data_source stringclasses 1
value | prompt listlengths 1 1 | ability stringclasses 1
value | reward_model dict | extra_info dict |
|---|---|---|---|---|
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [44, 19, 35], create an equation that equals 98. You can use basi... | math | {
"ground_truth": {
"numbers": [
44,
19,
35
],
"target": 98
},
"style": "rule"
} | {
"index": 0,
"solution": {
"numbers": [
44,
19,
35
],
"target": 98
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [63, 95, 96], create an equation that equals 64. You can use basi... | math | {
"ground_truth": {
"numbers": [
63,
95,
96
],
"target": 64
},
"style": "rule"
} | {
"index": 1,
"solution": {
"numbers": [
63,
95,
96
],
"target": 64
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [95, 11, 56], create an equation that equals 28. You can use basi... | math | {
"ground_truth": {
"numbers": [
95,
11,
56
],
"target": 28
},
"style": "rule"
} | {
"index": 2,
"solution": {
"numbers": [
95,
11,
56
],
"target": 28
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [19, 74, 45], create an equation that equals 48. You can use basi... | math | {
"ground_truth": {
"numbers": [
19,
74,
45
],
"target": 48
},
"style": "rule"
} | {
"index": 3,
"solution": {
"numbers": [
19,
74,
45
],
"target": 48
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [49, 41, 73], create an equation that equals 17. You can use basi... | math | {
"ground_truth": {
"numbers": [
49,
41,
73
],
"target": 17
},
"style": "rule"
} | {
"index": 4,
"solution": {
"numbers": [
49,
41,
73
],
"target": 17
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [48, 28, 42], create an equation that equals 62. You can use basi... | math | {
"ground_truth": {
"numbers": [
48,
28,
42
],
"target": 62
},
"style": "rule"
} | {
"index": 5,
"solution": {
"numbers": [
48,
28,
42
],
"target": 62
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [83, 78, 1, 39], create an equation that equals 82. You can use b... | math | {
"ground_truth": {
"numbers": [
83,
78,
1,
39
],
"target": 82
},
"style": "rule"
} | {
"index": 6,
"solution": {
"numbers": [
83,
78,
1,
39
],
"target": 82
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [67, 21, 31, 20], create an equation that equals 98. You can use ... | math | {
"ground_truth": {
"numbers": [
67,
21,
31,
20
],
"target": 98
},
"style": "rule"
} | {
"index": 7,
"solution": {
"numbers": [
67,
21,
31,
20
],
"target": 98
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [96, 74, 94, 54], create an equation that equals 40. You can use ... | math | {
"ground_truth": {
"numbers": [
96,
74,
94,
54
],
"target": 40
},
"style": "rule"
} | {
"index": 8,
"solution": {
"numbers": [
96,
74,
94,
54
],
"target": 40
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [81, 84, 62], create an equation that equals 65. You can use basi... | math | {
"ground_truth": {
"numbers": [
81,
84,
62
],
"target": 65
},
"style": "rule"
} | {
"index": 9,
"solution": {
"numbers": [
81,
84,
62
],
"target": 65
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [2, 54, 17], create an equation that equals 35. You can use basic... | math | {
"ground_truth": {
"numbers": [
2,
54,
17
],
"target": 35
},
"style": "rule"
} | {
"index": 10,
"solution": {
"numbers": [
2,
54,
17
],
"target": 35
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [26, 71, 58, 38], create an equation that equals 40. You can use ... | math | {
"ground_truth": {
"numbers": [
26,
71,
58,
38
],
"target": 40
},
"style": "rule"
} | {
"index": 11,
"solution": {
"numbers": [
26,
71,
58,
38
],
"target": 40
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [21, 39, 34, 36], create an equation that equals 16. You can use ... | math | {
"ground_truth": {
"numbers": [
21,
39,
34,
36
],
"target": 16
},
"style": "rule"
} | {
"index": 12,
"solution": {
"numbers": [
21,
39,
34,
36
],
"target": 16
