Rename configs from wr* to benchmark_wr* to match codebase convention

README.md

@@ -13,30 +13,30 @@ pretty_name: MAPF-FrozenLake Benchmark
 size_categories:
   - 1K<n<10K
 configs:
-  - config_name: wr025
+  - config_name: benchmark_wr025
     data_files:
       - split: 3_agents
-        path: wr025/3_agents.jsonl
+        path: benchmark_wr025/3_agents.jsonl
       - split: 4_agents
-        path: wr025/4_agents.jsonl
+        path: benchmark_wr025/4_agents.jsonl
       - split: 5_agents
-        path: wr025/5_agents.jsonl
-  - config_name: wr050
+        path: benchmark_wr025/5_agents.jsonl
+  - config_name: benchmark_wr050
     data_files:
       - split: 3_agents
-        path: wr050/3_agents.jsonl
+        path: benchmark_wr050/3_agents.jsonl
       - split: 4_agents
-        path: wr050/4_agents.jsonl
+        path: benchmark_wr050/4_agents.jsonl
       - split: 5_agents
-        path: wr050/5_agents.jsonl
-  - config_name: wr075
+        path: benchmark_wr050/5_agents.jsonl
+  - config_name: benchmark_wr075
     data_files:
       - split: 3_agents
-        path: wr075/3_agents.jsonl
+        path: benchmark_wr075/3_agents.jsonl
       - split: 4_agents
-        path: wr075/4_agents.jsonl
+        path: benchmark_wr075/4_agents.jsonl
       - split: 5_agents
-        path: wr075/5_agents.jsonl
+        path: benchmark_wr075/5_agents.jsonl
 ---
 
 # MAPF-FrozenLake Benchmark
@@ -44,9 +44,9 @@ configs:
 Evaluation benchmark for the paper
 **[From Trainee to Trainer: LLM-Designed Training Environment for RL with Multi-Agent Reasoning](https://github.com/LARK-AI-Lab/Trainee-to-Trainer)**.
 
-Three configs (`wr025` / `wr050` / `wr075`) correspond to the
-wait-ratio threshold of the underlying CBS-optimal solution
-(higher = more inter-agent coordination required).
+Three configs (`benchmark_wr025` / `benchmark_wr050` / `benchmark_wr075`)
+correspond to the wait-ratio threshold of the underlying CBS-optimal
+solution (higher = more inter-agent coordination required).
 Each config has three splits by agent count.
 
 ## Load
@@ -55,35 +55,29 @@ Each config has three splits by agent count.
 from datasets import load_dataset
 
 ds = load_dataset("LARK-Lab/MAPF-FrozenLake-Benchmark",
-                  name="wr075", split="5_agents")
+                  name="benchmark_wr075", split="5_agents")
 print(ds[0]["text"][:400])
 ```
 
 ## Run evaluation
 
-Place the benchmark under the
+Drop the downloaded folders directly into the
 [Trainee-to-Trainer](https://github.com/LARK-AI-Lab/Trainee-to-Trainer)
-repo root at one of `benchmark_wr025/` / `benchmark_wr050/` /
-`benchmark_wr075/`, with this exact layout:
+repo root &mdash; the directory names already match what the evaluators
+expect (`benchmark_wr025/` / `benchmark_wr050/` / `benchmark_wr075/`).
+Each one must contain `<N>_agents/dataset_nl.jsonl`.
 
-```
-benchmark_wr075/
-├── 3_agents/dataset_nl.jsonl
-├── 4_agents/dataset_nl.jsonl
-└── 5_agents/dataset_nl.jsonl
-```
-
-You can download + lay out everything in one go:
+One-shot download + layout:
 
 ```bash
 hf download LARK-Lab/MAPF-FrozenLake-Benchmark \
   --repo-type dataset --local-dir /tmp/mapf_bench
 
-for wr in wr025 wr050 wr075; do
+for wr in benchmark_wr025 benchmark_wr050 benchmark_wr075; do
   for n in 3 4 5; do
-    mkdir -p benchmark_${wr}/${n}_agents
+    mkdir -p ${wr}/${n}_agents
     cp /tmp/mapf_bench/${wr}/${n}_agents.jsonl \
-       benchmark_${wr}/${n}_agents/dataset_nl.jsonl
+       ${wr}/${n}_agents/dataset_nl.jsonl
   done
 done
 ```