Add Kaggle math
Browse files
app.py
CHANGED
@@ -57,10 +57,7 @@ def get_leaderboard_df(merge_values: bool = True):
|
|
57 |
elif task.lower() == "agieval":
|
58 |
value = data["results"]["all"]["acc_norm"]
|
59 |
# MATH reports qem
|
60 |
-
elif task.lower()
|
61 |
-
value = data["results"]["all"]["qem"]
|
62 |
-
# MINIMATH reports qem
|
63 |
-
elif task.lower() == "mini_math":
|
64 |
value = data["results"]["all"]["qem"]
|
65 |
else:
|
66 |
first_metric_key = next(
|
|
|
57 |
elif task.lower() == "agieval":
|
58 |
value = data["results"]["all"]["acc_norm"]
|
59 |
# MATH reports qem
|
60 |
+
elif task.lower() in ["math", "mini_math", "aimo_kaggle"]:
|
|
|
|
|
|
|
61 |
value = data["results"]["all"]["qem"]
|
62 |
else:
|
63 |
first_metric_key = next(
|
eval_results/AI-MO/mistral-7b-sft/aimo_v03.00/aimo_kaggle/results_2024-04-22T15-38-51.171312.json
DELETED
@@ -1,89 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"config_general": {
|
3 |
-
"lighteval_sha": "?",
|
4 |
-
"num_fewshot_seeds": 1,
|
5 |
-
"override_batch_size": 4,
|
6 |
-
"max_samples": null,
|
7 |
-
"job_id": "",
|
8 |
-
"start_time": 6210406.872827123,
|
9 |
-
"end_time": 6210648.874373582,
|
10 |
-
"total_evaluation_time_secondes": "242.0015464592725",
|
11 |
-
"model_name": "AI-MO/mistral-7b-sft",
|
12 |
-
"model_sha": "159047b1ab76bbb7c9369ee71bfef1d441fc029e",
|
13 |
-
"model_dtype": "torch.bfloat16",
|
14 |
-
"model_size": "13.99 GB",
|
15 |
-
"config": null
|
16 |
-
},
|
17 |
-
"results": {
|
18 |
-
"custom|aimo_kaggle|0": {
|
19 |
-
"qem": 0.34615384615384615,
|
20 |
-
"qem_stderr": 0.05421594160377287
|
21 |
-
},
|
22 |
-
"all": {
|
23 |
-
"qem": 0.34615384615384615,
|
24 |
-
"qem_stderr": 0.05421594160377287
|
25 |
-
}
|
26 |
-
},
|
27 |
-
"versions": {
|
28 |
-
"custom|aimo_kaggle|0": 0
|
29 |
-
},
|
30 |
-
"config_tasks": {
|
31 |
-
"custom|aimo_kaggle": {
|
32 |
-
"name": "aimo_kaggle",
|
33 |
-
"prompt_function": "kaggle_prompt_fn",
|
34 |
-
"hf_repo": "AI-MO/kaggle-validation-set-v1",
|
35 |
-
"hf_subset": "default",
|
36 |
-
"metric": [
|
37 |
-
"quasi_exact_match_math"
|
38 |
-
],
|
39 |
-
"hf_avail_splits": [
|
40 |
-
"train"
|
41 |
-
],
|
42 |
-
"evaluation_splits": [
|
43 |
-
"train"
|
44 |
-
],
|
45 |
-
"few_shots_split": null,
|
46 |
-
"few_shots_select": null,
|
47 |
-
"generation_size": 2048,
|
48 |
-
"stop_sequence": null,
|
49 |
-
"output_regex": null,
|
50 |
-
"frozen": false,
|
51 |
-
"suite": [
|
52 |
-
"custom"
|
53 |
-
],
|
54 |
-
"original_num_docs": 78,
|
55 |
-
"effective_num_docs": 78,
