nielsr HF Staff commited on
Commit
9119265
·
verified ·
1 Parent(s): f710177

Add MiniMax reported SWE-Bench Verified result

Browse files

This PR ensures the 80.2 score of the model card also shows up at https://huggingface.co/datasets/SWE-bench/SWE-bench_Verified.

.eval_results/swe_bench_verified.yaml CHANGED
@@ -6,4 +6,14 @@
6
  url: https://www.swebench.com/
7
  name: SWE-Bench official evaluation
8
  user: nielsr
9
- notes: high reasoning
 
 
 
 
 
 
 
 
 
 
 
6
  url: https://www.swebench.com/
7
  name: SWE-Bench official evaluation
8
  user: nielsr
9
+ notes: high reasoning, official
10
+
11
+ - dataset:
12
+ id: SWE-bench/SWE-bench_Verified
13
+ task_id: swe_bench_%_resolved
14
+ value: 80.2
15
+ source:
16
+ url: https://huggingface.co/MiniMaxAI/MiniMax-M2.5/
17
+ name: Model card
18
+ user: nielsr
19
+ notes: MiniMax reported number