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [19, 25, 89], create an equation that equals 95. You can use basi... | math | {
"ground_truth": {
"numbers": [
19,
25,
89
],
"target": 95
},
"style": "rule"
} | {
"index": 13,
"solution": {
"numbers": [
19,
25,
89
],
"target": 95
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [8, 62, 43], create an equation that equals 27. You can use basic... | math | {
"ground_truth": {
"numbers": [
8,
62,
43
],
"target": 27
},
"style": "rule"
} | {
"index": 14,
"solution": {
"numbers": [
8,
62,
43
],
"target": 27
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [74, 5, 20, 88], create an equation that equals 50. You can use b... | math | {
"ground_truth": {
"numbers": [
74,
5,
20,
88
],
"target": 50
},
"style": "rule"
} | {
"index": 15,
"solution": {
"numbers": [
74,
5,
20,
88
],
"target": 50
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [14, 45, 9, 1], create an equation that equals 18. You can use ba... | math | {
"ground_truth": {
"numbers": [
14,
45,
9,
1
],
"target": 18
},
"style": "rule"
} | {
"index": 16,
"solution": {
"numbers": [
14,
45,
9,
1
],
"target": 18
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [79, 8, 27, 47], create an equation that equals 91. You can use b... | math | {
"ground_truth": {
"numbers": [
79,
8,
27,
47
],
"target": 91
},
"style": "rule"
} | {
"index": 17,
"solution": {
"numbers": [
79,
8,
27,
47
],
"target": 91
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [21, 14, 56], create an equation that equals 84. You can use basi... | math | {
"ground_truth": {
"numbers": [
21,
14,
56
],
"target": 84
},
"style": "rule"
} | {
"index": 18,
"solution": {
"numbers": [
21,
14,
56
],
"target": 84
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [45, 20, 98], create an equation that equals 73. You can use basi... | math | {
"ground_truth": {
"numbers": [
45,
20,
98
],
"target": 73
},
"style": "rule"
} | {
"index": 19,
"solution": {
"numbers": [
45,
20,
98
],
"target": 73
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [35, 77, 61], create an equation that equals 19. You can use basi... | math | {
"ground_truth": {
"numbers": [
35,
77,
61
],
"target": 19
},
"style": "rule"
} | {
"index": 20,
"solution": {
"numbers": [
35,
77,
61
],
"target": 19
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [96, 39, 28, 32], create an equation that equals 15. You can use ... | math | {
"ground_truth": {
"numbers": [
96,
39,
28,
32
],
"target": 15
},
"style": "rule"
} | {
"index": 21,
"solution": {
"numbers": [
96,
39,
28,
32
],
"target": 15
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [87, 20, 72, 91], create an equation that equals 67. You can use ... | math | {
"ground_truth": {
"numbers": [
87,
20,
72,
91
],
"target": 67
},
"style": "rule"
} | {
"index": 22,
"solution": {
"numbers": [
87,
20,
72,
91
],
"target": 67
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [43, 68, 45], create an equation that equals 20. You can use basi... | math | {
"ground_truth": {
"numbers": [
43,
68,
45
],
"target": 20
},
"style": "rule"
} | {
"index": 23,
"solution": {
"numbers": [
43,
68,
45
],
"target": 20
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [80, 66, 48, 36], create an equation that equals 62. You can use ... | math | {
"ground_truth": {
"numbers": [
80,
66,
48,
36
],
"target": 62
},
"style": "rule"
} | {
"index": 24,
"solution": {
"numbers": [
80,
66,
48,
36
],
"target": 62
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [3, 35, 84], create an equation that equals 63. You can use basic... | math | {
"ground_truth": {
"numbers": [
3,
35,
84
],
"target": 63
},
"style": "rule"
} | {
"index": 25,
"solution": {
"numbers": [
3,
35,
84
],
"target": 63
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [1, 86, 85, 54], create an equation that equals 55. You can use b... | math | {
"ground_truth": {
"numbers": [
1,
86,
85,
54
],
"target": 55
},
"style": "rule"
