LR2Bench / results /DeepSeek-R1.json
UltraRonin's picture
add
f5894fd
Raw History Blame Contribute Delete
1.41 kB
{
"config": {
"model_name": "deepseek-ai/DeepSeek-R1",
"link": "https://huggingface.co/deepseek-ai/DeepSeek-R1",
"Params": 671
},
"results": {
"Overall": {
"CR": "100.0",
"S-Acc": "58.4",
"EM": "20.0",
"PM-0.5": "62.0",
"Tokens": "9856"
},
"Acrostic": {
"CR": "100.0",
"S-Acc": "62.2",
"EM": "0.0",
"PM-0.5": "83.0",
"Tokens": "10077"
},
"Crossword": {
"CR": "100.0",
"S-Acc": "75.3",
"EM": "16.7",
"PM-0.5": "94.0",
"Tokens": "9810"
},
"Cryptogram": {
"CR": "100.0",
"S-Acc": "26.0",
"EM": "4.0",
"PM-0.5": "21.0",
"Tokens": "10344"
},
"Logic_Puzzle": {
"CR": "100.0",
"S-Acc": "69.4",
"EM": "42.5",
"PM-0.5": "68.0",
"Tokens": "9205"
},
"Sudoku": {
"CR": "100.0",
"S-Acc": "70.3",
"EM": "50.0",
"PM-0.5": "64.0",
"Tokens": "8277"
},
"Drop_Quote": {
"CR": "100.0",
"S-Acc": "47.3",
"EM": "7.0",
"PM-0.5": "42.0",
"Tokens": "11422"
}
}
}