|
56 |
-
"trust_dataset": null,
|
57 |
-
"must_remove_duplicate_docs": null
|
58 |
-
}
|
59 |
-
},
|
60 |
-
"summary_tasks": {
|
61 |
-
"custom|aimo_kaggle|0": {
|
62 |
-
"hashes": {
|
63 |
-
"hash_examples": "df0b9e4031e0ac93",
|
64 |
-
"hash_full_prompts": "7dc3f63cdfbfe7fe",
|
65 |
-
"hash_input_tokens": "88b37a81252e9f7f",
|
66 |
-
"hash_cont_tokens": "6fa54d6858c4d02d"
|
67 |
-
},
|
68 |
-
"truncated": 78,
|
69 |
-
"non_truncated": 0,
|
70 |
-
"padded": 44,
|
71 |
-
"non_padded": 34,
|
72 |
-
"effective_few_shots": 0.0,
|
73 |
-
"num_truncated_few_shots": 0
|
74 |
-
}
|
75 |
-
},
|
76 |
-
"summary_general": {
|
77 |
-
"hashes": {
|
78 |
-
"hash_examples": "f5b0a912024169d0",
|
79 |
-
"hash_full_prompts": "4532c87f6934c354",
|
80 |
-
"hash_input_tokens": "8e967efb2f3aee9d",
|
81 |
-
"hash_cont_tokens": "8b4f673571e10c5d"
|
82 |
-
},
|
83 |
-
"truncated": 78,
|
84 |
-
"non_truncated": 0,
|
85 |
-
"padded": 44,
|
86 |
-
"non_padded": 34,
|
87 |
-
"num_truncated_few_shots": 0
|
88 |
-
}
|
89 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
eval_results/AI-MO/mistral-7b-sft/aimo_v03.00/mini_math/results_2024-04-23T12-07-13.400858.json
DELETED
@@ -1,291 +0,0 @@
|
|
1 |
-
{
|
2 |
-
"config_general": {
|
3 |
-
"lighteval_sha": "?",
|
4 |
-
"num_fewshot_seeds": 1,
|
5 |
-
"override_batch_size": 4,
|
6 |
-
"max_samples": null,
|
7 |
-
"job_id": "",
|
8 |
-
"start_time": 1691110.31978568,
|
9 |
-
"end_time": 1692074.373145221,
|
10 |
-
"total_evaluation_time_secondes": "964.0533595411107",
|
11 |
-
"model_name": "AI-MO/mistral-7b-sft",
|
12 |
-
"model_sha": "159047b1ab76bbb7c9369ee71bfef1d441fc029e",
|
13 |
-
"model_dtype": "torch.bfloat16",
|
14 |
-
"model_size": "13.99 GB",
|
15 |
-
"config": null
|
16 |
-
},
|
17 |
-
"results": {
|
18 |
-
"custom|mini_math:level_1|0": {
|
19 |
-
"qem": 0.6842105263157895,
|
20 |
-
"qem_stderr": 0.07641750223933595
|
21 |
-
},
|
22 |
-
"custom|mini_math:level_2|0": {
|
23 |
-
"qem": 0.4225352112676056,
|
24 |
-
"qem_stderr": 0.05903984205682581
|
25 |
-
},
|
26 |
-
"custom|mini_math:level_3|0": {
|
27 |
-
"qem": 0.28205128205128205,
|
28 |
-
"qem_stderr": 0.051282051282051246
|
29 |
-
},
|
30 |
-
"custom|mini_math:level_4|0": {
|
31 |
-
"qem": 0.24770642201834864,
|
32 |
-
"qem_stderr": 0.04153846393641014
|
33 |
-
},
|
34 |
-
"custom|mini_math:level_5|0": {
|
35 |
-
"qem": 0.06542056074766354,
|
36 |
-
"qem_stderr": 0.02401664339953487
|
37 |
-
},
|
38 |
-
"custom|mini_math:_average|0": {
|
39 |
-
"qem": 0.34038480048013786,
|
40 |
-