} | {
"index": 26,
"solution": {
"numbers": [
1,
86,
85,
54
],
"target": 55
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [8, 38, 62, 75], create an equation that equals 91. You can use b... | math | {
"ground_truth": {
"numbers": [
8,
38,
62,
75
],
"target": 91
},
"style": "rule"
} | {
"index": 27,
"solution": {
"numbers": [
8,
38,
62,
75
],
"target": 91
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [4, 50, 92, 11], create an equation that equals 57. You can use b... | math | {
"ground_truth": {
"numbers": [
4,
50,
92,
11
],
"target": 57
},
"style": "rule"
} | {
"index": 28,
"solution": {
"numbers": [
4,
50,
92,
11
],
"target": 57
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [15, 51, 73], create an equation that equals 37. You can use basi... | math | {
"ground_truth": {
"numbers": [
15,
51,
73
],
"target": 37
},
"style": "rule"
} | {
"index": 29,
"solution": {
"numbers": [
15,
51,
73
],
"target": 37
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [85, 45, 69, 74], create an equation that equals 94. You can use ... | math | {
"ground_truth": {
"numbers": [
85,
45,
69,
74
],
"target": 94
},
"style": "rule"
} | {
"index": 30,
"solution": {
"numbers": [
85,
45,
69,
74
],
"target": 94
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [98, 85, 33, 37], create an equation that equals 83. You can use ... | math | {
"ground_truth": {
"numbers": [
98,
85,
33,
37
],
"target": 83
},
"style": "rule"
} | {
"index": 31,
"solution": {
"numbers": [
98,
85,
33,
37
],
"target": 83
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [35, 2, 76, 27], create an equation that equals 72. You can use b... | math | {
"ground_truth": {
"numbers": [
35,
2,
76,
27
],
"target": 72
},
"style": "rule"
} | {
"index": 32,
"solution": {
"numbers": [
35,
2,
76,
27
],
"target": 72
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [28, 35, 77], create an equation that equals 84. You can use basi... | math | {
"ground_truth": {
"numbers": [
28,
35,
77
],
"target": 84
},
"style": "rule"
} | {
"index": 33,
"solution": {
"numbers": [
28,
35,
77
],
"target": 84
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [23, 58, 5, 23], create an equation that equals 63. You can use b... | math | {
"ground_truth": {
"numbers": [
23,
58,
5,
23
],
"target": 63
},
"style": "rule"
} | {
"index": 34,
"solution": {
"numbers": [
23,
58,
5,
23
],
"target": 63
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [51, 61, 91, 48], create an equation that equals 44. You can use ... | math | {
"ground_truth": {
"numbers": [
51,
61,
91,
48
],
"target": 44
},
"style": "rule"
} | {
"index": 35,
"solution": {
"numbers": [
51,
61,
91,
48
],
"target": 44
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [70, 53, 67], create an equation that equals 84. You can use basi... | math | {
"ground_truth": {
"numbers": [
70,
53,
67
],
"target": 84
},
"style": "rule"
} | {
"index": 36,
"solution": {
"numbers": [
70,
53,
67
],
"target": 84
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [14, 9, 34, 22], create an equation that equals 76. You can use b... | math | {
"ground_truth": {
"numbers": [
14,
9,
34,
22
],
"target": 76
},
"style": "rule"
} | {
"index": 37,
"solution": {
"numbers": [
14,
9,
34,
22
],
"target": 76
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [31, 7, 67, 6], create an equation that equals 37. You can use ba... | math | {
"ground_truth": {
"numbers": [
31,
7,
67,
6
],
"target": 37
},
"style": "rule"
} | {
"index": 38,
"solution": {
"numbers": [
31,
7,
67,
6
],
"target": 37
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [78, 41, 37], create an equation that equals 74. You can use basi... | math | {
"ground_truth": {
"numbers": [
78,
41,
37
],
"target": 74
},
"style": "rule"
} | {
"index": 39,
"solution": {
"numbers": [
78,
41,
37
],
"target": 74
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [6, 14, 42], create an equation that equals 42. You can use basic... | math | {
