hooneyskywalker/g1-motion-tracking-policies
Add the held-out, perturbation and transition results
Add student/policy.onnx
Add student/eval_push_teacher_vs_student.json
Add student/eval_transition.json
Add student/eval_heldout_content.json
Add student/eval_heldout_repetition.json
Document the distilled student and what it took
Add the distilled student that covers fourteen sequences
Add the distilled student that covers fourteen sequences
Add the distilled student that covers fourteen sequences
Add the distilled student that covers fourteen sequences
Use the metric names the papers use
Name the metrics and show completion without the three dead sequences
Label the sample size per row, since it is not the same
Show the sequence where the two simulators agree most
Lead the transfer section with the three numbers that answer it
Say why the pipeline exists
Read the per-sequence table instead of just printing it
Say what each column of the per-sequence table means
Put the per-sequence table in the card, not only in the image
Drop the sim-to-sim numbers the old protocol produced
Order the per-sequence table by completion, matching the results table
Re-measure sim-to-sim with both simulators on one protocol
Re-render the Isaac panel
Show the clips, not just the tables
Upload reset_walk3_subject4.gif with huggingface_hub
Upload reset_walk3_subject1.gif with huggingface_hub
Upload tracking.gif with huggingface_hub
Upload kobe.gif with huggingface_hub
Upload sim2sim.gif with huggingface_hub
Link the code and the video playlist
Upload eval_polysim/kobe_level1.json with huggingface_hub
Both pass criteria, the reference defects, and kobe
Score the same rollouts under both standards
Match the two columns to the same definition
Fill the ten-seed column
Replace with the full-length table
Upload README.md with huggingface_hub
Upload eval_polysim/walk4_subject1.json with huggingface_hub
Upload eval_polysim/walk3_subject5.json with huggingface_hub
Upload eval_polysim/walk3_subject4.json with huggingface_hub
Upload eval_polysim/walk3_subject2.json with huggingface_hub
Upload eval_polysim/walk3_subject1.json with huggingface_hub
Upload eval_polysim/walk2_subject4.json with huggingface_hub
Upload eval_polysim/walk2_subject3.json with huggingface_hub
Upload eval_polysim/walk2_subject1.json with huggingface_hub
Upload eval_polysim/walk1_subject5.json with huggingface_hub
Upload eval_polysim/walk1_subject2.json with huggingface_hub
Upload eval_polysim/walk1_subject1.json with huggingface_hub
Upload eval_polysim/run2_subject4.json with huggingface_hub