Hugging Face's logo

Datasets:
mlfoundations-dev
/
math_ablate_duplications_1

Modalities:
Text
Formats:
parquet
Size:
< 1K
Libraries:
Datasets
pandas
Dataset card Data Studio Files
xet
 Community
EtashGuha commited on
Commit
f853d39
·
verified ·
1 Parent(s): a10714e

Upload configuration file

Browse files
Files changed (1) hide show
  1. config/base_math_ablate_duplications.yaml +248 -0
config/base_math_ablate_duplications.yaml ADDED
@@ -0,0 +1,248 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ operators:
2
+ - id: seed_math_automath_1000000
3
+ config:
4
+ type: load_preexisting
5
+ framework_name: seed_math_automath_1000000
6
+ - id: drop_columns_automath_1000000
7
+ config:
8
+ type: function
9
+ function: data_strategies.commons.keep_only_columns
10
+ function_config:
11
+ columns_to_keep:
12
+ - instruction_seed
13
+ input_ids:
14
+ - seed_math_automath_1000000
15
+ - id: add_source_automath
16
+ config:
17
+ type: function
18
+ function: data_strategies.commons.add_constant_columns
19
+ function_config:
20
+ source: "automath"
21
+ input_ids:
22
+ - drop_columns_automath_1000000
23
+ - id: filter_out_proofs_automath_1000000
24
+ config:
25
+ type: function
26
+ function: data_strategies.MultipleSamples.utils.filter_out_proofs
27
+ function_config:
28
+ problem_column: instruction_seed
29
+ input_ids:
30
+ - add_source_automath
31
+ - id: sample_dataset_automath
32
+ config:
33
+ type: function
34
+ function: data_strategies.commons.uniform_sample_fixed
35
+ function_config:
36
+ num_samples: 4_000
37
+ input_ids:
38
+ - filter_out_proofs_automath_1000000
39
+
40
+ - id: seed_math_nvidia_math_exploit
41
+ config:
42
+ type: load_preexisting
43
+ framework_name: seed_math_nvidia_math_exploit
44
+ - id: drop_columns_nvidia_math_exploit
45
+ config:
46
+ type: function
47
+ function: data_strategies.commons.keep_only_columns
48
+ function_config:
49
+ columns_to_keep:
50
+ - instruction_seed
51
+ input_ids:
52
+ - seed_math_nvidia_math_exploit
53
+ - id: add_source_nvidia_math
54
+ config:
55
+ type: function
56
+ function: data_strategies.commons.add_constant_columns
57
+ function_config:
58
+ source: "nvidia_math"
59
+ input_ids:
60
+ - drop_columns_nvidia_math_exploit
61
+ - id: filter_out_proofs_nvidia_math
62
+ config:
63
+ type: function
64
+ function: data_strategies.MultipleSamples.utils.filter_out_proofs
65
+ function_config:
66
+ problem_column: instruction_seed
67
+ input_ids:
68
+ - add_source_nvidia_math
69
+ - id: sample_dataset_nvidia_math
70
+ config:
71
+ type: function
72
+ function: data_strategies.commons.uniform_sample_fixed
73
+ function_config:
74
+ num_samples: 4_000
75
+ input_ids:
76
+ - filter_out_proofs_nvidia_math
77
+
78
+
79
+ - id: seed_math_tiger_lab_math_exploit
80
+ config:
81
+ type: load_preexisting
82
+ framework_name: seed_math_tiger_lab_math_exploit
83
+ - id: drop_columns_tiger_lab_math_exploit
84
+ config:
85
+ type: function
86
+ function: data_strategies.commons.keep_only_columns
87
+ function_config:
88
+ columns_to_keep:
89
+ - instruction_seed
90
+ input_ids:
91
+ - seed_math_tiger_lab_math_exploit
92
+ - id: add_source_tiger_lab_math
93
+ config:
94
+ type: function
95
+ function: data_strategies.commons.add_constant_columns
96
+ function_config:
97
+ source: "tiger_lab_math"
98
+ input_ids:
99
+ - drop_columns_tiger_lab_math_exploit
100
+ - id: filter_out_proofs_tiger_lab_math
101
+ config:
102
+ type: function
103
+ function: data_strategies.MultipleSamples.utils.filter_out_proofs
104
+ function_config:
105
+ problem_column: instruction_seed
106
+ input_ids:
107
+ - add_source_tiger_lab_math
108
+ - id: sample_dataset_tiger_lab_math
109
+ config:
110
+ type: function
111
+ function: data_strategies.commons.uniform_sample_fixed
112
+ function_config:
113
+ num_samples: 4_000