"ground_truth": {
"numbers": [
6,
14,
42
],
"target": 42
},
"style": "rule"
} | {
"index": 40,
"solution": {
"numbers": [
6,
14,
42
],
"target": 42
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [70, 7, 15, 84], create an equation that equals 79. You can use b... | math | {
"ground_truth": {
"numbers": [
70,
7,
15,
84
],
"target": 79
},
"style": "rule"
} | {
"index": 41,
"solution": {
"numbers": [
70,
7,
15,
84
],
"target": 79
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [26, 4, 2], create an equation that equals 44. You can use basic ... | math | {
"ground_truth": {
"numbers": [
26,
4,
2
],
"target": 44
},
"style": "rule"
} | {
"index": 42,
"solution": {
"numbers": [
26,
4,
2
],
"target": 44
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [44, 56, 40, 38], create an equation that equals 78. You can use ... | math | {
"ground_truth": {
"numbers": [
44,
56,
40,
38
],
"target": 78
},
"style": "rule"
} | {
"index": 43,
"solution": {
"numbers": [
44,
56,
40,
38
],
"target": 78
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [39, 58, 22, 55], create an equation that equals 58. You can use ... | math | {
"ground_truth": {
"numbers": [
39,
58,
22,
55
],
"target": 58
},
"style": "rule"
} | {
"index": 44,
"solution": {
"numbers": [
39,
58,
22,
55
],
"target": 58
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [81, 8, 3], create an equation that equals 35. You can use basic ... | math | {
"ground_truth": {
"numbers": [
81,
8,
3
],
"target": 35
},
"style": "rule"
} | {
"index": 45,
"solution": {
"numbers": [
81,
8,
3
],
"target": 35
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [47, 14, 77, 97], create an equation that equals 53. You can use ... | math | {
"ground_truth": {
"numbers": [
47,
14,
77,
97
],
"target": 53
},
"style": "rule"
} | {
"index": 46,
"solution": {
"numbers": [
47,
14,
77,
97
],
"target": 53
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [4, 8, 37], create an equation that equals 39. You can use basic ... | math | {
"ground_truth": {
"numbers": [
4,
8,
37
],
"target": 39
},
"style": "rule"
} | {
"index": 47,
"solution": {
"numbers": [
4,
8,
37
],
"target": 39
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [94, 89, 94], create an equation that equals 89. You can use basi... | math | {
"ground_truth": {
"numbers": [
94,
89,
94
],
"target": 89
},
"style": "rule"
} | {
"index": 48,
"solution": {
"numbers": [
94,
89,
94
],
"target": 89
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [93, 4, 51, 61], create an equation that equals 99. You can use b... | math | {
"ground_truth": {
"numbers": [
93,
4,
51,
61
],
"target": 99
},
"style": "rule"
} | {
"index": 49,
"solution": {
"numbers": [
93,
4,
51,
61
],
"target": 99
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [85, 31, 37, 4], create an equation that equals 76. You can use b... | math | {
"ground_truth": {
"numbers": [
85,
31,
37,
4
],
"target": 76
},
"style": "rule"
} | {
"index": 50,
"solution": {
"numbers": [
85,
31,
37,
4
],
"target": 76
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [79, 54, 4, 18], create an equation that equals 25. You can use b... | math | {
"ground_truth": {
"numbers": [
79,
54,
4,
18
],
"target": 25
},
"style": "rule"
} | {
"index": 51,
"solution": {
"numbers": [
79,
54,
4,
18
],
"target": 25
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [30, 3, 32], create an equation that equals 65. You can use basic... | math | {
"ground_truth": {
"numbers": [
30,
3,
32
],
"target": 65
},
"style": "rule"
} | {
"index": 52,
"solution": {
"numbers": [
30,
3,
32
],
"target": 65
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [94, 14, 70], create an equation that equals 99. You can use basi... | math | {
"ground_truth": {
"numbers": [
94,
14,
70
],
"target": 99
},
"style": "rule"
} | {
"index": 53,
"solution": {
"numbers": [
