Dataset card: enable Viewer for the two public benchmark files (prompts, per_item_results); add size_categories n<1K; body unchanged

README.md

@@ -17,6 +17,20 @@ tags:
 - expert-advisor
 - code-generation
 - benchmark
+
+size_categories:
+- n<1K
+
+configs:
+- config_name: prompts
+  default: true
+  data_files:
+  - split: test
+    path: prompts.jsonl
+- config_name: per_item_results
+  data_files:
+  - split: test
+    path: per_item_results.jsonl
 ---
 
 # CompilingThings Compile Benchmark for MQL5®