"qem_stderr": 0.0504589005828316
|
41 |
-
},
|
42 |
-
"all": {
|
43 |
-
"qem": 0.34038480048013786,
|
44 |
-
"qem_stderr": 0.0504589005828316
|
45 |
-
}
|
46 |
-
},
|
47 |
-
"versions": {
|
48 |
-
"custom|mini_math:level_1|0": 0,
|
49 |
-
"custom|mini_math:level_2|0": 0,
|
50 |
-
"custom|mini_math:level_3|0": 0,
|
51 |
-
"custom|mini_math:level_4|0": 0,
|
52 |
-
"custom|mini_math:level_5|0": 0
|
53 |
-
},
|
54 |
-
"config_tasks": {
|
55 |
-
"custom|mini_math:level_1": {
|
56 |
-
"name": "mini_math:level_1",
|
57 |
-
"prompt_function": "math",
|
58 |
-
"hf_repo": "HuggingFaceH4/lighteval-mini-math",
|
59 |
-
"hf_subset": "Level 1",
|
60 |
-
"metric": [
|
61 |
-
"quasi_exact_match_math"
|
62 |
-
],
|
63 |
-
"hf_avail_splits": [
|
64 |
-
"train",
|
65 |
-
"test"
|
66 |
-
],
|
67 |
-
"evaluation_splits": [
|
68 |
-
"test"
|
69 |
-
],
|
70 |
-
"few_shots_split": null,
|
71 |
-
"few_shots_select": null,
|
72 |
-
"generation_size": 2048,
|
73 |
-
"stop_sequence": null,
|
74 |
-
"output_regex": null,
|
75 |
-
"frozen": false,
|
76 |
-
"suite": [
|
77 |
-
"custom",
|
78 |
-
"mini_math"
|
79 |
-
],
|
80 |
-
"original_num_docs": 38,
|
81 |
-
"effective_num_docs": 38,
|
82 |
-
"trust_dataset": null,
|
83 |
-
"must_remove_duplicate_docs": null
|
84 |
-
},
|
85 |
-
"custom|mini_math:level_2": {
|
86 |
-
"name": "mini_math:level_2",
|
87 |
-
"prompt_function": "math",
|
88 |
-
"hf_repo": "HuggingFaceH4/lighteval-mini-math",
|
89 |
-
"hf_subset": "Level 2",
|
90 |
-
"metric": [
|
91 |
-
"quasi_exact_match_math"
|
92 |
-
],
|
93 |
-
"hf_avail_splits": [
|
94 |
-
"train",
|
95 |
-
"test"
|
96 |
-
],
|
97 |
-
"evaluation_splits": [
|
98 |
-
"test"
|
99 |
-
],
|
100 |
-
"few_shots_split": null,
|
101 |
-
"few_shots_select": null,
|
102 |
-
"generation_size": 2048,
|
103 |
-
"stop_sequence": null,
|
104 |
-
"output_regex": null,
|
105 |
-
"frozen": false,
|
106 |
-
"suite": [
|
107 |
-
"custom",
|
108 |
-
"mini_math"
|
109 |
-
],
|
110 |
-
"original_num_docs": 71,
|
111 |
-
"effective_num_docs": 71,
|
112 |
-
"trust_dataset": null,
|
113 |
-
"must_remove_duplicate_docs": null
|
114 |
-
},
|
115 |
-
"custom|mini_math:level_3": {
|
116 |
-
"name": "mini_math:level_3",
|
117 |
-
"prompt_function": "math",
|
118 |
-
"hf_repo": "HuggingFaceH4/lighteval-mini-math",
|
119 |
-
"hf_subset": "Level 3",
|
120 |
-
"metric": [
|
121 |
-
"quasi_exact_match_math"
|
122 |
-
],
|
123 |
-
"hf_avail_splits": [
|
124 |
-
"train",
|
125 |
-
"test"
|
126 |
-
],
|
127 |
-
"evaluation_splits": [
|
128 |
-
"test"
|
129 |
-
],
|
130 |
-