94,
14,
70
],
"target": 99
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [7, 2, 43, 63], create an equation that equals 97. You can use ba... | math | {
"ground_truth": {
"numbers": [
7,
2,
43,
63
],
"target": 97
},
"style": "rule"
} | {
"index": 54,
"solution": {
"numbers": [
7,
2,
43,
63
],
"target": 97
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [81, 25, 9], create an equation that equals 34. You can use basic... | math | {
"ground_truth": {
"numbers": [
81,
25,
9
],
"target": 34
},
"style": "rule"
} | {
"index": 55,
"solution": {
"numbers": [
81,
25,
9
],
"target": 34
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [65, 31, 40], create an equation that equals 74. You can use basi... | math | {
"ground_truth": {
"numbers": [
65,
31,
40
],
"target": 74
},
"style": "rule"
} | {
"index": 56,
"solution": {
"numbers": [
65,
31,
40
],
"target": 74
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [51, 68, 10, 19], create an equation that equals 46. You can use ... | math | {
"ground_truth": {
"numbers": [
51,
68,
10,
19
],
"target": 46
},
"style": "rule"
} | {
"index": 57,
"solution": {
"numbers": [
51,
68,
10,
19
],
"target": 46
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [70, 70, 91], create an equation that equals 90. You can use basi... | math | {
"ground_truth": {
"numbers": [
70,
70,
91
],
"target": 90
},
"style": "rule"
} | {
"index": 58,
"solution": {
"numbers": [
70,
70,
91
],
"target": 90
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [78, 26, 74, 10], create an equation that equals 94. You can use ... | math | {
"ground_truth": {
"numbers": [
78,
26,
74,
10
],
"target": 94
},
"style": "rule"
} | {
"index": 59,
"solution": {
"numbers": [
78,
26,
74,
10
],
"target": 94
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [98, 88, 25], create an equation that equals 35. You can use basi... | math | {
"ground_truth": {
"numbers": [
98,
88,
25
],
"target": 35
},
"style": "rule"
} | {
"index": 60,
"solution": {
"numbers": [
98,
88,
25
],
"target": 35
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [96, 66, 94], create an equation that equals 33. You can use basi... | math | {
"ground_truth": {
"numbers": [
96,
66,
94
],
"target": 33
},
"style": "rule"
} | {
"index": 61,
"solution": {
"numbers": [
96,
66,
94
],
"target": 33
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [39, 31, 57, 70], create an equation that equals 56. You can use ... | math | {
"ground_truth": {
"numbers": [
39,
31,
57,
70
],
"target": 56
},
"style": "rule"
} | {
"index": 62,
"solution": {
"numbers": [
39,
31,
57,
70
],
"target": 56
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [93, 99, 15], create an equation that equals 90. You can use basi... | math | {
"ground_truth": {
"numbers": [
93,
99,
15
],
"target": 90
},
"style": "rule"
} | {
"index": 63,
"solution": {
"numbers": [
93,
99,
15
],
"target": 90
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [56, 23, 59], create an equation that equals 20. You can use basi... | math | {
"ground_truth": {
"numbers": [
56,
23,
59
],
"target": 20
},
"style": "rule"
} | {
"index": 64,
"solution": {
"numbers": [
56,
23,
59
],
"target": 20
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [6, 6, 87, 45], create an equation that equals 30. You can use ba... | math | {
"ground_truth": {
"numbers": [
6,
6,
87,
45
],
"target": 30
},
"style": "rule"
} | {
"index": 65,
"solution": {
"numbers": [
6,
6,
87,
45
],
"target": 30
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [95, 97, 78], create an equation that equals 39. You can use basi... | math | {
"ground_truth": {
"numbers": [
95,
97,
78
],
"target": 39
},
"style": "rule"
} | {
"index": 66,
"solution": {
"numbers": [
95,
97,
78
],
"target": 39
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [18, 2, 44], create an equation that equals 28. You can use basic... | math | {
"ground_truth": {
"numbers": [
