Locutusque
commited on
Commit
•
adc60c4
1
Parent(s):
57f52a5
Create TM.py
Browse files
TM.py
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import datasets
|
2 |
+
from datasets import DatasetBuilder, DatasetInfo
|
3 |
+
import pandas as pd
|
4 |
+
|
5 |
+
class TMDataset(datasets.GeneratorBasedBuilder):
|
6 |
+
def __init__(self):
|
7 |
+
self.downloads = [
|
8 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_1.parquet",
|
9 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_10.parquet",
|
10 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_11.parquet",
|
11 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_12.parquet",
|
12 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_13.parquet",
|
13 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_14.parquet",
|
14 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_15.parquet",
|
15 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_16.parquet",
|
16 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_17.parquet",
|
17 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_18.parquet",
|
18 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_2.parquet",
|
19 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_3.parquet",
|
20 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_4.parquet",
|
21 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_5.parquet",
|
22 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_6.parquet",
|
23 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_7.parquet",
|
24 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_8.parquet",
|
25 |
+
"https://huggingface.co/datasets/Locutusque/InstructMix-V2/resolve/main/data/combined_datasets_batch_9.parquet"
|
26 |
+
|
27 |
+
]
|
28 |
+
VERSION = datasets.Version("1.0.0")
|
29 |
+
|
30 |
+
def _info(self):
|
31 |
+
|
32 |
+
features = datasets.Features({
|
33 |
+
"text": datasets.Value("string"),
|
34 |
+
})
|
35 |
+
|
36 |
+
return datasets.DatasetInfo(
|
37 |
+
description="Combination of text completion datasets",
|
38 |
+
features=features,
|
39 |
+
supervised_keys=None
|
40 |
+
)
|
41 |
+
|
42 |
+
def _split_generators(self, dl_manager):
|
43 |
+
|
44 |
+
data_dir = dl_manager.download_and_extract(self.downloads)
|
45 |
+
return [
|
46 |
+
datasets.SplitGenerator(name=datasets.Split.TRAIN, gen_kwargs={"filepath": data_dir})
|
47 |
+
]
|
48 |
+
|
49 |
+
def _generate_examples(self, filepath):
|
50 |
+
for file in filepath:
|
51 |
+
df = pd.read_parquet(file)
|
52 |
+
|
53 |
+
for idx, row in df.iterrows():
|
54 |
+
yield idx, {
|
55 |
+
"text": row["text"],
|
56 |
+
}
|