Reinforcement Learning
PyTorch
English
Chinese
robotics
motion-tracking
mujoco
unitree-g1
SuperMimic_Dyn / release.json
unpackableorange's picture
Release E009 40k high-dynamic tracking checkpoint
669542a verified
Raw History Blame Contribute Delete
684 Bytes
{
"experiment": "E009",
"iteration": 40000,
"filename": "ckpt_0040000.pt",
"size_bytes": 221362447,
"sha256": "182b7a5d05b6c3e4364d6223d1cf52ba18e6573563e961f6e019d6d445340c3c",
"original_checkpoint_sha256": "d6fed9645ab4f934dfb3a64d49e90d37509b73b70081b6cc17eeef4bd6ebfe24",
"original_size_bytes": 221366063,
"metadata_changes": [
"config.log_root",
"config.data.library",
"config.data.sampling_manifest"
],
"tensor_count": 166,
"all_tensors_identical": true,
"cpu_actor_output_check": {
"batch_size": 4,
"obs_dim": 930,
"goal_dim": 689,
"action_dim": 29,
"bitwise_equal": true
},
"physics_reevaluation_performed": false
}