hatemestinbejaia's picture
Upload dataset
1de0d84 verified
metadata
dataset_info:
  - config_name: KD_train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA
    features:
      - name: pos_score
        dtype: float64
      - name: neg_score
        dtype: float64
      - name: query
        dtype: string
      - name: pos
        dtype: string
      - name: neg
        dtype: string
      - name: label
        dtype: float64
    splits:
      - name: train0
        num_bytes: 5841404323
        num_examples: 5000000
      - name: train1
        num_bytes: 5841395510
        num_examples: 5000000
      - name: train2
        num_bytes: 5840318955
        num_examples: 5000000
      - name: test
        num_bytes: 11701618
        num_examples: 10000
    download_size: 5981881490
    dataset_size: 17534820406
  - config_name: KDtrain5Mx2tokenizedAraDPR
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 15600000000
        num_examples: 10000000
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1722531359
    dataset_size: 15631200000
  - config_name: KDtrain5Mx2tokenizedAraELECTRA
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 15600000000
        num_examples: 10000000
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1768987166
    dataset_size: 15631200000
  - config_name: KDtrain5Mx2tokenizedmMiniLML12v2
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 15600000000
        num_examples: 10000000
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1949360776
    dataset_size: 15631200000
  - config_name: collection-arabic-embedding-KDaradpr
    features:
      - name: id
        dtype: int32
      - name: embedding
        sequence: float32
    splits:
      - name: collection
        num_bytes: 27232814840
        num_examples: 8841823
    download_size: 32738291533
    dataset_size: 27232814840
  - config_name: collection-arabic-embedding-KDaraelectra
    features:
      - name: id
        dtype: int32
      - name: embedding
        sequence: float32
    splits:
      - name: collection
        num_bytes: 27232814840
        num_examples: 8841823
    download_size: 32738244587
    dataset_size: 27232814840
  - config_name: collection-arabic-embedding-KDmMinilm
    features:
      - name: id
        dtype: int32
      - name: embedding
        sequence: float32
    splits:
      - name: collection
        num_bytes: 13651774712
        num_examples: 8841823
    download_size: 19151445507
    dataset_size: 13651774712
  - config_name: collection-arabic-embedding-NOKDaradpr
    features:
      - name: id
        dtype: int32
      - name: embedding
        sequence: float32
    splits:
      - name: collection
        num_bytes: 27232814840
        num_examples: 8841823
    download_size: 32738586906
    dataset_size: 27232814840
  - config_name: collection-arabic-embedding-NOKDaraelectra
    features:
      - name: id
        dtype: int32
      - name: embedding
        sequence: float32
    splits:
      - name: collection
        num_bytes: 27232814840
        num_examples: 8841823
    download_size: 32738175487
    dataset_size: 27232814840
  - config_name: collection-arabic-embedding-NOKDmMinilm
    features:
      - name: id
        dtype: int32
      - name: embedding
        sequence: float32
    splits:
      - name: collection
        num_bytes: 13651774712
        num_examples: 8841823
    download_size: 19151951299
    dataset_size: 13651774712
  - config_name: collection-arabic-preprocess-for-araelectra
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
    splits:
      - name: collection
        num_bytes: 4664115376
        num_examples: 8841823
    download_size: 2104594788
    dataset_size: 4664115376
  - config_name: collection-arabic-preprocess-for-bert-base-arabertv2
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
    splits:
      - name: collection
        num_bytes: 5237059499
        num_examples: 8841823
    download_size: 2708420681
    dataset_size: 5237059499
  - config_name: hard_train1HardNegative_and_test_tokenizedAraELECTRA
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train1
        num_bytes: 11738691120
        num_examples: 7524802
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1286601979
    dataset_size: 11769891120
  - config_name: mmarco_eval10k_preprossesd_for_AraBERT
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train
        num_bytes: 25758000
        num_examples: 10000
    download_size: 2319539
    dataset_size: 25758000
  - config_name: mmarco_eval10k_preprossesd_for_AraELECTRA
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train
        num_bytes: 13602000
        num_examples: 10000
    download_size: 1786837
    dataset_size: 13602000
