Hugging Face's logo

Datasets:
AndrewRWilliams
/
time-mmd-DC

Modalities:
Text
Formats:
json
Size:
10K - 100K
Libraries:
Datasets
pandas
Dataset card Data Studio Files
xet
Community
1
time-mmd-DC / README.md
AndrewRWilliams's picture
AndrewRWilliams
Upload folder using huggingface_hub
ba9481d verified
preview code
|
raw
history blame contribute delete
14.5 kB
metadata 
configs:
  - config_name: newyork_aqi_day_predLen_48
    data_files:
      - split: train
        path: data/newyork_aqi_day_predLen_48/train.jsonl
      - split: validation
        path: data/newyork_aqi_day_predLen_48/validation.jsonl
      - split: test
        path: data/newyork_aqi_day_predLen_48/test.jsonl
  - config_name: newyork_aqi_day_predLen_96
    data_files:
      - split: train
        path: data/newyork_aqi_day_predLen_96/train.jsonl
      - split: validation
        path: data/newyork_aqi_day_predLen_96/validation.jsonl
      - split: test
        path: data/newyork_aqi_day_predLen_96/test.jsonl
  - config_name: newyork_aqi_day_predLen_192
    data_files:
      - split: train
        path: data/newyork_aqi_day_predLen_192/train.jsonl
      - split: validation
        path: data/newyork_aqi_day_predLen_192/validation.jsonl
      - split: test
        path: data/newyork_aqi_day_predLen_192/test.jsonl
  - config_name: newyork_aqi_day_predLen_336
    data_files:
      - split: train
        path: data/newyork_aqi_day_predLen_336/train.jsonl
      - split: validation
        path: data/newyork_aqi_day_predLen_336/validation.jsonl
      - split: test
        path: data/newyork_aqi_day_predLen_336/test.jsonl
  - config_name: unadj_unemploymentrate_all_processed_predLen_6
    data_files:
      - split: train
        path: data/unadj_unemploymentrate_all_processed_predLen_6/train.jsonl
      - split: validation
        path: data/unadj_unemploymentrate_all_processed_predLen_6/validation.jsonl
      - split: test
        path: data/unadj_unemploymentrate_all_processed_predLen_6/test.jsonl
  - config_name: unadj_unemploymentrate_all_processed_predLen_8
    data_files:
      - split: train
        path: data/unadj_unemploymentrate_all_processed_predLen_8/train.jsonl
      - split: validation
        path: data/unadj_unemploymentrate_all_processed_predLen_8/validation.jsonl
      - split: test
        path: data/unadj_unemploymentrate_all_processed_predLen_8/test.jsonl
  - config_name: unadj_unemploymentrate_all_processed_predLen_10
    data_files:
      - split: train
        path: data/unadj_unemploymentrate_all_processed_predLen_10/train.jsonl
      - split: validation
        path: data/unadj_unemploymentrate_all_processed_predLen_10/validation.jsonl
      - split: test
        path: data/unadj_unemploymentrate_all_processed_predLen_10/test.jsonl
  - config_name: unadj_unemploymentrate_all_processed_predLen_12
    data_files:
      - split: train
        path: data/unadj_unemploymentrate_all_processed_predLen_12/train.jsonl
      - split: validation
        path: data/unadj_unemploymentrate_all_processed_predLen_12/validation.jsonl
      - split: test
        path: data/unadj_unemploymentrate_all_processed_predLen_12/test.jsonl
  - config_name: us_femagrant_month_predLen_6
    data_files:
      - split: train
        path: data/us_femagrant_month_predLen_6/train.jsonl
      - split: validation
        path: data/us_femagrant_month_predLen_6/validation.jsonl
      - split: test
        path: data/us_femagrant_month_predLen_6/test.jsonl
  - config_name: us_femagrant_month_predLen_8
    data_files:
      - split: train
        path: data/us_femagrant_month_predLen_8/train.jsonl
      - split: validation
        path: data/us_femagrant_month_predLen_8/validation.jsonl
      - split: test
        path: data/us_femagrant_month_predLen_8/test.jsonl
  - config_name: us_femagrant_month_predLen_10
    data_files:
      - split: train
        path: data/us_femagrant_month_predLen_10/train.jsonl
      - split: validation
        path: data/us_femagrant_month_predLen_10/validation.jsonl
      - split: test
        path: data/us_femagrant_month_predLen_10/test.jsonl
