Bofang
dddddsds31341
AI & ML interests
None yet
Recent Activity
published a dataset 3 minutes ago
dddddsds31341/longwam-rollout-20261113-154752-sync-3741484-ee63fd99 updated a collection 3 minutes ago
Kai0 Piper Rollouts updated a dataset 3 minutes ago
dddddsds31341/longwam-rollout-20261113-154333-sync-3702409-b29492b3Organizations
None yet
Long-WAM putting-tennis evaluation checkpoints.
Intermediate RECAP value-model evaluation weights with matching configurations, data statistics and code.
-
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-500
Robotics • Updated • 20 -
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-1000
Robotics • Updated • 20 -
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-1500
Robotics • Updated • 20 -
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-2000
Robotics • Updated • 21
SIP evaluation models for nesting Russian dolls.
Evaluation weights from the RLinf RLT Stage 1 Insert Tube experiment, with model config, task normalization and code; no optimizer state.
Evaluation weights from the standardized RLinf RLT Stage 1 bread experiment; excludes optimizer state.
Inference weights from SIP run fold-towel-3ep-epochckpt-20261030-023552, with its configuration, normalization statistics and code.
Kai0 JAX π0.5 joint fine-tuning on putting tennis, nesting doll, fold towel and sorting dolls; inference checkpoints from step 5,000 to 40,005.
Long-WAM-YAM evaluation checkpoints fine-tuned on wash_cup.
Franka whiteboard-wiping checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Long-WAM-YAM fine-tuning on 100 nesting-doll demonstrations: epoch 1, 2 and 3 evaluation checkpoints.
Franka tennis-ball checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Franka robot manipulation datasets in LeRobot and RLDS formats.
Public OpenPI π0 and π0.5 checkpoints for nesting_3dolls_100, organized by training step.
Public VLA-Adapter checkpoints for pick_balls (100 demos), organized by training step from 10k through 150k.
OpenPI checkpoints for nesting_doll, organized by model version, demo count, and training step.
Public VLA-Adapter checkpoints for nesting_3dolls (99 valid demos), organized by training step from 10k through 150k.
OpenPI checkpoints for handover_cup, organized by model version, demo count, and training step.
OpenPI checkpoints for sorting_dolls, organized by model version, demo count, and training step.
Public OpenPI checkpoints for bottle_100, organized by model version and training step (5k, 10k, 15k, and 20k).
OpenPI checkpoints for sweep_trash, organized by model version, demo count, and training step.
OpenPI checkpoints for fold_towel, organized by model version, demo count, and training step.
Evaluation weights for Franka Test Tube RL Token Stage 1, with matching configuration, normalization statistics and source code.
Long-WAM fold-towel evaluation checkpoints from run 20261107.
Native DCP training states with optimizer, scheduler, RNG, task configuration and source code.
Evaluation weights from the RLinf RLT Stage 1 Wipe Whiteboard experiment, with model config, task normalization and code; no optimizer state.
Evaluation weights from sorting-dolls, with exact configuration, normalization statistics and code.
-
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-5000
Robotics • Updated • 35 -
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-10000
Robotics • Updated • 35 -
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-15000
Robotics • Updated • 32 -
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-20000
Robotics • Updated • 41
Public Kai0 Piper robot policy rollouts in LeRobot v2.1 format. Each dataset contains one recorded rollout.
-
dddddsds31341/kai0-rollout-20261104t063715z-rtc-b9d9051e-b45bcd0b
ALOHA • Updated • 1 episodes • 188 -
dddddsds31341/kai0-rollout-20261104t064214z-rtc-7ae671c8-21d3e22d
ALOHA • Updated • 1 episodes • 203 -
dddddsds31341/kai0-rollout-20261104t064539z-rtc-91be670a-427986ba
ALOHA • Updated • 1 episodes • 195 -
dddddsds31341/kai0-rollout-20261104t064411z-rtc-a3d1de0a-7bfdb198
ALOHA • Updated • 1 episodes • 198
Inference weights from SIP run nesting-doll-unclipped-30k-20261103-020414, with its configuration, normalization statistics and code.
-
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-30k-Step-5000
Robotics • Updated • 22 -
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-Step-5000
Robotics • Updated • 16 -
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-30k-Step-10000
Robotics • Updated • 24 -
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-Step-10000
Robotics • Updated • 17
Long-WAM-YAM evaluation checkpoints fine-tuned on fold_towel.
Long-WAM-YAM evaluation checkpoints fine-tuned on sweep_trash.
Franka test-tube placement checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Franka gray-tube insertion checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Franka Bread checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Public OpenPI π0 and π0.5 checkpoints for pick_balls_100, organized by training step.
OpenPI checkpoints for pen_uncap, organized by model version, demo count, and training step.
OpenPI checkpoints for putting_tennis, organized by model version, demo count, and training step.
Public OpenPI π0 and π0.5 checkpoints for food_100, organized by training step.
OpenPI checkpoints for water_pouring, organized by model version, demo count, and training step.
OpenPI checkpoints for wash_cup, organized by model version, demo count, and training step.
Public VLA-Adapter checkpoints for food (100 demos), organized by training step from 10k through 150k.
Public VLA-Adapter checkpoints for bottle (100 demos), organized by training step from 10k through the final 150005 checkpoint.
