Upload folder using huggingface_hub

README.md

@@ -1,3 +1,33 @@
----
-license: cc-by-sa-4.0
----
+---
+dataset_info:
+  features:
+    - name: text
+      dtype: string
+    - name: word_count
+      dtype: int32
+  splits:
+    - name: train
+      num_examples: 2883231 
+  dataset_size: 4395542393
+configs:
+  - config_name: default
+    data_files:
+      - split: train
+        path: "temp/*.parquet"
+
+license: apache-2.0
+task_categories:
+  - text2text-generation
+  - text-generation
+language:
+  - pt
+pretty_name: 'Subset Corpus Itau-Unibanco/aroeira: 1B tokens (portuguese PT-BR)'
+size_categories:
+  - 1K<n<10K
+---
+
+# **Subset Corpus Itau-Unibanco/aroeira: 1B tokens (portuguese PT-BR)**
+
+Subset Corpus Itau-Unibanco/aroeira: 1B tokens (portuguese PT-BR)
+
+subset-Itau-Unibanco-aroeira-1B-tokens 

aroeira_subset.py

@@ -0,0 +1,125 @@
+import os
+import logging
+import pyarrow as pa
+import pyarrow.parquet as pq
+from datasets import load_dataset
+
+# Configuração de logging
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s - %(levelname)s - %(message)s"
+)
+
+# Threshold de palavras
+WORD_THRESHOLD = 4_000_000_000  # 3 bilhão
+
+
+def count_words(text: str) -> int:
+    """
+    Conta a quantidade de palavras em `text` de forma eficiente,
+    usando contagem de espaços para evitar alocações de lista.
+
+    Args:
+        text (str): Texto a ser analisado.
+
+    Returns:
+        int: Número de palavras.
+    """
+    return text.count(" ") + 1 if text else 0
+
+
+def process_and_write(
+    dataset,
+    output_folder: str,
+    batch_size: int = 100_000
+) -> None:
+    """
+    Itera em um `IterableDataset` streaming registro a registro, agrupa em batches,
+    conta palavras e escreve cada batch em um arquivo Parquet separado.
+    Para evitar gerar arquivos além do limite, interrompe ao alcançar 1 bilhão de palavras.
+
+    Args:
+        dataset: HuggingFace streaming Dataset.
+        output_folder (str): Diretório onde serão salvos os arquivos .parquet.
+        batch_size (int): Número de registros por batch.
+    """
+    # Garante que a pasta exista
+    os.makedirs(output_folder, exist_ok=True)
+
+    batch_records = []
+    batch_idx = 0
+    cumulative_words = 0
+    stop_processing = False
+
+    for record in dataset:
+        batch_records.append(record)
+        # Ao atingir o tamanho do batch, processa e escreve
+        if len(batch_records) >= batch_size:
+            batch_idx += 1
+            texts = [rec["text"] for rec in batch_records]
+            word_counts = [count_words(t) for t in texts]
+            batch_word_sum = sum(word_counts)
+
+            table = pa.Table.from_pydict({
+                "text": texts,
+                "word_count": word_counts,
+            })
+
+            # Define caminho do arquivo para este batch
+            file_path = os.path.join(output_folder, f"batch_{batch_idx}.parquet")
+            pq.write_table(table, file_path)
+            cumulative_words += batch_word_sum
+            logging.info(
+                f"Batch {batch_idx} salvo em '{file_path}': "
+                f"{len(texts)} registros, {batch_word_sum} palavras, total acumulado: {cumulative_words}"
+            )
+
+            # Verifica threshold
+            if cumulative_words >= WORD_THRESHOLD:
+                logging.info(
+                    f"Limite de {WORD_THRESHOLD} palavras atingido. "
+                    f"Parando processamento.")
+                stop_processing = True
+                break
+
+            batch_records = []
+
+    # Processa o lote restante, se houver e se não interrompido
+    if not stop_processing and batch_records:
+        batch_idx += 1
+        texts = [rec["text"] for rec in batch_records]
+        word_counts = [count_words(t) for t in texts]
+        batch_word_sum = sum(word_counts)
+
+        table = pa.Table.from_pydict({
+            "text": texts,
+            "word_count": word_counts,
+        })
+
+        file_path = os.path.join(output_folder, f"batch_{batch_idx}.parquet")
+        pq.write_table(table, file_path)
+        cumulative_words += batch_word_sum
+        logging.info(
+            f"Batch {batch_idx} (restante) salvo em '{file_path}': "
+            f"{len(texts)} registros, {batch_word_sum} palavras, total acumulado: {cumulative_words}"
+        )
+
+    logging.info(f"Processamento finalizado. Total de batches: {batch_idx}, palavras totais: {cumulative_words}")
+
+
+def main() -> None:
+    """
+    Ponto de entrada: carrega o dataset em streaming e chama o processamento.
+    """
+    logging.info("Iniciando processamento e escrita de dados em streaming...")
+    ds = load_dataset("Itau-Unibanco/aroeira", split="train", streaming=True)
+
+    script_dir = os.path.dirname(os.path.abspath(__file__))
+    output_folder = os.path.join(script_dir, "data")
+
+    process_and_write(ds, output_folder)
+    logging.info("Processamento concluído com sucesso.")
+
+
+if __name__ == "__main__":
+    main()

