Datasets:
Upload stats_p1_35b_think.json with huggingface_hub
Browse files- stats_p1_35b_think.json +15 -0
stats_p1_35b_think.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"phase": "p1_35b_think",
|
| 3 |
+
"model": "Qwen3.5-35B-A3B",
|
| 4 |
+
"thinking": "true",
|
| 5 |
+
"trajectories": 38136,
|
| 6 |
+
"distinct_tasks_covered": 9534,
|
| 7 |
+
"k_effective_vs_covered_tasks": 4.0,
|
| 8 |
+
"k_effective_vs_full_corpus": 4.0,
|
| 9 |
+
"screenshots": 2346434,
|
| 10 |
+
"shards": 326,
|
| 11 |
+
"skipped_below_min_reward": 0,
|
| 12 |
+
"mean_reward": 0.1253,
|
| 13 |
+
"golden_rate": 0.2275,
|
| 14 |
+
"perfect_1_0_rate": 0.0503
|
| 15 |
+
}
|