Hugging Face's logo

Datasets:
mlfoundations-dev
/
no_pipeline

Modalities:
Text
Formats:
parquet
Size:
1M - 10M
Libraries:
Datasets
Dask
Dataset card Data Studio Files
xet
 Community
ryanmarten commited on
Commit
34b2c40
·
verified ·
1 Parent(s): 7c90b5d

Upload configuration file

Browse files
Files changed (1) hide show
  1. config/a1_science_stackexchange_biology.yaml +267 -0
config/a1_science_stackexchange_biology.yaml ADDED
@@ -0,0 +1,267 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ operators:
2
+ - id: cs_stack_exchange_0
3
+ config:
4
+ type: function
5
+ function: data_strategies.StackExchange.utils.analyze_post_history
6
+ function_config:
7
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
8
+ split_index: 0
9
+ split_factor: 15
10
+ - id: cs_stack_exchange_1
11
+ config:
12
+ type: function
13
+ function: data_strategies.StackExchange.utils.analyze_post_history
14
+ function_config:
15
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
16
+ split_index: 1
17
+ split_factor: 15
18
+ - id: cs_stack_exchange_2
19
+ config:
20
+ type: function
21
+ function: data_strategies.StackExchange.utils.analyze_post_history
22
+ function_config:
23
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
24
+ split_index: 2
25
+ split_factor: 15
26
+ - id: cs_stack_exchange_3
27
+ config:
28
+ type: function
29
+ function: data_strategies.StackExchange.utils.analyze_post_history
30
+ function_config:
31
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
32
+ split_index: 3
33
+ split_factor: 15
34
+ - id: cs_stack_exchange_4
35
+ config:
36
+ type: function
37
+ function: data_strategies.StackExchange.utils.analyze_post_history
38
+ function_config:
39
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
40
+ split_index: 4
41
+ split_factor: 15
42
+ - id: cs_stack_exchange_5
43
+ config:
44
+ type: function
45
+ function: data_strategies.StackExchange.utils.analyze_post_history
46
+ function_config:
47
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
48
+ split_index: 5
49
+ split_factor: 15
50
+ - id: cs_stack_exchange_6
51
+ config:
52
+ type: function
53
+ function: data_strategies.StackExchange.utils.analyze_post_history
54
+ function_config:
55
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
56
+ split_index: 6
57
+ split_factor: 15
58
+ - id: cs_stack_exchange_7
59
+ config:
60
+ type: function
61
+ function: data_strategies.StackExchange.utils.analyze_post_history
62
+ function_config:
63
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
64
+ split_index: 7
65
+ split_factor: 15
66
+ - id: cs_stack_exchange_8
67
+ config:
68
+ type: function
69
+ function: data_strategies.StackExchange.utils.analyze_post_history
70
+ function_config:
71
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
72
+ split_index: 8
73
+ split_factor: 15
74
+ - id: cs_stack_exchange_9
75
+ config:
76
+ type: function
77
+ function: data_strategies.StackExchange.utils.analyze_post_history
78
+ function_config:
79
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
80
+ split_index: 9
81
+ split_factor: 15
82
+ - id: cs_stack_exchange_10
83
+ config:
84
+ type: function
85
+ function: data_strategies.StackExchange.utils.analyze_post_history
86
+ function_config:
87
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
88
+ split_index: 10
89
+ split_factor: 15
90
+ - id: cs_stack_exchange_11
91
+ config:
92
+ type: function
93
+ function: data_strategies.StackExchange.utils.analyze_post_history
94
+ function_config:
95
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
96
+ split_index: 11
97
+ split_factor: 15
98
+ - id: cs_stack_exchange_12
99
+ config:
100
+ type: function
101
+ function: data_strategies.StackExchange.utils.analyze_post_history
102
+ function_config:
103
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
104
+ split_index: 12
105
+ split_factor: 15
106
+ - id: cs_stack_exchange_13
107
+ config:
108
+ type: function
109
+ function: data_strategies.StackExchange.utils.analyze_post_history
110
+ function_config:
111
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
112
+ split_index: 13
113
+ split_factor: 15
114
+ - id: cs_stack_exchange_14
115
+ config:
116
+ type: function
117