18,
2,
44
],
"target": 28
},
"style": "rule"
} | {
"index": 67,
"solution": {
"numbers": [
18,
2,
44
],
"target": 28
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [86, 7, 33, 35], create an equation that equals 22. You can use b... | math | {
"ground_truth": {
"numbers": [
86,
7,
33,
35
],
"target": 22
},
"style": "rule"
} | {
"index": 68,
"solution": {
"numbers": [
86,
7,
33,
35
],
"target": 22
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [3, 33, 3], create an equation that equals 27. You can use basic ... | math | {
"ground_truth": {
"numbers": [
3,
33,
3
],
"target": 27
},
"style": "rule"
} | {
"index": 69,
"solution": {
"numbers": [
3,
33,
3
],
"target": 27
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [3, 42, 95], create an equation that equals 50. You can use basic... | math | {
"ground_truth": {
"numbers": [
3,
42,
95
],
"target": 50
},
"style": "rule"
} | {
"index": 70,
"solution": {
"numbers": [
3,
42,
95
],
"target": 50
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [17, 4, 55, 3], create an equation that equals 16. You can use ba... | math | {
"ground_truth": {
"numbers": [
17,
4,
55,
3
],
"target": 16
},
"style": "rule"
} | {
"index": 71,
"solution": {
"numbers": [
17,
4,
55,
3
],
"target": 16
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [39, 62, 39, 72], create an equation that equals 56. You can use ... | math | {
"ground_truth": {
"numbers": [
39,
62,
39,
72
],
"target": 56
},
"style": "rule"
} | {
"index": 72,
"solution": {
"numbers": [
39,
62,
39,
72
],
"target": 56
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [71, 13, 45, 68], create an equation that equals 34. You can use ... | math | {
"ground_truth": {
"numbers": [
71,
13,
45,
68
],
"target": 34
},
"style": "rule"
} | {
"index": 73,
"solution": {
"numbers": [
71,
13,
45,
68
],
"target": 34
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [3, 10, 78], create an equation that equals 85. You can use basic... | math | {
"ground_truth": {
"numbers": [
3,
10,
78
],
"target": 85
},
"style": "rule"
} | {
"index": 74,
"solution": {
"numbers": [
3,
10,
78
],
"target": 85
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [50, 24, 12], create an equation that equals 14. You can use basi... | math | {
"ground_truth": {
"numbers": [
50,
24,
12
],
"target": 14
},
"style": "rule"
} | {
"index": 75,
"solution": {
"numbers": [
50,
24,
12
],
"target": 14
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [90, 48, 58], create an equation that equals 16. You can use basi... | math | {
"ground_truth": {
"numbers": [
90,
48,
58
],
"target": 16
},
"style": "rule"
} | {
"index": 76,
"solution": {
"numbers": [
90,
48,
58
],
"target": 16
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [78, 16, 51], create an equation that equals 11. You can use basi... | math | {
"ground_truth": {
"numbers": [
78,
16,
51
],
"target": 11
},
"style": "rule"
} | {
"index": 77,
"solution": {
"numbers": [
78,
16,
51
],
"target": 11
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [48, 25, 95, 23], create an equation that equals 94. You can use ... | math | {
"ground_truth": {
"numbers": [
48,
25,
95,
23
],
"target": 94
},
"style": "rule"
} | {
"index": 78,
"solution": {
"numbers": [
48,
25,
95,
23
],
"target": 94
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [3, 82, 18, 47], create an equation that equals 41. You can use b... | math | {
"ground_truth": {
"numbers": [
3,
82,
18,
47
],
"target": 41
},
"style": "rule"
} | {
"index": 79,
"solution": {
"numbers": [
3,
82,
18,
47
],
"target": 41
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [41, 60, 21], create an equation that equals 80. You can use basi... | math | {
"ground_truth": {
"numbers": [
41,
60,
21
],
"target": 80
},
"style": "rule"
} | {
"index": 80,
"solution": {
"numbers": [
41,
60,
21
],
"target": 80
