metadata

language:
  - en
license: apache-2.0
source_datasets: vivym/midjourney-messages
task_categories:
  - text-generation
dataset_info:
  - config_name: deduped
    features:
      - name: id
        dtype: string
      - name: channel_id
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 2538669745
        num_examples: 14828769
    download_size: 1585207687
    dataset_size: 2538669745
  - config_name: default
    features:
      - name: id
        dtype: string
      - name: channel_id
        dtype: string
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 3575844717.3610477
        num_examples: 19716685
    download_size: 1514418407
    dataset_size: 3575844717.3610477
configs:
  - config_name: deduped
    data_files:
      - split: train
        path: deduped/train-*
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
tags:
  - midjourney

midjourney-messages-cleaned

This is vivym/midjourney-messages but with the following cleaning steps:

remove most columns (keep id columns for reference vs. original)
Apply clean-text to all rows (keep casing)
rename content to text (ffs)
remove intermediate ID/tag (???) in angle brackets at the end, remove double asterisks **
remove exact duplicate rows

dataset structure

overall:

DatasetDict({
    train: Dataset({
        features: ['id', 'channel_id', 'text'],
        num_rows: 19738964
    })
})

A single example looks like this:

random.choice(dataset['train'])

{'id': '1108635049391308879',
 'channel_id': '1008571088919343124',
 'text': 'Warhammer 40k Chaos Space Marine with pink Armor and a guitar'}

details

585M GPT-4 tiktoken tokens.

token_count
count  1.971668e+07
mean   2.971651e+01
std    3.875208e+01
min    1.000000e+00
25%    1.000000e+01
50%    1.900000e+01
75%    3.400000e+01
max    2.077000e+03