  - config_name: us_femagrant_month_predLen_12
    data_files:
      - split: train
        path: data/us_femagrant_month_predLen_12/train.jsonl
      - split: validation
        path: data/us_femagrant_month_predLen_12/validation.jsonl
      - split: test
        path: data/us_femagrant_month_predLen_12/test.jsonl
  - config_name: us_fluratio_week_predLen_12
    data_files:
      - split: train
        path: data/us_fluratio_week_predLen_12/train.jsonl
      - split: validation
        path: data/us_fluratio_week_predLen_12/validation.jsonl
      - split: test
        path: data/us_fluratio_week_predLen_12/test.jsonl
  - config_name: us_fluratio_week_predLen_24
    data_files:
      - split: train
        path: data/us_fluratio_week_predLen_24/train.jsonl
      - split: validation
        path: data/us_fluratio_week_predLen_24/validation.jsonl
      - split: test
        path: data/us_fluratio_week_predLen_24/test.jsonl
  - config_name: us_fluratio_week_predLen_36
    data_files:
      - split: train
        path: data/us_fluratio_week_predLen_36/train.jsonl
      - split: validation
        path: data/us_fluratio_week_predLen_36/validation.jsonl
      - split: test
        path: data/us_fluratio_week_predLen_36/test.jsonl
  - config_name: us_fluratio_week_predLen_48
    data_files:
      - split: train
        path: data/us_fluratio_week_predLen_48/train.jsonl
      - split: validation
        path: data/us_fluratio_week_predLen_48/validation.jsonl
      - split: test
        path: data/us_fluratio_week_predLen_48/test.jsonl
  - config_name: us_gasolineprice_week_predLen_12
    data_files:
      - split: train
        path: data/us_gasolineprice_week_predLen_12/train.jsonl
      - split: validation
        path: data/us_gasolineprice_week_predLen_12/validation.jsonl
      - split: test
        path: data/us_gasolineprice_week_predLen_12/test.jsonl
  - config_name: us_gasolineprice_week_predLen_24
    data_files:
      - split: train
        path: data/us_gasolineprice_week_predLen_24/train.jsonl
      - split: validation
        path: data/us_gasolineprice_week_predLen_24/validation.jsonl
      - split: test
        path: data/us_gasolineprice_week_predLen_24/test.jsonl
  - config_name: us_gasolineprice_week_predLen_36
    data_files:
      - split: train
        path: data/us_gasolineprice_week_predLen_36/train.jsonl
      - split: validation
        path: data/us_gasolineprice_week_predLen_36/validation.jsonl
      - split: test
        path: data/us_gasolineprice_week_predLen_36/test.jsonl
  - config_name: us_gasolineprice_week_predLen_48
    data_files:
      - split: train
        path: data/us_gasolineprice_week_predLen_48/train.jsonl
      - split: validation
        path: data/us_gasolineprice_week_predLen_48/validation.jsonl
      - split: test
        path: data/us_gasolineprice_week_predLen_48/test.jsonl
  - config_name: us_precipitation_month_predLen_6
    data_files:
      - split: train
        path: data/us_precipitation_month_predLen_6/train.jsonl
      - split: validation
        path: data/us_precipitation_month_predLen_6/validation.jsonl
      - split: test
        path: data/us_precipitation_month_predLen_6/test.jsonl
  - config_name: us_precipitation_month_predLen_8
    data_files:
      - split: train
        path: data/us_precipitation_month_predLen_8/train.jsonl
      - split: validation
        path: data/us_precipitation_month_predLen_8/validation.jsonl
      - split: test
        path: data/us_precipitation_month_predLen_8/test.jsonl
  - config_name: us_precipitation_month_predLen_10
    data_files:
      - split: train
        path: data/us_precipitation_month_predLen_10/train.jsonl
      - split: validation
        path: data/us_precipitation_month_predLen_10/validation.jsonl
      - split: test
        path: data/us_precipitation_month_predLen_10/test.jsonl
  - config_name: us_precipitation_month_predLen_12
    data_files:
      - split: train
        path: data/us_precipitation_month_predLen_12/train.jsonl
      - split: validation
