Add Reports Based on "Llemma: An Open Language Model For Mathematics"
#23
by
wlchen
- opened
- contamination_report.csv +7 -0
- postprocessing.py +4 -0
contamination_report.csv
CHANGED
@@ -163,6 +163,9 @@ gigaword;;togethercomputer/RedPajama-Data-V2;;corpus;;;2.82;data-based;https://a
|
|
163 |
|
164 |
gsm8k;;BAAI/Aquila2-34B;;model;;;100.0;model-based;https://huggingface.co/BAAI/Aquila2-34B/blob/main/README.md;21
|
165 |
gsm8k;;BAAI/AquilaChat2-34B;;model;;;100.0;model-based;https://huggingface.co/BAAI/AquilaChat2-34B/blob/main/README.md;21
|
|
|
|
|
|
|
166 |
gsm8k;;GPT-4;;model;100.0;;1.0;data-based;https://arxiv.org/abs/2303.08774;11
|
167 |
gsm8k;;GPT-4;;model;79.00;;;model-based;https://arxiv.org/abs/2311.06233;8
|
168 |
|
@@ -176,6 +179,10 @@ health_fact;;allenai/c4;;corpus;;;7.53;data-based;https://arxiv.org/abs/2310.207
|
|
176 |
health_fact;;oscar-corpus/OSCAR-2301;;corpus;;;3.4;data-based;https://arxiv.org/abs/2310.20707;2
|
177 |
health_fact;;togethercomputer/RedPajama-Data-V2;;corpus;;;18.7;data-based;https://arxiv.org/abs/2310.20707;2
|
178 |
|
|
|
|
|
|
|
|
|
179 |
hlgd;;EleutherAI/pile;;corpus;;;0.0;data-based;https://arxiv.org/abs/2310.20707;2
|
180 |
hlgd;;allenai/c4;;corpus;;;0.0;data-based;https://arxiv.org/abs/2310.20707;2
|
181 |
hlgd;;oscar-corpus/OSCAR-2301;;corpus;;;0.0;data-based;https://arxiv.org/abs/2310.20707;2
|
|
|
163 |
|
164 |
gsm8k;;BAAI/Aquila2-34B;;model;;;100.0;model-based;https://huggingface.co/BAAI/Aquila2-34B/blob/main/README.md;21
|
165 |
gsm8k;;BAAI/AquilaChat2-34B;;model;;;100.0;model-based;https://huggingface.co/BAAI/AquilaChat2-34B/blob/main/README.md;21
|
166 |
+
gsm8k;;EleutherAI/llemma_34b;;model;;;0.15;data-based;https://openreview.net/forum?id=4WnqRR915j;23
|
167 |
+
gsm8k;;EleutherAI/llemma_7b;;model;;;0.15;data-based;https://openreview.net/forum?id=4WnqRR915j;23
|
168 |
+
gsm8k;;EleutherAI/proof-pile-2;;corpus;;;0.15;data-based;https://openreview.net/forum?id=4WnqRR915j;23
|
169 |
gsm8k;;GPT-4;;model;100.0;;1.0;data-based;https://arxiv.org/abs/2303.08774;11
|
170 |
gsm8k;;GPT-4;;model;79.00;;;model-based;https://arxiv.org/abs/2311.06233;8
|
171 |
|
|
|
179 |
health_fact;;oscar-corpus/OSCAR-2301;;corpus;;;3.4;data-based;https://arxiv.org/abs/2310.20707;2
|
180 |
health_fact;;togethercomputer/RedPajama-Data-V2;;corpus;;;18.7;data-based;https://arxiv.org/abs/2310.20707;2
|
181 |
|
182 |
+
hendrycks/competition_math;;EleutherAI/llemma_34b;;model;;;7.72;data-based;https://openreview.net/forum?id=4WnqRR915j;23
|
183 |
+
hendrycks/competition_math;;EleutherAI/llemma_7b;;model;;;7.72;data-based;https://openreview.net/forum?id=4WnqRR915j;23
|
184 |
+
hendrycks/competition_math;;EleutherAI/proof-pile-2;;corpus;;;7.72;data-based;https://openreview.net/forum?id=4WnqRR915j;23
|
185 |
+
|
186 |
hlgd;;EleutherAI/pile;;corpus;;;0.0;data-based;https://arxiv.org/abs/2310.20707;2
|
187 |
hlgd;;allenai/c4;;corpus;;;0.0;data-based;https://arxiv.org/abs/2310.20707;2
|
188 |
hlgd;;oscar-corpus/OSCAR-2301;;corpus;;;0.0;data-based;https://arxiv.org/abs/2310.20707;2
|
postprocessing.py
CHANGED
@@ -17,6 +17,9 @@ def remove_duplicates(data):
|
|
17 |
def fix_arxiv_links(data):
|
18 |
return [[*item[:-2], item[-2].replace("arxiv.org/pdf", "arxiv.org/abs"), item[-1]] for item in data]
|
19 |
|
|
|
|
|
|
|
20 |
def sort_data(data):
|
21 |
return sorted(data, key=lambda x: (x[0], x[1], x[2], x[3], x[-1]))
|
22 |
|
@@ -25,6 +28,7 @@ def main():
|
|
25 |
data = sort_data(data)
|
26 |
data = remove_duplicates(data)
|
27 |
data = fix_arxiv_links(data)
|
|
|
28 |
print("Total datapoints:", len(data))
|
29 |
|
30 |
with open("contamination_report.csv", 'w') as f:
|
|
|
17 |
def fix_arxiv_links(data):
|
18 |
return [[*item[:-2], item[-2].replace("arxiv.org/pdf", "arxiv.org/abs"), item[-1]] for item in data]
|
19 |
|
20 |
+
def fix_openreview_links(data):
|
21 |
+
return [[*item[:-2], item[-2].replace("openreview.net/pdf", "openreview.net/forum"), item[-1]] for item in data]
|
22 |
+
|
23 |
def sort_data(data):
|
24 |
return sorted(data, key=lambda x: (x[0], x[1], x[2], x[3], x[-1]))
|
25 |
|
|
|
28 |
data = sort_data(data)
|
29 |
data = remove_duplicates(data)
|
30 |
data = fix_arxiv_links(data)
|
31 |
+
data = fix_openreview_links(data)
|
32 |
print("Total datapoints:", len(data))
|
33 |
|
34 |
with open("contamination_report.csv", 'w') as f:
|