open-thought/reasoning-gym
1
0
Fork 0
mirror of https://github.com/open-thought/reasoning-gym.git synced 2026-04-26 17:13:17 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

[eval-v1] add timer

Browse source
This commit is contained in:
rishabhranawat 2025-02-10 21:48:44 -08:00
parent 247464a47d
commit df5438498e
2 changed files with 40 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

39
eval/results/summary_google_gemini-2.0-flash-001_20250210_214631.json Normal file
View file

@ -0,0 +1,39 @@
[
{
"dataset_name": "letter_counting",
"model": "google/gemini-2.0-flash-001",
"average_score": 0.059,
"total_examples": 10,
"timestamp": "2025-02-10T21:46:27.185026",
"config": {
"min_words": 5,
"max_words": 15,
"size": 10,
"seed": 42
}
},
{
"dataset_name": "propositional_logic",
"model": "google/gemini-2.0-flash-001",
"average_score": 0.059,
"total_examples": 10,
"timestamp": "2025-02-10T21:46:31.805110",
"config": {
"size": 10,
"seed": 42
}
},
{
"dataset_name": "leg_counting",
"model": "google/gemini-2.0-flash-001",
"average_score": 0.40199999999999997,
"total_examples": 10,
"timestamp": "2025-02-10T21:46:31.805665",
"config": {
"min_animals": 3,
"max_animals": 8,
"size": 10,
"seed": 42
}
}
]