Datasets:
Tasks:
Text Generation
Modalities:
Text
Sub-tasks:
language-modeling
Languages:
English
Size:
100K - 1M
License:
Zhangir Azerbayev
commited on
Commit
•
187fe1c
1
Parent(s):
105a9ab
think I'm finally done
Browse files
test.py
CHANGED
@@ -4,12 +4,19 @@ import sys
|
|
4 |
import time
|
5 |
from tqdm import tqdm
|
6 |
|
|
|
7 |
for x in ["arxiv", "books", "formal", "stack-exchange", "wiki", "math-dataset"]:
|
8 |
dataset = load_dataset("./proof-pile.py", x)
|
|
|
|
|
9 |
print(x.upper())
|
10 |
print(dataset)
|
11 |
then = time.time()
|
12 |
for x in tqdm(dataset["train"]):
|
13 |
pass
|
14 |
now = time.time()
|
15 |
-
print(f"
|
|
|
|
|
|
|
|
|
|
4 |
import time
|
5 |
from tqdm import tqdm
|
6 |
|
7 |
+
total_size = 0
|
8 |
for x in ["arxiv", "books", "formal", "stack-exchange", "wiki", "math-dataset"]:
|
9 |
dataset = load_dataset("./proof-pile.py", x)
|
10 |
+
size = dataset["train"].dataset_size / 2**30
|
11 |
+
total_size += size
|
12 |
print(x.upper())
|
13 |
print(dataset)
|
14 |
then = time.time()
|
15 |
for x in tqdm(dataset["train"]):
|
16 |
pass
|
17 |
now = time.time()
|
18 |
+
print(f"{size} GB TRAIN TOTAL")
|
19 |
+
print(f"TRAVERSED IN {now-then} SECONDS")
|
20 |
+
|
21 |
+
print(f"{total_size} GB ACROSS ALL CONFIGS")
|
22 |
+
|