OpenPI checkpoints for wash_dish, organized by model version, demo count, and training step.
Evaluation weights for Franka Test Tube RL Token Stage 1, with matching configuration, normalization statistics and source code.
Long-WAM fold-towel evaluation checkpoints from run 20261107.
Long-WAM putting-tennis evaluation checkpoints.
Intermediate RECAP value-model evaluation weights with matching configurations, data statistics and code.
-
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-500
Robotics • Updated • 20 -
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-1000
Robotics • Updated • 20 -
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-1500
Robotics • Updated • 20 -
dddddsds31341/RLinf-RECAP-Franka-Bread-Value-Step-2000
Robotics • Updated • 21
Native DCP training states with optimizer, scheduler, RNG, task configuration and source code.
SIP evaluation models for nesting Russian dolls.
Evaluation weights from the RLinf RLT Stage 1 Wipe Whiteboard experiment, with model config, task normalization and code; no optimizer state.
Evaluation weights from the RLinf RLT Stage 1 Insert Tube experiment, with model config, task normalization and code; no optimizer state.
Evaluation weights from sorting-dolls, with exact configuration, normalization statistics and code.
-
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-5000
Robotics • Updated • 35 -
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-10000
Robotics • Updated • 35 -
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-15000
Robotics • Updated • 32 -
dddddsds31341/Long-WAM-Sorting-Dolls-20k-Step-20000
Robotics • Updated • 41
Evaluation weights from the standardized RLinf RLT Stage 1 bread experiment; excludes optimizer state.
Public Kai0 Piper robot policy rollouts in LeRobot v2.1 format. Each dataset contains one recorded rollout.
-
dddddsds31341/kai0-rollout-20261104t063715z-rtc-b9d9051e-b45bcd0b
ALOHA • Updated • 1 episodes • 188 -
dddddsds31341/kai0-rollout-20261104t064214z-rtc-7ae671c8-21d3e22d
ALOHA • Updated • 1 episodes • 203 -
dddddsds31341/kai0-rollout-20261104t064539z-rtc-91be670a-427986ba
ALOHA • Updated • 1 episodes • 195 -
dddddsds31341/kai0-rollout-20261104t064411z-rtc-a3d1de0a-7bfdb198
ALOHA • Updated • 1 episodes • 198
Inference weights from SIP run fold-towel-3ep-epochckpt-20261030-023552, with its configuration, normalization statistics and code.
Inference weights from SIP run nesting-doll-unclipped-30k-20261103-020414, with its configuration, normalization statistics and code.
-
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-30k-Step-5000
Robotics • Updated • 22 -
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-Step-5000
Robotics • Updated • 16 -
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-30k-Step-10000
Robotics • Updated • 24 -
dddddsds31341/Long-WAM-Nesting-Doll-Unclipped-Step-10000
Robotics • Updated • 17
Kai0 JAX π0.5 joint fine-tuning on putting tennis, nesting doll, fold towel and sorting dolls; inference checkpoints from step 5,000 to 40,005.
Long-WAM-YAM evaluation checkpoints fine-tuned on fold_towel.
Long-WAM-YAM evaluation checkpoints fine-tuned on wash_cup.
Long-WAM-YAM evaluation checkpoints fine-tuned on sweep_trash.
Franka whiteboard-wiping checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Franka test-tube placement checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Long-WAM-YAM fine-tuning on 100 nesting-doll demonstrations: epoch 1, 2 and 3 evaluation checkpoints.
Franka gray-tube insertion checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Franka tennis-ball checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Franka Bread checkpoints fine-tuned from Physical Intelligence π0.5 base on Franka demonstrations.
Franka robot manipulation datasets in LeRobot and RLDS formats.
Public OpenPI π0 and π0.5 checkpoints for pick_balls_100, organized by training step.
Public OpenPI π0 and π0.5 checkpoints for nesting_3dolls_100, organized by training step.
OpenPI checkpoints for pen_uncap, organized by model version, demo count, and training step.
Public VLA-Adapter checkpoints for pick_balls (100 demos), organized by training step from 10k through 150k.
OpenPI checkpoints for putting_tennis, organized by model version, demo count, and training step.
OpenPI checkpoints for nesting_doll, organized by model version, demo count, and training step.
Public OpenPI π0 and π0.5 checkpoints for food_100, organized by training step.
Public VLA-Adapter checkpoints for nesting_3dolls (99 valid demos), organized by training step from 10k through 150k.
OpenPI checkpoints for water_pouring, organized by model version, demo count, and training step.
OpenPI checkpoints for handover_cup, organized by model version, demo count, and training step.
OpenPI checkpoints for wash_cup, organized by model version, demo count, and training step.
OpenPI checkpoints for sorting_dolls, organized by model version, demo count, and training step.
Public VLA-Adapter checkpoints for food (100 demos), organized by training step from 10k through 150k.
Public OpenPI checkpoints for bottle_100, organized by model version and training step (5k, 10k, 15k, and 20k).
Public VLA-Adapter checkpoints for bottle (100 demos), organized by training step from 10k through the final 150005 checkpoint.
OpenPI checkpoints for sweep_trash, organized by model version, demo count, and training step.
OpenPI checkpoints for wash_dish, organized by model version, demo count, and training step.
OpenPI checkpoints for fold_towel, organized by model version, demo count, and training step.