"few_shots_split": null,
|
131 |
-
"few_shots_select": null,
|
132 |
-
"generation_size": 2048,
|
133 |
-
"stop_sequence": null,
|
134 |
-
"output_regex": null,
|
135 |
-
"frozen": false,
|
136 |
-
"suite": [
|
137 |
-
"custom",
|
138 |
-
"mini_math"
|
139 |
-
],
|
140 |
-
"original_num_docs": 78,
|
141 |
-
"effective_num_docs": 78,
|
142 |
-
"trust_dataset": null,
|
143 |
-
"must_remove_duplicate_docs": null
|
144 |
-
},
|
145 |
-
"custom|mini_math:level_4": {
|
146 |
-
"name": "mini_math:level_4",
|
147 |
-
"prompt_function": "math",
|
148 |
-
"hf_repo": "HuggingFaceH4/lighteval-mini-math",
|
149 |
-
"hf_subset": "Level 4",
|
150 |
-
"metric": [
|
151 |
-
"quasi_exact_match_math"
|
152 |
-
],
|
153 |
-
"hf_avail_splits": [
|
154 |
-
"train",
|
155 |
-
"test"
|
156 |
-
],
|
157 |
-
"evaluation_splits": [
|
158 |
-
"test"
|
159 |
-
],
|
160 |
-
"few_shots_split": null,
|
161 |
-
"few_shots_select": null,
|
162 |
-
"generation_size": 2048,
|
163 |
-
"stop_sequence": null,
|
164 |
-
"output_regex": null,
|
165 |
-
"frozen": false,
|
166 |
-
"suite": [
|
167 |
-
"custom",
|
168 |
-
"mini_math"
|
169 |
-
],
|
170 |
-
"original_num_docs": 109,
|
171 |
-
"effective_num_docs": 109,
|
172 |
-
"trust_dataset": null,
|
173 |
-
"must_remove_duplicate_docs": null
|
174 |
-
},
|
175 |
-
"custom|mini_math:level_5": {
|
176 |
-
"name": "mini_math:level_5",
|
177 |
-
"prompt_function": "math",
|
178 |
-
"hf_repo": "HuggingFaceH4/lighteval-mini-math",
|
179 |
-
"hf_subset": "Level 5",
|
180 |
-
"metric": [
|
181 |
-
"quasi_exact_match_math"
|
182 |
-
],
|
183 |
-
"hf_avail_splits": [
|
184 |
-
"train",
|
185 |
-
"test"
|
186 |
-
],
|
187 |
-
"evaluation_splits": [
|
188 |
-
"test"
|
189 |
-
],
|
190 |
-
"few_shots_split": null,
|
191 |
-
"few_shots_select": null,
|
192 |
-
"generation_size": 2048,
|
193 |
-
"stop_sequence": null,
|
194 |
-
"output_regex": null,
|
195 |
-
"frozen": false,
|
196 |
-
"suite": [
|
197 |
-
"custom",
|
198 |
-
"mini_math"
|
199 |
-
],
|
200 |
-
"original_num_docs": 107,
|
201 |
-
"effective_num_docs": 107,
|
202 |
-
"trust_dataset": null,
|
203 |
-
"must_remove_duplicate_docs": null
|
204 |
-
}
|
205 |
-
},
|
206 |
-
"summary_tasks": {
|
207 |
-
"custom|mini_math:level_1|0": {
|
208 |
-
"hashes": {
|
209 |
-
"hash_examples": "781a25d9a2f29e87",
|
210 |
-
"hash_full_prompts": "ab476c3f1e6189c8",
|
211 |
-
"hash_input_tokens": "748b0dfb1d890562",
|
212 |
-
"hash_cont_tokens": "719d7c50adc91f8f"
|
213 |
-
},
|
214 |
-
"truncated": 38,
|
215 |
-
"non_truncated": 0,
|
216 |
-