        path: data/us_precipitation_month_predLen_12/validation.jsonl
      - split: test
        path: data/us_precipitation_month_predLen_12/test.jsonl
  - config_name: us_retailbroilercomposite_month_predLen_6
    data_files:
      - split: train
        path: data/us_retailbroilercomposite_month_predLen_6/train.jsonl
      - split: validation
        path: data/us_retailbroilercomposite_month_predLen_6/validation.jsonl
      - split: test
        path: data/us_retailbroilercomposite_month_predLen_6/test.jsonl
  - config_name: us_retailbroilercomposite_month_predLen_8
    data_files:
      - split: train
        path: data/us_retailbroilercomposite_month_predLen_8/train.jsonl
      - split: validation
        path: data/us_retailbroilercomposite_month_predLen_8/validation.jsonl
      - split: test
        path: data/us_retailbroilercomposite_month_predLen_8/test.jsonl
  - config_name: us_retailbroilercomposite_month_predLen_10
    data_files:
      - split: train
        path: data/us_retailbroilercomposite_month_predLen_10/train.jsonl
      - split: validation
        path: data/us_retailbroilercomposite_month_predLen_10/validation.jsonl
      - split: test
        path: data/us_retailbroilercomposite_month_predLen_10/test.jsonl
  - config_name: us_retailbroilercomposite_month_predLen_12
    data_files:
      - split: train
        path: data/us_retailbroilercomposite_month_predLen_12/train.jsonl
      - split: validation
        path: data/us_retailbroilercomposite_month_predLen_12/validation.jsonl
      - split: test
        path: data/us_retailbroilercomposite_month_predLen_12/test.jsonl
  - config_name: us_tradebalance_month_predLen_6
    data_files:
      - split: train
        path: data/us_tradebalance_month_predLen_6/train.jsonl
      - split: validation
        path: data/us_tradebalance_month_predLen_6/validation.jsonl
      - split: test
        path: data/us_tradebalance_month_predLen_6/test.jsonl
  - config_name: us_tradebalance_month_predLen_8
    data_files:
      - split: train
        path: data/us_tradebalance_month_predLen_8/train.jsonl
      - split: validation
        path: data/us_tradebalance_month_predLen_8/validation.jsonl
      - split: test
        path: data/us_tradebalance_month_predLen_8/test.jsonl
  - config_name: us_tradebalance_month_predLen_10
    data_files:
      - split: train
        path: data/us_tradebalance_month_predLen_10/train.jsonl
      - split: validation
        path: data/us_tradebalance_month_predLen_10/validation.jsonl
      - split: test
        path: data/us_tradebalance_month_predLen_10/test.jsonl
  - config_name: us_tradebalance_month_predLen_12
    data_files:
      - split: train
        path: data/us_tradebalance_month_predLen_12/train.jsonl
      - split: validation
        path: data/us_tradebalance_month_predLen_12/validation.jsonl
      - split: test
        path: data/us_tradebalance_month_predLen_12/test.jsonl
  - config_name: us_vmt_month_predLen_6
    data_files:
      - split: train
        path: data/us_vmt_month_predLen_6/train.jsonl
      - split: validation
        path: data/us_vmt_month_predLen_6/validation.jsonl
      - split: test
        path: data/us_vmt_month_predLen_6/test.jsonl
  - config_name: us_vmt_month_predLen_8
    data_files:
      - split: train
        path: data/us_vmt_month_predLen_8/train.jsonl
      - split: validation
        path: data/us_vmt_month_predLen_8/validation.jsonl
      - split: test
        path: data/us_vmt_month_predLen_8/test.jsonl
  - config_name: us_vmt_month_predLen_10
    data_files:
      - split: train
        path: data/us_vmt_month_predLen_10/train.jsonl
      - split: validation
        path: data/us_vmt_month_predLen_10/validation.jsonl
      - split: test
        path: data/us_vmt_month_predLen_10/test.jsonl
  - config_name: us_vmt_month_predLen_12
    data_files:
      - split: train
        path: data/us_vmt_month_predLen_12/train.jsonl
      - split: validation
        path: data/us_vmt_month_predLen_12/validation.jsonl
      - split: test
        path: data/us_vmt_month_predLen_12/test.jsonl

