Sourab Mangrulkar
commited on
Commit
•
8601501
1
Parent(s):
f52011b
updating code
Browse files- chat-t5-remove.py +0 -81
chat-t5-remove.py
DELETED
@@ -1,81 +0,0 @@
|
|
1 |
-
import os
|
2 |
-
import sys
|
3 |
-
import pandas as pd
|
4 |
-
import numpy as np
|
5 |
-
import datasets
|
6 |
-
import pathlib
|
7 |
-
|
8 |
-
_DESCRIPTION = """\
|
9 |
-
This script is used to generate a dataset for the chatbot task. The dataset is combination of datasets used to train BlenderBot."""
|
10 |
-
|
11 |
-
_VERSION = "1.0.0"
|
12 |
-
|
13 |
-
|
14 |
-
class ChatT5Config(datasets.BuilderConfig):
|
15 |
-
"""BuilderConfig for TapacoConfig."""
|
16 |
-
|
17 |
-
def __init__(self, **kwargs):
|
18 |
-
super(ChatT5Config, self).__init__(version=datasets.Version(_VERSION, ""), **kwargs),
|
19 |
-
|
20 |
-
|
21 |
-
class ChatT5(datasets.GeneratorBasedBuilder):
|
22 |
-
|
23 |
-
BUILDER_CONFIGS = [
|
24 |
-
ChatT5Config(
|
25 |
-
name="chat_t5",
|
26 |
-
description="The dataset is combination of datasets used to train BlenderBot.",
|
27 |
-
)
|
28 |
-
]
|
29 |
-
BUILDER_CONFIG_CLASS = ChatT5Config
|
30 |
-
DEFAULT_CONFIG_NAME = "chat_t5"
|
31 |
-
|
32 |
-
def _info(self):
|
33 |
-
features = datasets.Features({"texts": datasets.Value("string"), "labels": datasets.Value("string")})
|
34 |
-
|
35 |
-
return datasets.DatasetInfo(description=_DESCRIPTION, features=features, supervised_keys=None)
|
36 |
-
|
37 |
-
def _split_generators(self, dl_manager):
|
38 |
-
"""Returns SplitGenerators."""
|
39 |
-
train_files = [
|
40 |
-
"empathetic_dialogues_train.parquet",
|
41 |
-
"blended_skill_talk_train.parquet",
|
42 |
-
"wizard_of_wikipedia_train.parquet",
|
43 |
-
"conv_ai_2_train.parquet",
|
44 |
-
]
|
45 |
-
validation_files = [
|
46 |
-
"conv_ai_2_validation.parquet",
|
47 |
-
"empathetic_dialogues_validation.parquet",
|
48 |
-
"blended_skill_talk_validation.parquet",
|
49 |
-
"wizard_of_wikipedia_validation.parquet",
|
50 |
-
]
|
51 |
-
test_files = [
|
52 |
-
"blended_skill_talk_test.parquet",
|
53 |
-
"empathetic_dialogues_test.parquet",
|
54 |
-
"wizard_of_wikipedia_test.parquet",
|
55 |
-
]
|
56 |
-
urls_to_download = "./data/train/conv_ai_2_train.parquet"
|
57 |
-
downloaded_files = dl_manager.download_and_extract(urls_to_download)
|
58 |
-
return [
|
59 |
-
datasets.SplitGenerator(
|
60 |
-
name=datasets.Split.TRAIN,
|
61 |
-
gen_kwargs={"data_dir": downloaded_files["train"]},
|
62 |
-
),
|
63 |
-
datasets.SplitGenerator(
|
64 |
-
name=datasets.Split.TEST,
|
65 |
-
gen_kwargs={"data_dir": downloaded_files["test"]},
|
66 |
-
),
|
67 |
-
datasets.SplitGenerator(
|
68 |
-
name=datasets.Split.VALIDATION,
|
69 |
-
gen_kwargs={"data_dir": downloaded_files["validation"]},
|
70 |
-
),
|
71 |
-
]
|
72 |
-
|
73 |
-
def _generate_examples(self, data_dir):
|
74 |
-
"""Yields examples."""
|
75 |
-
data = pd.DataFrame()
|
76 |
-
for file_path in os.listdir(data_dir):
|
77 |
-
data = pd.concat([data, pd.read_parquet(os.path.join(data_dir, file_path))])
|
78 |
-
id_ = -1
|
79 |
-
for _, row in data.iterrows():
|
80 |
-
id_ += 1
|
81 |
-
yield id_, {"texts": row["texts"], "labels": row["labels"]}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|