+ function: data_strategies.StackExchange.utils.analyze_post_history
118
+ function_config:
119
+ url: https://archive.org/download/stackexchange/biology.stackexchange.com.7z
120
+ split_index: 14
121
+ split_factor: 15
122
+ - id: uniform_sample
123
+ config:
124
+ type: high_memory_function
125
+ memory: 200
126
+ function: data_strategies.commons.uniform_sample_fixed
127
+ function_config:
128
+ num_samples: 50000
129
+ input_ids:
130
+ - cs_stack_exchange_0
131
+ - cs_stack_exchange_1
132
+ - cs_stack_exchange_2
133
+ - cs_stack_exchange_3
134
+ - cs_stack_exchange_4
135
+ - cs_stack_exchange_5
136
+ - cs_stack_exchange_6
137
+ - cs_stack_exchange_7
138
+ - cs_stack_exchange_8
139
+ - cs_stack_exchange_9
140
+ - cs_stack_exchange_10
141
+ - cs_stack_exchange_11
142
+ - cs_stack_exchange_12
143
+ - cs_stack_exchange_13
144
+ - cs_stack_exchange_14
145
+ - id: rename_task_column
146
+ config:
147
+ type: function
148
+ function: data_strategies.commons.force_rename_columns
149
+ function_config:
150
+ column_maps:
151
+ instruction: instruction_seed
152
+ input_ids:
153
+ - uniform_sample
154
+ - id: sample_dataset_pre
155
+ config:
156
+ type: function
157
+ function: data_strategies.commons.uniform_sample_fixed
158
+ function_config:
159
+ num_samples: 40_000
160
+ input_ids:
161
+ - rename_task_column
162
+ - id: annotate_r1_distill_70b
163
+ config:
164
+ type: completions
165
+ map: deepseek_reasoner
166
+ map_config:
167
+ input_problem_column: instruction_seed
168
+ model: deepseek-reasoner
169
+ batch: false
170
+ temperature: 1.0
171
+ require_all_responses: false
172
+ backend: openai
173
+ backend_params:
174
+ max_requests_per_minute: 500
175
+ max_tokens_per_minute: 1_000_000_000
176
+ base_url: "https://api.deepseek.com/"
177
+ invalid_finish_reasons: ['content_filter']
178
+ input_ids:
179
+ - sample_dataset_pre
180
+ - id: add_source_name
181
+ config:
182
+ type: function
183
+ function: data_strategies.commons.add_constant_columns
184
+ function_config:
185
+ source: "stackexchange-biology"
186
+ input_ids:
187
+ - annotate_r1_distill_70b
188
+ - id: decontaminate
189
+ config:
190
+ type: cpu_function
191
+ sharded: false
192
+ num_cpus: 32
193
+ function: data_strategies._A1_Source.utils.decontaminate_fuzzy_w_ngram
194
+ function_config:
195
+ column: instruction_seed
196
+ eval_datasets:
197
+ - HuggingFaceH4/MATH-500
198
+ - Maxwell-Jia/AIME_2024
199
+ - AI-MO/aimo-validation-amc
200
+ - livecodebench/code_generation_lite
201
+ - mlfoundations-dev/AIME2025_combined
202
+ - cais/hle
203
+ - open-r1/codeforces
204
+ - Idavidrein/gpqa
205
+ - daman1209arora/jeebench
206
+ - mlfoundations-dev/mmlu_pro_eval_full
207
+ - Qwen/CodeElo
208
+ - open-r1/ioi
209
+ eval_columns:
210
+ - problem
211
+ - Problem
212
+ - problem
213
+ - question_content
214
+ - question
215
+ - question
216
+ - description
217
+ - Question
218
+ - question
219
+ - prompt
220
+ - description
221
+ - statement
222
+ eval_splits:
223
+ - test
224
+ - train
225
+ - train
226
+ - test
227
+ - train
228
+ - test
229
+ - test
230
+ - train
231
+ - test
232
+ - test
233
+ - test
234
+ - test
235
+ eval_subsets:
236
+ Idavidrein/gpqa: gpqa_diamond
237
+ similarity_threshold: 75.0
238
+ ngram_size: 13
239
+ input_ids:
240
+ - add_source_name
241
+ - id: sample_dataset
242
+ config:
243
+ type: function
244
+ function: data_strategies.commons.uniform_sample_limited
245
+ function_config:
246
+ num_samples: 31_600 # 1.25% of 10000 target from stage 1
247
+ input_ids: ["decontaminate"]
248
+ - id: convert_reasoning_trace_to_final
249
+ config:
250
+ type: function
251
+ function: data_strategies.commons.convert_reasoning_trace_to_final
252
+ function_config:
253
+ reasoning_column: reasoning
254
+ solution_column: deepseek_solution
255
+ output_column: final_reasoning_trace
256
+ input_ids:
257
+ - sample_dataset
258
+ - id: convert_to_sharegpt
259
+ config:
260
+ type: function
261
+ function: data_strategies.commons.convert_instruction_response_to_sharegpt
262
+ function_config:
263
+ input_instruction_column: instruction_seed
264
+ input_response_column: final_reasoning_trace
265
+ output_sharegpt_column: conversations
266
+ input_ids:
267
+ - convert_reasoning_trace_to_final