  - config_name: mmarco_train10M_preprossesd_for_AraBERT
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train
        num_bytes: 21841398000
        num_examples: 10000000
    download_size: 2197994660
    dataset_size: 21841398000
  - config_name: mmarco_train10M_preprossesd_for_AraELECTRA
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train
        num_bytes: 13831290000
        num_examples: 10000000
    download_size: 1744039517
    dataset_size: 13831290000
  - config_name: queries-arabic-embedding-NOKDaradpr
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
      - name: embedding
        sequence: float32
    splits:
      - name: dev
        num_bytes: 21890112
        num_examples: 6980
    download_size: 26059284
    dataset_size: 21890112
  - config_name: queries-arabic-embedding-NOKDaraelectra
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
      - name: embedding
        sequence: float32
    splits:
      - name: dev
        num_bytes: 21890112
        num_examples: 6980
    download_size: 26059237
    dataset_size: 21890112
  - config_name: queries-arabic-embedding-NOKDmMinilm
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
      - name: embedding
        sequence: float32
    splits:
      - name: dev
        num_bytes: 11168832
        num_examples: 6980
    download_size: 15336592
    dataset_size: 11168832
  - config_name: queries-arabic-embedding-kdaradpr
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
      - name: embedding
        sequence: float32
    splits:
      - name: dev
        num_bytes: 21890112
        num_examples: 6980
    download_size: 26059362
    dataset_size: 21890112
  - config_name: queries-arabic-embedding-kdaraelectra
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
      - name: embedding
        sequence: float32
    splits:
      - name: dev
        num_bytes: 21890112
        num_examples: 6980
    download_size: 26059864
    dataset_size: 21890112
  - config_name: queries-arabic-embedding-kdmMinilm
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
      - name: embedding
        sequence: float32
    splits:
      - name: dev
        num_bytes: 11168832
        num_examples: 6980
    download_size: 15336372
    dataset_size: 11168832
  - config_name: queries-arabic-preprocess-for-araelectra
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 51742703
        num_examples: 808731
      - name: dev.full
        num_bytes: 6511940
        num_examples: 101093
      - name: dev
        num_bytes: 419632
        num_examples: 6980
    download_size: 30345567
    dataset_size: 58674275
  - config_name: queries-arabic-preprocess-for-bert-base-arabertv2
    features:
      - name: id
        dtype: int32
      - name: text
        dtype: string
    splits:
      - name: train
        num_bytes: 56696908
        num_examples: 808731
      - name: dev.full
        num_bytes: 7134521
        num_examples: 101093
      - name: dev
        num_bytes: 458603
        num_examples: 6980
    download_size: 31339968
    dataset_size: 64290032
  - config_name: train0_5Mx2_and_test_tokenized512base_AraBERTv2
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 30960000000
        num_examples: 10000000
      - name: test
        num_bytes: 61920000
        num_examples: 20000
    download_size: 2007459042
    dataset_size: 31021920000
  - config_name: train0_5Mx2_and_test_tokenizedbase_AraBERTv2
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 15600000000
        num_examples: 10000000
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1891858601
    dataset_size: 15631200000
  - config_name: train1HardNegative_and_test_tokenizedAraDPR
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train1
        num_bytes: 1775042880
        num_examples: 1137848
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 220643710
    dataset_size: 1806242880
  - config_name: train1HardNegative_and_test_tokenizedAraELECTRA
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train1
        num_bytes: 1625413920
        num_examples: 1041932
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 203915399
    dataset_size: 1656613920
  - config_name: train5Mx2tokenizedAraDPR
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 15600000000
        num_examples: 10000000
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1630145787
    dataset_size: 15631200000
  - config_name: train5Mx2tokenizedAraELECTRA
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 15600000000
        num_examples: 10000000
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1676601513
    dataset_size: 15631200000
  - config_name: train5Mx2tokenizedmMiniLML12v2
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train0