data/batch_1.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e56025063ed37ee9fef37ce755c62aaaa8b88827e5ad9daecde0fb72eb229f2
+size 161970581

data/batch_10.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c52cf3e77dd7ab9fb6253c6c38ce0e75fbe32c56acd84d686d6f972fb5e602c7
+size 150502460

data/batch_100.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ddb800a6662e7af2eeb6dd1ee157838a3397cd8a678c5f25509d0f866dc73d6
+size 146993509

data/batch_101.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:edaaa853bf8f7950b8a479d7e04add6b074243af82e37d260d38208b95950d11
+size 147496878

data/batch_102.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1d0aa3ff9b455cd87304490c5939c2f1418676ad25e4e7e25353d7ce354738f
+size 147426839

data/batch_103.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c3683aa8bd50ec1ecc7b8ee9841f6372fa61bdad1de61f4e2a12e061e57498a
+size 146656083

data/batch_104.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ae45d167464fe72c26d9cea74333300bc34c6784252a6b39b29b8d0910482073
+size 148750118

data/batch_105.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0f542efee6488048ad193c941ca3b5e4a4b212c9441f9842d2947f5190f3579
+size 147225586

data/batch_106.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0497b839b6d9d0acfc556b684acc2fb4c74670b717736211e665ebad873d1aa2
+size 147551571

data/batch_107.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a74a03461dab206fd2e8dde74559db5ba84a46ebac9d1110e0ef6cdeb094579a
+size 147597767

data/batch_108.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62ef2e185a4f475a535bbd3530adb533b4246ef95eebdce3a0324f5bce5e474f
+size 147892353

data/batch_109.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dec40446673ea8eb03ab39efdcae5e272b4f320393f3603370fa1823149ef47
+size 147832782

data/batch_11.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2800fcc0769d9277b7bef0930cac033d2ab787cebff8d54b7228ace9eee2f49
+size 123371165

data/batch_110.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fc759823173614dfb07e0e790c69fe71b565141d53ee66e3743b2c3c3e84971b
+size 147468575

data/batch_12.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1607aa01a63f90a7a20a6fec9aed51c3ea0e00eb6db3757939e24f9add4ef2d
+size 124721389

data/batch_13.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:995270a086eeb498482f3da44d9392a6747189ee98e5078a9186348afdaa40d9
+size 128241633

data/batch_14.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7ab12952c9ce796b635fbaedfdb96e2eeed7043b74aba9bb07c4aebea96baab4
+size 128427021

data/batch_15.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f2c578b5150940ecfc47c0e7bcdf3c71d0a9ac9e72bd24ab256797389d9a0266
+size 130228780

data/batch_16.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dd88ee01a65039531b520804d3aac94f681c1a352f3a1ffe3b9b7e0d95e3eae
+size 130688539

data/batch_17.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b15cdb064f38bdb17165e66aca9080b59e459bdd2d91f90ad965ae42f979ed0e
+size 133420209