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [20, 15, 82, 33], create an equation that equals 54. You can use ... | math | {
"ground_truth": {
"numbers": [
20,
15,
82,
33
],
"target": 54
},
"style": "rule"
} | {
"index": 81,
"solution": {
"numbers": [
20,
15,
82,
33
],
"target": 54
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [22, 25, 35, 3], create an equation that equals 26. You can use b... | math | {
"ground_truth": {
"numbers": [
22,
25,
35,
3
],
"target": 26
},
"style": "rule"
} | {
"index": 82,
"solution": {
"numbers": [
22,
25,
35,
3
],
"target": 26
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [73, 27, 9], create an equation that equals 76. You can use basic... | math | {
"ground_truth": {
"numbers": [
73,
27,
9
],
"target": 76
},
"style": "rule"
} | {
"index": 83,
"solution": {
"numbers": [
73,
27,
9
],
"target": 76
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [13, 32, 21, 56], create an equation that equals 25. You can use ... | math | {
"ground_truth": {
"numbers": [
13,
32,
21,
56
],
"target": 25
},
"style": "rule"
} | {
"index": 84,
"solution": {
"numbers": [
13,
32,
21,
56
],
"target": 25
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [10, 47, 3, 50], create an equation that equals 62. You can use b... | math | {
"ground_truth": {
"numbers": [
10,
47,
3,
50
],
"target": 62
},
"style": "rule"
} | {
"index": 85,
"solution": {
"numbers": [
10,
47,
3,
50
],
"target": 62
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [25, 1, 3], create an equation that equals 27. You can use basic ... | math | {
"ground_truth": {
"numbers": [
25,
1,
3
],
"target": 27
},
"style": "rule"
} | {
"index": 86,
"solution": {
"numbers": [
25,
1,
3
],
"target": 27
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [88, 24, 34], create an equation that equals 98. You can use basi... | math | {
"ground_truth": {
"numbers": [
88,
24,
34
],
"target": 98
},
"style": "rule"
} | {
"index": 87,
"solution": {
"numbers": [
88,
24,
34
],
"target": 98
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [99, 78, 22, 42], create an equation that equals 50. You can use ... | math | {
"ground_truth": {
"numbers": [
99,
78,
22,
42
],
"target": 50
},
"style": "rule"
} | {
"index": 88,
"solution": {
"numbers": [
99,
78,
22,
42
],
"target": 50
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [27, 30, 18, 27], create an equation that equals 99. You can use ... | math | {
"ground_truth": {
"numbers": [
27,
30,
18,
27
],
"target": 99
},
"style": "rule"
} | {
"index": 89,
"solution": {
"numbers": [
27,
30,
18,
27
],
"target": 99
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [21, 30, 6, 98], create an equation that equals 82. You can use b... | math | {
"ground_truth": {
"numbers": [
21,
30,
6,
98
],
"target": 82
},
"style": "rule"
} | {
"index": 90,
"solution": {
"numbers": [
21,
30,
6,
98
],
"target": 82
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [65, 50, 70, 9], create an equation that equals 46. You can use b... | math | {
"ground_truth": {
"numbers": [
65,
50,
70,
9
],
"target": 46
},
"style": "rule"
} | {
"index": 91,
"solution": {
"numbers": [
65,
50,
70,
9
],
"target": 46
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [66, 2, 85], create an equation that equals 47. You can use basic... | math | {
"ground_truth": {
"numbers": [
66,
2,
85
],
"target": 47
},
"style": "rule"
} | {
"index": 92,
"solution": {
"numbers": [
66,
2,
85
],
"target": 47
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [97, 31, 25], create an equation that equals 41. You can use basi... | math | {
"ground_truth": {
"numbers": [
97,
31,
25
],
"target": 41
},
"style": "rule"
} | {
"index": 93,
"solution": {
"numbers": [
97,
31,
25
],
"target": 41
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [41, 12, 16], create an equation that equals 69. You can use basi... | math | {
"ground_truth": {
"numbers": [
41,
12,
16
],
"target": 69
},
"style": "rule"
} | {
"index": 94,
