metadata

language:
  - en
task_categories:
  - summarization

AutoTrain Dataset for project: test-summarization

Dataset Description

This dataset has been automatically processed by AutoTrain for project test-summarization.

Languages

The BCP-47 code for the dataset's language is en.

Dataset Structure

Data Instances

A sample from this dataset looks as follows:

[
  {
    "feat_id": "train_1087",
    "text": "#Person1#: Hello sir, how can I help you?\n#Person2#: Yes, I need this prescription please.\n#Person1#: Let's see. Okay, so 50 mg of Prozac, would you prefer this in capsule or tablet?\n#Person2#: Capsules are fine.\n#Person1#: Okay, you should take 1 capsule 3 times a day. Be sure not to take it on an empty stomach, and also, don't ever mix it with alcohol!\n#Person2#: Yes, I know. It's not the first time I'm taking this! Don't worry, I won't overdose!\n#Person1#: Okay, anything else I can get you?\n#Person2#: Oh, yes, I almost forgot! Can I also get some eye drops and um, some condoms?\n#Person1#: Sure. Darn condoms aren't registered in our system.\n#Person2#: Oh, well that's okay, I'll get some later, thanks. . . Really it's no problem.\n#Person1#: Just hang on there a sec. Can I get a price check on ' Fun Times Ribbed Condoms ' please!",
    "target": "#Person1# gives #Person2# #Person2#'s prescription. #Person2# also wants some eye drops and some condoms but is told that darn condoms are registered in their system.",
    "feat_topic": "the pharmacy"
  },
  {
    "feat_id": "train_1193",
    "text": "#Person1#: Don't be too sad. If you really think that you have no feeling with him, then, in my opinion, getting divorced maybe is the best way to solve the problem. \n#Person2#: I know clearly at the bottom of my heart. I just can't set my mind at rest because of the child. She's little. She cannot understand us and accept such truth. \n#Person1#: Yeah, child is the matter. Don't tell Jenny the truth, only tell her the white lie. When she grows up, you find the suitable opportunity to tell her. \n#Person2#: I see. OK. ",
    "target": "#Person1# suggests #Person2# get divorced if #Person2# has no feeling with a man and tell their daughter the white lie.",
    "feat_topic": "divorce"
  }
]

Dataset Fields

The dataset has the following fields (also called "features"):

{
  "feat_id": "Value(dtype='string', id=None)",
  "text": "Value(dtype='string', id=None)",
  "target": "Value(dtype='string', id=None)",
  "feat_topic": "Value(dtype='string', id=None)"
}

Dataset Splits

This dataset is split into a train and validation split. The split sizes are as follow:

Split name	Num samples
train	1599
valid	400