data/batch_18.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:89e71bdfd5e82a005a84d78c58788f29b99d4e8db593907cb9029a632002d90b
+size 134931040

data/batch_19.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5694a49939c9241652902d66d47efb5b207842e6bef791c61e23101b8d8cbb1
+size 141729812

data/batch_2.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a4da3f429ca53fae814138522b9cc74e708f75828c680ca85a4776a93b1d6d2
+size 157655555

data/batch_20.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c3eb14bce2c1f8913e1439ab4b158ea4eec21b21abd6bf8ccefb185c7e487302
+size 146977519

data/batch_21.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:468b0745b79dcbff1387381802a13eb240a3115828c8ec9c26649b74d8d71854
+size 147493118

data/batch_22.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8890d610315322d1fe9ba54b0e4dcb9c7c83768f406aaa4636709a4bd04f09a2
+size 153472310

data/batch_23.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:756fdbae2400f437fea5ce30ed86f1d37074866d65a1ede24541604a44bbf33a
+size 153518839

data/batch_24.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0a2b34548d0619e59b93309c8e1d9c5860d1a97da303f701197eecffc9f4241f
+size 152635817

data/batch_25.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee3624c1e373a55979c23ea34922d63bc9e1b84cd4581ec1b569ec2bb91dfddf
+size 150468246

data/batch_26.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:313fd94231e1c7936c117d96df29eb2d062823c81bbdca942f3ca491ae788af5
+size 150374704

data/batch_27.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:754799628d0c75643c8a518170ef204f4c1d8740460e706cf25d7a0f6135386d
+size 151113726

data/batch_28.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:308ecf4fa0c60c4f8613fa9201912b9f1dab823cf2017406f2857f063c2459c4
+size 151046758

data/batch_29.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d8d944abde4227b20a37b0803cda81bf3bcd1001182ea82d6385cce6b53304c
+size 150967441

data/batch_3.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f6fc89d17484a1086cfd50837335ee2d0c046660ec793b750760089a610424f
+size 157925452

data/batch_30.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92dbad584dc4cb943667f55f354b91e2d8aa024b9c96c0ccf6a22a41ead81a57
+size 152464235

data/batch_31.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fa59def5f755dc8f9303fe687ec519a055d756d9ded909845912c0bbd16870dd
+size 152952896

data/batch_32.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee0ac841d949e72c0e2f291272b2dbb31be2857a404d16e58003d0e17c6fcdf0
+size 150829658

data/batch_33.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fac175205fdad8d2c83d1a6599a08cf1adc88c46806dceecfaa4e636e5ff78e7
+size 149736238

data/batch_34.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f39a41016088dee83f8e2d8bd90ad7451d0a997c06e1f55d4400017a344dcec
+size 150080133

data/batch_35.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9135958f16a09af24db8f9460c63e873d9348a232dfc71bb58e059e3089f21ea
+size 150177327

data/batch_36.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff1dffb963e27ad4b6853070b597d233d1860e2f737e17543ea096184eefa3d4
+size 148790412

data/batch_37.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:642234995dd5fff1465c6cf43b0820adb20c777cb22b41630da88d4a8aa1748c
+size 148663076

data/batch_38.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba1174a7d36456cf5cdb26848d5d18b0ea1c94a1d9daeeb33e930520ec680ea8
+size 149445457

data/batch_39.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2482a6b42176f73df58a46cbb2847c058a99ebc8f9872da44dfd42bbf1f603d1
+size 148330416

data/batch_4.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:379d4bbed7cad0e7b2b58158c1ae98301e9525b581a04c3f0bd26b2c67f28bd5
+size 159768482

data/batch_40.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1334fbd7a1ac54a4caaeff92be786ff199c4328cd3ed41c152f5f72575da8acf
+size 149562742

data/batch_41.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f4d1c6c2e2bfb25511d5d2ceba36abeebbcd9b952bfe35f3c2ab3079de24041
+size 149531170

data/batch_42.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:65c7bf70a3ecf57d46326e4dcf68fcad481e37988e1aefc6519b66ec10586418
+size 149769197