Datasets:
Tasks:
Text Generation
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
< 1K
Tags:
shakespeare
License:
Fixed main.py
Browse files- src/main.py +24 -15
src/main.py
CHANGED
@@ -1,22 +1,31 @@
|
|
|
|
1 |
import datasets
|
2 |
-
import glob
|
3 |
|
4 |
-
|
5 |
-
|
|
|
|
|
6 |
|
7 |
-
print(txt_files)
|
8 |
-
|
9 |
-
# Read each file and store its content in a list
|
10 |
texts = []
|
11 |
-
for
|
12 |
-
|
13 |
-
|
|
|
|
|
14 |
|
15 |
-
# Create a Hugging Face dataset from the list of texts
|
16 |
dataset = datasets.Dataset.from_dict({"text": texts})
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
|
18 |
-
|
19 |
-
'train': train_split,
|
20 |
-
'validation': validation_split,
|
21 |
-
'test': test_split
|
22 |
-
})
|
|
|
1 |
+
import os
|
2 |
import datasets
|
|
|
3 |
|
4 |
+
script_path = os.path.abspath(__file__)
|
5 |
+
src_path = os.path.dirname(script_path)
|
6 |
+
root_path = os.path.dirname(src_path)
|
7 |
+
res_path = root_path+"/res"
|
8 |
|
|
|
|
|
|
|
9 |
texts = []
|
10 |
+
for name in os.listdir(res_path):
|
11 |
+
if name.endswith(".txt"):
|
12 |
+
with open(res_path+"/"+name, "r") as file:
|
13 |
+
text = file.read()
|
14 |
+
texts.append(text)
|
15 |
|
|
|
16 |
dataset = datasets.Dataset.from_dict({"text": texts})
|
17 |
+
dataset = dataset.train_test_split(train_size=None, test_size=4, shuffle=False)
|
18 |
+
|
19 |
+
train_val_dataset = dataset["train"]
|
20 |
+
train_val_dataset = train_val_dataset.train_test_split(train_size=None, test_size=2, shuffle=False)
|
21 |
+
|
22 |
+
train_dataset = train_val_dataset["train"]
|
23 |
+
val_dataset = train_val_dataset["test"]
|
24 |
+
test_dataset = dataset["test"]
|
25 |
+
dataset = datasets.DatasetDict({
|
26 |
+
"train": train_dataset,
|
27 |
+
"val": val_dataset,
|
28 |
+
"test": test_dataset
|
29 |
+
})
|
30 |
|
31 |
+
dataset.push_to_hub("shakespearefirstfolio")
|
|
|
|
|
|
|
|