Time-MMD Multi-Config

Updated September 2025. This dataset repository packages multiple long-horizon forecasting configs for the Time-MMD evaluation suite.

Manual Configuration

This repository follows Hugging Face's manual configuration layout. Each config name maps to the files under data/<config_id>/. The canonical list of configs lives in configs.json, which the Hub reads to expose selectable config_name options.

Available Configs

The following configs are available (see configs.json for the source of truth):

  • newyork_aqi_day_predLen_48

  • newyork_aqi_day_predLen_96

  • newyork_aqi_day_predLen_192

  • newyork_aqi_day_predLen_336

  • unadj_unemploymentrate_all_processed_predLen_6

  • unadj_unemploymentrate_all_processed_predLen_8

  • unadj_unemploymentrate_all_processed_predLen_10

  • unadj_unemploymentrate_all_processed_predLen_12

  • us_femagrant_month_predLen_6

  • us_femagrant_month_predLen_8

  • us_femagrant_month_predLen_10

  • us_femagrant_month_predLen_12

  • us_fluratio_week_predLen_12

  • us_fluratio_week_predLen_24

  • us_fluratio_week_predLen_36

  • us_fluratio_week_predLen_48

  • us_gasolineprice_week_predLen_12

  • us_gasolineprice_week_predLen_24

  • us_gasolineprice_week_predLen_36

  • us_gasolineprice_week_predLen_48

  • us_precipitation_month_predLen_6

  • us_precipitation_month_predLen_8

  • us_precipitation_month_predLen_10

  • us_precipitation_month_predLen_12

  • us_retailbroilercomposite_month_predLen_6

  • us_retailbroilercomposite_month_predLen_8

  • us_retailbroilercomposite_month_predLen_10

  • us_retailbroilercomposite_month_predLen_12

  • us_tradebalance_month_predLen_6

  • us_tradebalance_month_predLen_8

  • us_tradebalance_month_predLen_10

  • us_tradebalance_month_predLen_12

  • us_vmt_month_predLen_6

  • us_vmt_month_predLen_8

  • us_vmt_month_predLen_10

  • us_vmt_month_predLen_12

Each entry corresponds to a directory under data/ containing three splits: train.jsonl, validation.jsonl, and test.jsonl.

Record Schema

Every JSONL record is a single evaluation window with these keys:

  • batch_x: observed history values.
  • batch_y: prediction horizon targets.
  • batch_text: natural-language prompt context.
  • batch_x_mark / batch_y_mark: positional encodings.
  • batch_x_timestamps_datetime64_ns / batch_y_timestamps_datetime64_ns: absolute timestamps in nanoseconds since the Unix epoch.

Using From datasets 

from datasets import load_dataset
ds = load_dataset("AndrewRWilliams/time-mmd-DC", config_name="newyork_aqi_day_predLen_192", split="train")