Added more expert shards, updating readme
Browse files- pusht_dataset/README.md +1 -1
pusht_dataset/README.md
CHANGED
|
@@ -6,7 +6,7 @@ Video + action data from the [gym-pusht](https://github.com/huggingface/gym-push
|
|
| 6 |
|-------|----------|-------------|-------|-------------|
|
| 7 |
| `smooth/` | ~38,900 | 300 | ~324 hrs | Random smooth movement (Ornstein-Uhlenbeck process) |
|
| 8 |
| `goal/` | ~8,700 | 298 | ~73 hrs | Heuristic goal-directed policy (keypoint matching) |
|
| 9 |
-
| `expert/` | ~
|
| 10 |
|
| 11 |
## File format
|
| 12 |
|
|
|
|
| 6 |
|-------|----------|-------------|-------|-------------|
|
| 7 |
| `smooth/` | ~38,900 | 300 | ~324 hrs | Random smooth movement (Ornstein-Uhlenbeck process) |
|
| 8 |
| `goal/` | ~8,700 | 298 | ~73 hrs | Heuristic goal-directed policy (keypoint matching) |
|
| 9 |
+
| `expert/` | ~21,800 | 228 | ~138 hrs | Pretrained diffusion policy, ~74% success rate |
|
| 10 |
|
| 11 |
## File format
|
| 12 |
|