000
  • Initializer: pi05_base_pytorch
  • Adaptation: full VLM / full vision / full action path, no LoRA
  • VLM prefix: L18
  • Mixer: Transformer depth4
  • Context interface: CC-Gate, pooled ctxD gate
  • Training: batch 16, grad accumulation 1, seed 42
  • LR schedule: cosine from 5e-5 to 5e-6 over 30k steps
  • LIBERO eval: seed 7, 50 trials per task, 10 denoising steps

LIBERO Results

Mixer Context Spatial Object Goal LIBERO-10 Average Action expert core Train time Median latency
Transformer CC-Gate 82.8 90.6 86.8 82.6 85.7 0.030B 13.69 h 74.22 ms

Use

--config-name pi05_transformer_ctxd18_gated_fullvlm_libero_5090
--checkpoint-dir /path/to/checkpoint

The checkpoint includes the LIBERO norm stats under assets/physical-intelligence/libero.

Downloads last month

-

Downloads are not tracked for this model. How to track
Safetensors
Model size
3B params
Tensor type
F32
·
BF16
·
Video Preview
loading

Dataset used to train giakhuyendihoc/transformer-gated-l18-depth4-libero-nolora-fullvlm