Manipulation on DeepMind Manipulation tasks (train)
204Avg ReturnDrQ-v2
Evaluation Results
| Method | Links | |
|---|---|---|
| DrQ-v2Zero-shot=false, Training Steps=500k, Action Repeat=22022.12 | 204 | |
| PIE-GZero-shot=false, Backbone=ResNet-18, Training Steps=500k, Action Repeat=22022.12 | 199 | |
| DrQZero-shot=false, Training Steps=500k, Action Repeat=22022.12 | 130 | |
| TLDAZero-shot=false, Training Steps=500k, Action Repeat=22022.12 | 124 | |
| SVEAZero-shot=false, Training Steps=500k, Action Repeat=22022.12 | 49 | |
| SACZero-shot=false, Training Steps=500k, Action Repeat=22022.12 | 2.5 |