"padded": 23,
|
217 |
-
"non_padded": 15,
|
218 |
-
"effective_few_shots": 0.0,
|
219 |
-
"num_truncated_few_shots": 0
|
220 |
-
},
|
221 |
-
"custom|mini_math:level_2|0": {
|
222 |
-
"hashes": {
|
223 |
-
"hash_examples": "471a4a8568c70994",
|
224 |
-
"hash_full_prompts": "8ec479c7a8a3d59b",
|
225 |
-
"hash_input_tokens": "79bfecc8111dea84",
|
226 |
-
"hash_cont_tokens": "238f82a182eb52a4"
|
227 |
-
},
|
228 |
-
"truncated": 71,
|
229 |
-
"non_truncated": 0,
|
230 |
-
"padded": 35,
|
231 |
-
"non_padded": 36,
|
232 |
-
"effective_few_shots": 0.0,
|
233 |
-
"num_truncated_few_shots": 0
|
234 |
-
},
|
235 |
-
"custom|mini_math:level_3|0": {
|
236 |
-
"hashes": {
|
237 |
-
"hash_examples": "aa6979a90c9a7776",
|
238 |
-
"hash_full_prompts": "0c55bb3e608c2221",
|
239 |
-
"hash_input_tokens": "152332152ec83113",
|
240 |
-
"hash_cont_tokens": "15e129814ba3f8ce"
|
241 |
-
},
|
242 |
-
"truncated": 78,
|
243 |
-
"non_truncated": 0,
|
244 |
-
"padded": 36,
|
245 |
-
"non_padded": 42,
|
246 |
-
"effective_few_shots": 0.0,
|
247 |
-
"num_truncated_few_shots": 0
|
248 |
-
},
|
249 |
-
"custom|mini_math:level_4|0": {
|
250 |
-
"hashes": {
|
251 |
-
"hash_examples": "2e9e3be067de6ec6",
|
252 |
-
"hash_full_prompts": "e471b8649404f9a7",
|
253 |
-
"hash_input_tokens": "b0305a77648ed6a6",
|
254 |
-
"hash_cont_tokens": "0d133d0776517b35"
|
255 |
-
},
|
256 |
-
"truncated": 108,
|
257 |
-
"non_truncated": 1,
|
258 |
-
"padded": 36,
|
259 |
-
"non_padded": 73,
|
260 |
-
"effective_few_shots": 0.0,
|
261 |
-
"num_truncated_few_shots": 0
|
262 |
-
},
|
263 |
-
"custom|mini_math:level_5|0": {
|
264 |
-
"hashes": {
|
265 |
-
"hash_examples": "18b402b510f9a746",
|
266 |
-
"hash_full_prompts": "12e58a2a352d2763",
|
267 |
-
"hash_input_tokens": "5c4b6f568f76445a",
|
268 |
-
"hash_cont_tokens": "c138c90600d234a1"
|
269 |
-
},
|
270 |
-
"truncated": 107,
|
271 |
-
"non_truncated": 0,
|
272 |
-
"padded": 39,
|
273 |
-
"non_padded": 68,
|
274 |
-
"effective_few_shots": 0.0,
|
275 |
-
"num_truncated_few_shots": 0
|
276 |
-
}
|
277 |
-
},
|
278 |
-
"summary_general": {
|
279 |
-
"hashes": {
|
280 |
-
"hash_examples": "470ce01cab0383c9",
|
281 |
-
"hash_full_prompts": "3a272d494f912b3f",
|
282 |
-
"hash_input_tokens": "7d3b36c2ba40899d",
|
283 |
-
"hash_cont_tokens": "017293ac30be3dc4"
|
284 |
-
},
|
285 |
-
"truncated": 402,
|
286 |
-
"non_truncated": 1,
|
287 |
-
"padded": 169,
|
288 |
-
"non_padded": 234,
|
289 |
-
"num_truncated_few_shots": 0
|
290 |
-
}
|
291 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|