"solution": {
"numbers": [
41,
12,
16
],
"target": 69
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [78, 38, 9], create an equation that equals 31. You can use basic... | math | {
"ground_truth": {
"numbers": [
78,
38,
9
],
"target": 31
},
"style": "rule"
} | {
"index": 95,
"solution": {
"numbers": [
78,
38,
9
],
"target": 31
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [11, 76, 29, 99], create an equation that equals 96. You can use ... | math | {
"ground_truth": {
"numbers": [
11,
76,
29,
99
],
"target": 96
},
"style": "rule"
} | {
"index": 96,
"solution": {
"numbers": [
11,
76,
29,
99
],
"target": 96
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [18, 58, 38, 37], create an equation that equals 74. You can use ... | math | {
"ground_truth": {
"numbers": [
18,
58,
38,
37
],
"target": 74
},
"style": "rule"
} | {
"index": 97,
"solution": {
"numbers": [
18,
58,
38,
37
],
"target": 74
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [23, 56, 39, 61], create an equation that equals 87. You can use ... | math | {
"ground_truth": {
"numbers": [
23,
56,
39,
61
],
"target": 87
},
"style": "rule"
} | {
"index": 98,
"solution": {
"numbers": [
23,
56,
39,
61
],
"target": 87
}
} |
DeepScaleR | [
{
"content": "A conversation between User and Assistant. The user asks a question, and the Assistant solves it. The assistant first thinks about the reasoning process in the mind and then provides the user with the answer. User: Using the numbers [45, 61, 37], create an equation that equals 53. You can use basi... | math | {
"ground_truth": {
"numbers": [
45,
61,
37
],
"target": 53
},
"style": "rule"
} | {
"index": 99,
"solution": {
"numbers": [
45,
61,
37
],
"target": 53
}
} |
End of preview. Expand in Data Studio
Countdown Dataset for Verl - DeepSeek-R1 "Aha Moment" Reproduction
This dataset is prepared for reproducing DeepSeek-R1's "aha moment" using the Verl framework. It uses the Countdown task from Jiayi-Pan/Countdown-Tasks-3to4.
Overview
The "aha moment" was first observed in DeepSeek-R1-Zero when training on the Countdown task — the model learns to allocate more thinking time and re-evaluate its initial approach rather than rushing to an answer.
Files
| File | Format | Rows | Size |
|---|---|---|---|
train.parquet |
Parquet | 327,680 | 20 MB |
test.parquet |
Parquet | 1,024 | 72 KB |
train.arrow |
Arrow IPC | 327,680 | 214 MB |
test.arrow |
Arrow IPC | 1,024 | 688 KB |
Schema (Verl RLHFDataset format)
Each row contains:
{
"data_source": "DeepScaleR",
"prompt": [{"role": "user", "content": "..."}],
"ability": "math",
"reward_model": {
"style": "rule",
"ground_truth": {"target": 98, "numbers": [44, 19, 35]}
},
"extra_info": {
"index": 0,
"solution": {"target": 98, "numbers": [44, 19, 35]}
}
}
Usage with Verl
from datasets import load_dataset
# Load from Hugging Face (parquet)
train_ds = load_dataset('sdfffafag3/countdown-verl-aha-moment', data_files='train.parquet', split='train')
# Load from Hugging Face (test)
test_ds = load_dataset('sdfffafag3/countdown-verl-aha-moment', data_files='test.parquet', split='train')
Prompt Template
The prompt uses the base template (works with any base model):
A conversation between User and Assistant... Using the numbers [N1, N2, ...], create an equation that equals T. Show your work in `` tags. Return the final answer in
<answer> </answer>tags.
For Qwen-Instruct models, use the --template_type qwen-instruct flag when running prepare_dataset.py.
Reward Function
The rule-based scorer (reward_score/countdown.py):
- Extracts the equation from
<answer>tags - Validates each number is used exactly once
- Evaluates the arithmetic expression
- Returns 1.0 for correct, 0.1 for format-only, 0 for invalid
Preparation Script
python prepare_dataset.py --save_arrow
Source
Built from Jiayi-Pan/Countdown-Tasks-3to4 (490K countdown problems with 3-4 operands).
- Downloads last month
- 27