        num_bytes: 15600000000
        num_examples: 10000000
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 1856975313
    dataset_size: 15631200000
  - config_name: train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA
    features:
      - name: query
        dtype: string
      - name: pos
        dtype: string
      - name: neg
        dtype: string
    splits:
      - name: train0
        num_bytes: 5721404323
        num_examples: 5000000
      - name: train1
        num_bytes: 5721395510
        num_examples: 5000000
      - name: test
        num_bytes: 11461618
        num_examples: 10000
    download_size: 3765291207
    dataset_size: 11454261451
  - config_name: train_dataset_t05M_t15m_eval_10k_preprocessedbase_AraBERTv2
    features:
      - name: query
        dtype: string
      - name: pos
        dtype: string
      - name: neg
        dtype: string
    splits:
      - name: train0
        num_bytes: 6201597264
        num_examples: 5000000
      - name: train1
        num_bytes: 6197574906
        num_examples: 5000000
      - name: test
        num_bytes: 12439665
        num_examples: 10000
    download_size: 4166515408
    dataset_size: 12411611835
  - config_name: very_hard_train1HardNegative_and_test_tokenizedmMiniLML12v2
    features:
      - name: input_ids
        sequence: int32
      - name: token_type_ids
        sequence: int8
      - name: attention_mask
        sequence: int8
      - name: labels
        sequence: float64
    splits:
      - name: train1
        num_bytes: 2783339520
        num_examples: 1784192
      - name: test
        num_bytes: 31200000
        num_examples: 20000
    download_size: 366604431
    dataset_size: 2814539520
configs:
  - config_name: KD_train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA
    data_files:
      - split: train0
        path: KD_train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA/train0-*
      - split: train1
        path: KD_train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA/train1-*
      - split: train2
        path: KD_train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA/train2-*
      - split: test
        path: KD_train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA/test-*
  - config_name: KDtrain5Mx2tokenizedAraDPR
    data_files:
      - split: train0
        path: KDtrain5Mx2tokenizedAraDPR/train0-*
      - split: test
        path: KDtrain5Mx2tokenizedAraDPR/test-*
  - config_name: KDtrain5Mx2tokenizedAraELECTRA
    data_files:
      - split: train0
        path: KDtrain5Mx2tokenizedAraELECTRA/train0-*
      - split: test
        path: KDtrain5Mx2tokenizedAraELECTRA/test-*
  - config_name: KDtrain5Mx2tokenizedmMiniLML12v2
    data_files:
      - split: train0
        path: KDtrain5Mx2tokenizedmMiniLML12v2/train0-*
      - split: test
        path: KDtrain5Mx2tokenizedmMiniLML12v2/test-*
  - config_name: collection-arabic-embedding-KDaradpr
    data_files:
      - split: collection
        path: collection-arabic-embedding-KDaradpr/collection-*
  - config_name: collection-arabic-embedding-KDaraelectra
    data_files:
      - split: collection
        path: collection-arabic-embedding-KDaraelectra/collection-*
  - config_name: collection-arabic-embedding-KDmMinilm
    data_files:
      - split: collection
        path: collection-arabic-embedding-KDmMinilm/collection-*
  - config_name: collection-arabic-embedding-NOKDaradpr
    data_files:
      - split: collection
        path: collection-arabic-embedding-NOKDaradpr/collection-*
  - config_name: collection-arabic-embedding-NOKDaraelectra
    data_files:
      - split: collection
        path: collection-arabic-embedding-NOKDaraelectra/collection-*
  - config_name: collection-arabic-embedding-NOKDmMinilm
    data_files:
      - split: collection
        path: collection-arabic-embedding-NOKDmMinilm/collection-*
  - config_name: collection-arabic-preprocess-for-araelectra
    data_files:
      - split: collection
        path: collection-arabic-preprocess-for-araelectra/collection-*
  - config_name: collection-arabic-preprocess-for-bert-base-arabertv2
    data_files:
      - split: collection
        path: collection-arabic-preprocess-for-bert-base-arabertv2/collection-*
  - config_name: hard_train1HardNegative_and_test_tokenizedAraELECTRA
    data_files:
      - split: train1
        path: hard_train1HardNegative_and_test_tokenizedAraELECTRA/train1-*
      - split: test
        path: hard_train1HardNegative_and_test_tokenizedAraELECTRA/test-*
  - config_name: mmarco_eval10k_preprossesd_for_AraBERT
    data_files:
      - split: train
        path: mmarco_eval10k_preprossesd_for_AraBERT/train-*
  - config_name: mmarco_eval10k_preprossesd_for_AraELECTRA
