Add AntMaze benchmark for SAC-N and EDAC (tinkoff-ai#2)

- They were not originally benchmarked on these datasets, so further hyperparam search is certainly needed
lucas-yyy000 · Oct 2, 2022 · 9069495 · 9069495
1 parent 0be50c0
commit 9069495
Showing 1 changed file with 8 additions and 8 deletions.
diff --git a/README.md b/README.md
@@ -69,11 +69,11 @@ For learning curves and all the details, you can check the links above. Here, we
 #### Antmaze
 | **Task-Name**|BC|BC-10%|TD3 + BC|CQL|IQL|AWAC|SAC-N|EDAC|DT |
 |------------------------------|------------|--------|--------|-----|-----|------|-------|------|----|
-|antmaze-umaze-v0 | 51.50±8.81 | 0.00±0.00 | 93.25±1.50 | 67.00±6.24 | 74.50±11.03 | 63.50±9.33 | TBD±TBD | TBD±TBD | 52.75±11.47|
-|antmaze-medium-play-v0 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 68.00±12.77 | 0.00±0.00 | TBD±TBD | TBD±TBD | 0.00±0.00|
-|antmaze-large-play-v0 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 45.00±11.53 | 0.00±0.00 | TBD±TBD | TBD±TBD | 0.00±0.00|
+|antmaze-umaze-v0 | 51.50±8.81 | 0.00±0.00 | 93.25±1.50 | 63.75±8.26 | 74.50±11.03 | 63.50±9.33 | 0.00±0.00 | 29.25±33.35 | 52.75±11.47|
+|antmaze-medium-play-v0 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 71.50±12.56 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00|
+|antmaze-large-play-v0 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 40.75±12.69 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00|
 |                              |            |        |        |     |     |      |       |      |    |
-| **antmaze average**           | 17.17 | 0.00 | 31.08 | 22.33 | 62.50 | 21.17 | TBD | TBD | 17.58 |
+| **antmaze average**           | 17.17 | 0.00 | 31.08 | 21.25 | 62.25 | 21.17 | 0.00 | 9.75 | 17.58 |
 
 ### Best Scores
 #### Gym-MuJoCo
@@ -104,11 +104,11 @@ For learning curves and all the details, you can check the links above. Here, we
 #### Antmaze
 | **Task-Name**|BC|BC-10%|TD3 + BC|CQL|IQL|AWAC|SAC-N|EDAC|DT |
 |------------------------------|------------|--------|--------|-----|-----|------|-------|------|----|
-|antmaze-umaze-v0 | 71.25±9.07 | 0.00±0.00 | 97.75±1.50 | 78.33±4.73 | 87.00±2.94 | 74.75±8.77 | TBD±TBD | TBD±TBD | 65.50±8.96|
-|antmaze-medium-play-v0 | 4.75±2.22 | 0.00±0.00 | 6.00±2.00 | 2.00±1.41 | 85.33±2.08 | 14.00±11.80 | TBD±TBD | TBD±TBD | 1.00±2.00|
-|antmaze-large-play-v0 | 0.75±0.50 | 0.00±0.00 | 0.33±0.58 | 0.00±0.00 | 56.00±4.00 | 0.00±0.00 | TBD±TBD | TBD±TBD | 0.00±0.00|
+|antmaze-umaze-v0 | 71.25±9.07 | 0.00±0.00 | 97.75±1.50 | 76.75±4.99 | 87.00±2.94 | 74.75±8.77 | 0.00±0.00 | 75.00±27.51 | 65.50±8.96|
+|antmaze-medium-play-v0 | 4.75±2.22 | 0.00±0.00 | 6.00±2.00 | 1.75±0.96 | 86.00±2.16 | 14.00±11.80 | 0.00±0.00 | 0.00±0.00 | 1.00±2.00|
+|antmaze-large-play-v0 | 0.75±0.50 | 0.00±0.00 | 0.50±0.58 | 0.00±0.00 | 53.00±6.83 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00 | 0.00±0.00|
 |                              |            |        |        |     |     |      |       |      |    |
-| **antmaze average**           | 25.58 | 0.00 | 34.69 | 26.78 | 76.11 | 29.58 | TBD | TBD | 22.17 |
+| **antmaze average**           | 25.58 | 0.00 | 34.75 | 26.17 | 75.33 | 29.58 | 0.00 | 25.00 | 22.17 |
 
 ## Citing CORL