YANG-Cheng
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -127,13 +127,13 @@ And statistics of the evaluation data of 9 environments are as follows:
|
|
127 |
|
128 |
| | AlfWorld | ScienceWorld | BabyAI | Jericho | PDDL | WebShop | WebArena | Tool-Query | Tool-Operation |
|
129 |
|-------|----------|--------------|--------|---------|------|---------|----------|------------|----------------|
|
130 |
-
| **Progress Rate Metric** | subgoal | subgoal | subgoal | subgoal | match | match | match | subgoal | subgoal/match |
|
131 |
-
| **\#Avg. Turn** | 6 | 15 | 10 | 20 | 20 | 3 | 25 | 5 | 6 |
|
132 |
-
| **\#Avg. Action Space** | 13 | 21 | 8 | 150 | 8 | 2 | 12 | 15 | 16 |
|
133 |
-
| **\#Avg. Context Length** | 900 | 2800 | 1800 | 1500 | 2700 | 1200 | 15000 | 2100 | 4300 |
|
134 |
-
| **\#Avg. Subgoals** | 3 | 5 | 4 | 6 | 6 | 4 | 6 | 5 | 5 |
|
135 |
| **\#Environment** | 134 | 90 | 112 | 20 | 60 | 251 | 245 | 60 | 40 |
|
136 |
-
|
|
|
|
|
|
|
|
|
|
|
137 |
|
138 |
To help researchers quickly understand evaluation data of each task, we provide **Dataset Viewer** at Huggingface Dataset: [🤗 AgentBoard](https://huggingface.co/datasets/hkust-nlp/agentboard/).
|
139 |
|
|
|
127 |
|
128 |
| | AlfWorld | ScienceWorld | BabyAI | Jericho | PDDL | WebShop | WebArena | Tool-Query | Tool-Operation |
|
129 |
|-------|----------|--------------|--------|---------|------|---------|----------|------------|----------------|
|
|
|
|
|
|
|
|
|
|
|
130 |
| **\#Environment** | 134 | 90 | 112 | 20 | 60 | 251 | 245 | 60 | 40 |
|
131 |
+
| **\#Turn** | 6 | 15 | 10 | 20 | 20 | 3 | 25 | 5 | 6 |
|
132 |
+
| **\#Action Space** | 13 | 21 | 8 | 150 | 8 | 2 | 12 | 15 | 16 |
|
133 |
+
| **\#Context Length** | 900 | 2800 | 1800 | 1500 | 2700 | 1200 | 15000 | 2100 | 4300 |
|
134 |
+
| **Progress Rate** | subgoal | subgoal | subgoal | subgoal | match | match | match | subgoal | subgoal/match |
|
135 |
+
| **\#Avg. Subgoals** | 3 | 5 | 4 | 6 | 6 | 4 | 6 | 5 | 5 |
|
136 |
+
| **Hard/Easy Cutoff** | 3 | 3 | 3 | 4 | 6 | 1 | 4 | 4 | 4 |
|
137 |
|
138 |
To help researchers quickly understand evaluation data of each task, we provide **Dataset Viewer** at Huggingface Dataset: [🤗 AgentBoard](https://huggingface.co/datasets/hkust-nlp/agentboard/).
|
139 |
|