Datasets:

maxzt
/

OpenJA

Tasks:

Text Generation

Text Classification

Modalities:

Formats:

Languages:

Size:

Tags:

continued-pretraining

text-generation

Libraries:

License:

Dataset card Data Studio Files Files and versions

maxzt commited on 5 days ago

Commit

e9f2599

·

verified ·

1 Parent(s): f9294bf

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -25,7 +25,7 @@ size_categories:
 **OpenJA** is a dataset of clean, officially published parliamentary transcripts of the Japanese language. It is characterized by high-quality text (without web noise, HTML, advertising) and reliable metadata, but it represents one narrow language register (official/parliamentary speech) and is better suited as an addition to more diverse corpora than as the only source for a general-purpose pretrain.
 ### 📊 Dataset Statistics
-- **Total Records**: 3,244,877.
 - **Total Size**: ~679,496,535 million tokens (tokenized with `Qwen/Qwen3.5-9B`).
 - **Format**: Parquet.

 **OpenJA** is a dataset of clean, officially published parliamentary transcripts of the Japanese language. It is characterized by high-quality text (without web noise, HTML, advertising) and reliable metadata, but it represents one narrow language register (official/parliamentary speech) and is better suited as an addition to more diverse corpora than as the only source for a general-purpose pretrain.
 ### 📊 Dataset Statistics
+- **Total Records**: 3,244,877.
 - **Total Size**: ~679,496,535 million tokens (tokenized with `Qwen/Qwen3.5-9B`).
 - **Format**: Parquet.