    data_files:
      - split: train
        path: mmarco_eval10k_preprossesd_for_AraELECTRA/train-*
  - config_name: mmarco_train10M_preprossesd_for_AraBERT
    data_files:
      - split: train
        path: mmarco_train10M_preprossesd_for_AraBERT/train-*
  - config_name: mmarco_train10M_preprossesd_for_AraELECTRA
    data_files:
      - split: train
        path: mmarco_train10M_preprossesd_for_AraELECTRA/train-*
  - config_name: queries-arabic-embedding-NOKDaradpr
    data_files:
      - split: dev
        path: queries-arabic-embedding-NOKDaradpr/dev-*
  - config_name: queries-arabic-embedding-NOKDaraelectra
    data_files:
      - split: dev
        path: queries-arabic-embedding-NOKDaraelectra/dev-*
  - config_name: queries-arabic-embedding-NOKDmMinilm
    data_files:
      - split: dev
        path: queries-arabic-embedding-NOKDmMinilm/dev-*
  - config_name: queries-arabic-embedding-kdaradpr
    data_files:
      - split: dev
        path: queries-arabic-embedding-kdaradpr/dev-*
  - config_name: queries-arabic-embedding-kdaraelectra
    data_files:
      - split: dev
        path: queries-arabic-embedding-kdaraelectra/dev-*
  - config_name: queries-arabic-embedding-kdmMinilm
    data_files:
      - split: dev
        path: queries-arabic-embedding-kdmMinilm/dev-*
  - config_name: queries-arabic-preprocess-for-araelectra
    data_files:
      - split: train
        path: queries-arabic-preprocess-for-araelectra/train-*
      - split: dev.full
        path: queries-arabic-preprocess-for-araelectra/dev.full-*
      - split: dev
        path: queries-arabic-preprocess-for-araelectra/dev-*
  - config_name: queries-arabic-preprocess-for-bert-base-arabertv2
    data_files:
      - split: train
        path: queries-arabic-preprocess-for-bert-base-arabertv2/train-*
      - split: dev.full
        path: queries-arabic-preprocess-for-bert-base-arabertv2/dev.full-*
      - split: dev
        path: queries-arabic-preprocess-for-bert-base-arabertv2/dev-*
  - config_name: train0_5Mx2_and_test_tokenized512base_AraBERTv2
    data_files:
      - split: train0
        path: train0_5Mx2_and_test_tokenized512base_AraBERTv2/train0-*
      - split: test
        path: train0_5Mx2_and_test_tokenized512base_AraBERTv2/test-*
  - config_name: train0_5Mx2_and_test_tokenizedbase_AraBERTv2
    data_files:
      - split: train0
        path: train0_5Mx2_and_test_tokenizedbase_AraBERTv2/train0-*
      - split: test
        path: train0_5Mx2_and_test_tokenizedbase_AraBERTv2/test-*
  - config_name: train1HardNegative_and_test_tokenizedAraDPR
    data_files:
      - split: train1
        path: train1HardNegative_and_test_tokenizedAraDPR/train1-*
      - split: test
        path: train1HardNegative_and_test_tokenizedAraDPR/test-*
  - config_name: train1HardNegative_and_test_tokenizedAraELECTRA
    data_files:
      - split: train1
        path: train1HardNegative_and_test_tokenizedAraELECTRA/train1-*
      - split: test
        path: train1HardNegative_and_test_tokenizedAraELECTRA/test-*
  - config_name: train5Mx2tokenizedAraDPR
    data_files:
      - split: train0
        path: train5Mx2tokenizedAraDPR/train0-*
      - split: test
        path: train5Mx2tokenizedAraDPR/test-*
  - config_name: train5Mx2tokenizedAraELECTRA
    data_files:
      - split: train0
        path: train5Mx2tokenizedAraELECTRA/train0-*
      - split: test
        path: train5Mx2tokenizedAraELECTRA/test-*
  - config_name: train5Mx2tokenizedmMiniLML12v2
    data_files:
      - split: train0
        path: train5Mx2tokenizedmMiniLML12v2/train0-*
      - split: test
        path: train5Mx2tokenizedmMiniLML12v2/test-*
  - config_name: train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA
    data_files:
      - split: train0
        path: train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA/train0-*
      - split: train1
        path: train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA/train1-*
      - split: test
        path: train_dataset_t05M_t15m_eval_10k_preprocessedAraELECTRA/test-*
  - config_name: train_dataset_t05M_t15m_eval_10k_preprocessedbase_AraBERTv2
    data_files:
      - split: train0
        path: train_dataset_t05M_t15m_eval_10k_preprocessedbase_AraBERTv2/train0-*
      - split: train1
        path: train_dataset_t05M_t15m_eval_10k_preprocessedbase_AraBERTv2/train1-*
      - split: test
        path: train_dataset_t05M_t15m_eval_10k_preprocessedbase_AraBERTv2/test-*
  - config_name: very_hard_train1HardNegative_and_test_tokenizedmMiniLML12v2
    data_files:
      - split: train1
        path: very_hard_train1HardNegative_and_test_tokenizedmMiniLML12v2/train1-*
      - split: test
        path: very_hard_train1HardNegative_and_test_tokenizedmMiniLML12v2/test-*