114
+ input_ids:
115
+ - filter_out_proofs_tiger_lab_math
116
+
117
+ - id: seed_math_tiger_math_exploit
118
+ config:
119
+ type: load_preexisting
120
+ framework_name: seed_math_tiger_math_exploit
121
+ - id: drop_columns_tiger_math_exploit
122
+ config:
123
+ type: function
124
+ function: data_strategies.commons.keep_only_columns
125
+ function_config:
126
+ columns_to_keep:
127
+ - instruction_seed
128
+ input_ids:
129
+ - seed_math_tiger_math_exploit
130
+ - id: add_source_tiger_math
131
+ config:
132
+ type: function
133
+ function: data_strategies.commons.add_constant_columns
134
+ function_config:
135
+ source: "tiger_math"
136
+ input_ids:
137
+ - drop_columns_tiger_math_exploit
138
+ - id: filter_out_proofs_tiger_math
139
+ config:
140
+ type: function
141
+ function: data_strategies.MultipleSamples.utils.filter_out_proofs
142
+ function_config:
143
+ problem_column: instruction_seed
144
+ input_ids:
145
+ - add_source_tiger_math
146
+ - id: sample_dataset_tiger_math
147
+ config:
148
+ type: function
149
+ function: data_strategies.commons.uniform_sample_fixed
150
+ function_config:
151
+ num_samples: 4_000
152
+ input_ids:
153
+ - filter_out_proofs_tiger_math
154
+ - id: sample_dataset_all
155
+ config:
156
+ type: function
157
+ function: data_strategies.commons.uniform_sample_fixed
158
+ function_config:
159
+ num_samples: 16_000
160
+ input_ids:
161
+ - sample_dataset_tiger_math
162
+ - sample_dataset_automath
163
+ - sample_dataset_nvidia_math
164
+ - sample_dataset_tiger_lab_math
165
+
166
+ - id: rename_instruction
167
+ config:
168
+ type: function
169
+ function: data_strategies.commons.force_rename_columns
170
+ function_config:
171
+ column_maps:
172
+ instruction_seed: problem
173
+ input_ids:
174
+ - sample_dataset_all
175
+
176
+ - id: dedup_topics
177
+ config:
178
+ type: function
179
+ function: data_strategies.commons.dedup_on_columns
180
+ function_config:
181
+ dedup_columns:
182
+ - problem
183
+ input_ids:
184
+ - rename_instruction
185
+
186
+ - id: sample_dataset
187
+ config:
188
+ type: function
189
+ function: data_strategies.commons.uniform_sample_fixed
190
+ function_config:
191
+ num_samples: 1
192
+ input_ids:
193
+ - dedup_topics
194
+ - id: add_row_ids
195
+ config:
196
+ type: function
197
+ function: data_strategies.commons.add_row_ids
198
+ function_config:
199
+ column_name: id
200
+ input_ids:
201
+ - sample_dataset
202
+ - id: duplicate
203
+ config:
204
+ type: function
205
+ function: data_strategies.TestTime.generator.duplicate_rows
206
+ function_config:
207
+ n_copies: 12
208
+ input_ids:
209
+ - add_row_ids
210
+ - id: annotate_r1_distill_70b
211
+ config:
212
+ type: completions
213
+ map: deepseek_reasoner
214
+ map_config:
215
+ input_problem_column: problem
216
+ model: deepseek-reasoner
217
+ batch: False
218
+ temperature: 1.0
219
+ require_all_responses: false
220
+ backend: openai
221
+ backend_params:
222
+ max_requests_per_minute: 2_500
223
+ max_tokens_per_minute: 1_000_000_000
224
+ api_key: sk-66c9461d203144cc8bda1d0e71e08799
225
+ base_url: "https://api.deepseek.com/"
226
+ invalid_finish_reasons: ['content_filter']
227
+ input_ids:
228
+ - duplicate
229
+ - id: convert_reasoning_trace_to_final
230
+ config:
231
+ type: function
232
+ function: data_strategies.commons.convert_reasoning_trace_to_final
233
+ function_config:
234
+ reasoning_column: reasoning
235
+ solution_column: deepseek_solution
236
+ output_column: final_reasoning_trace
237
+ input_ids:
238
+ - annotate_r1_distill_70b
239
+ - id: merge_together
240
+ config:
241
+ type: function
242
+ function: data_strategies.TestTime.generator.merge_duplicate_rows
243
+ function_config:
244
+ diff_columns:
245
+ - final_reasoning_trace
246
+ - id
247
+ input_ids:
248
+ - convert_reasoning_trace_to_final