--- task_categories: - text-generation language: - en - de - fr - es - it pretty_name: Red Pajama V2 Dataset size_categories: - 100M bool: """ function returns True if the sample complies with Gopher rules """ signals = json.loads(sample["quality_signals"]) # rule 1: number of words between 50 and 10'000 word_count = signals["rps_doc_word_count"][0][2] if word_count < 50 or word_count > 10_000: return False # rule 2: mean word length between 3 and 10 mean_word_length = signals["rps_doc_mean_word_length"][0][2] if mean_word_length < 3 or mean_word_length > 10: return False # rule 2: symbol to word ratio below 0.1 symbol_word_ratio = signals["rps_doc_symbol_to_word_ratio"][0][2] if symbol_word_ratio > 0.1: return False # rule 3: 90% of lines need to start without a bullet point n_lines = signals["ccnet_nlines"][0][2] n_lines_bulletpoint_start = sum(map(lambda ln: ln[2], signals["rps_lines_start_with_bulletpoint"])) if n_lines_bulletpoint_start / n_lines > 0.9: return False # rule 4: the ratio between characters in the most frequent 2-gram and the total number # of characters must be below 0.2 top_2_gram_frac = signals["rps_doc_frac_chars_top_2gram"][0][2] if top_2_gram_frac > 0.2: return False # rule 5: ... return True ``` Filtering the RedPajama-V2 dataset with this set of rules is then as easy as: ```python ds_iterator = load_dataset( "togethercomputer/RedPajama-Data-V2", snapshots=["2023-14"], languages=["en"], name="default", streaming=True ) filtered_dataset = [] for sample in ds_iterator["train"]: if not gopher_rules_pass(sample): continue filtered_dataset.append(sample) ``` ### Dataset Summary RedPajama-V2 is an open dataset for training large language models and includes over 100B text documents. Out of these, 30B documents come with quality annotations. Out of these, there are 20B unique documents. #### Quality Annotations | Annotation Tag | Description | Category | Reference | |------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|------------------|-------------------------------------------------------------------------------------------------------------------------------| | ccnet_bucket | head, middle or tail bucket of the perplexity score | CCNet | [CCNet](https://github.com/facebookresearch/cc_net) | | ccnet_language_score | score of the language identification model | CCNet | [CCNet](https://github.com/facebookresearch/cc_net) | | ccnet_length | number of characters | CCNet | [CCNet](https://github.com/facebookresearch/cc_net) | | ccnet_nlines | number of lines | CCNet | [CCNet](https://github.com/facebookresearch/cc_net) | | ccnet_original_length | number of characters before in-document line deduplication | CCNet | [CCNet](https://github.com/facebookresearch/cc_net) | | ccnet_original_nlines | number of lines before in-document line deduplication | CCNet | [CCNet](https://github.com/facebookresearch/cc_net) | | ccnet_perplexity | perplexity of an LM trained on Wikipedia | CCNet | [CCNet](https://github.com/facebookresearch/cc_net) | | rps_doc_books_importance | Given a bag of {1,2}-wordgram model trained on Books p, and a model trained on the source domain q, This is the logarithm of the ratio p(doc)/q(doc). | ML Heuristics | [Importance Resampling (Xie et al.)](https://arxiv.org/abs/2302.03169) | | rps_doc_openwebtext_importance | Given a bag of {1,2}-wordgram model trained on OpenWebText p, and a model trained on the source domain q, this is the logarithm of the ratio p(doc)/q(doc). | ML Heuristics | [Importance Resampling (Xie et al.)](https://arxiv.org/abs/2302.03169) | | rps_doc_wikipedia_importance | Given a bag of {1,2}-wordgram model trained on Wikipedia articles p, and a model trained on the source domain q, this is the logarithm of the ratio p(doc)/q(doc). | ML Heuristics | [Importance Resampling (Xie et al.)](https://arxiv.org/abs/2302.03169) | | rps_doc_ml_wikiref_score | Fasttext classifier prediction for the document being a Wikipedia reference. This is the same fasttext model used in the RedPajama-1T dataset. Only applies to English data.. | ML Heuristics | [LLaMA](https://arxiv.org/abs/2302.13971), [RedPajama-1T](https://huggingface.co/datasets/togethercomputer/RedPajama-Data-1T) | | rps_doc_ml_palm_score | Fasttext classifier prediction for the document being a Wikipedia article, OpenWebText sample or a RedPajama-V1 book. Only for English data. | ML Heuristics | [PALM](https://arxiv.org/abs/2204.02311), [GLaM](https://arxiv.org/abs/2112.06905) | | rps_doc_ml_wikipedia_score | Fasttext classifier prediction for the document being a Wikipedia article. This is used for non-English data | ML Heuristics | - | | rps_doc_curly_bracket | The ratio between the number of occurrences of '{' or '}' and the number of characters in the raw text. | Natural Language | [C4](https://arxiv.org/abs/1910.10683) | | rps_doc_frac_all_caps_words | The fraction of words in the content that only consist of uppercase letters. This is based on the raw content. | Natural Language | [Pretrainer’s Guide](https://arxiv.org/abs/2305.13169) | | rps_doc_frac_lines_end_with_ellipsis | The fraction of lines that end with an ellipsis, where an ellipsis is defined as either "..." or "…". | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_no_alph_words | The fraction of words that contain no alphabetical character. | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_lorem_ipsum | The ratio between the number of occurrences of 'lorem ipsum' and the number of characters in the content after normalisation. | Natural Language | [C4](https://arxiv.org/abs/1910.10683) | | rps_doc_mean_word_length | The mean length of words in the content after normalisation. | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_stop_word_fraction | The ratio between the number of stop words and the number of words in the document. Stop words are obtained from the [stopwords-json](https://github.com/6/stopwords-json) repo. | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_symbol_to_word_ratio | The ratio of symbols to words in the content.. Symbols are defined "#", "...", and "…". | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_unique_words | The fraction of unique words in the content. This is also known as the degeneracy of a text sample. Calculated based on the normalised content. | Natural Language | [Pretrainer’s Guide](https://arxiv.org/abs/2305.13169) | | rps_doc_unigram_entropy | The entropy of the unigram distribution of the content. This measures the diversity of the content and is computed using sum(-x / total * log(x / total)) where the sum is taken over counts of unique words in the normalised content. | Natural Language | - | | rps_doc_word_count | The number of words in the content after normalisation. | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_lines_ending_with_terminal_punctution_mark | Indicates whether a line ends with a terminal punctuation mark. A terminal punctation mark is defined as one of: ".", "!", "?", "”". | Natural Language | [C4](https://arxiv.org/abs/1910.10683) | | rps_lines_javascript_counts | The number of occurrences of the word "javascript" in each line. | Natural Language | [C4](https://arxiv.org/abs/1910.10683) | | rps_lines_num_words | The number of words in each line. This is computed based on the normalised text. | Natural Language | [C4](https://arxiv.org/abs/1910.10683) , [RefinedWeb](https://arxiv.org/abs/2306.01116) | | rps_lines_numerical_chars_fraction | The ratio between the number of numerical characters and total number of characters in each line. This is based on the normalised content. | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116) | | rps_lines_start_with_bulletpoint | Whether the lines that start with a bullet point symbol. The following set of unicodes are considered a bullet point: \u2022 (bullet point), \u2023 (triangular bullet point), \u25B6 (black right pointing triangle), \u25C0 (black left pointing triangle), \u25E6 (white bullet point), \u25A0 (black square), \u25A1 (white square), \u25AA (black small square), \u25AB (white small square), \u2013 (en dash). | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_lines_uppercase_letter_fraction | The ratio between the number of uppercase letters and total number of characters in each line. This is based on the raw text. | Natural Language | [RefinedWeb](https://arxiv.org/abs/2306.01116) | | rps_doc_num_sentences | The number of sentences in the content. This is calculated using the regular expression `r'\b[^.!?]+[.!?]*'`. | Natural Language | [C4](https://arxiv.org/abs/1910.10683) | | rps_doc_frac_chars_dupe_10grams | The fraction of characters in duplicate word 10grams. This operates on the lower-cased, punctuation removed content. It is also ensured that characters in overlapping ngrams are only counted once. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_dupe_5grams | The fraction of characters in duplicate word 5grams. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_dupe_6grams | The fraction of characters in duplicate word 6grams. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_dupe_7grams | The fraction of characters in duplicate word 7grams. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_dupe_8grams | The fraction of characters in duplicate word 8grams. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_dupe_9grams | The fraction of characters in duplicate word 9grams. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_top_2gram | The fraction of characters in the top word 2gram. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_top_3gram | The fraction of characters in the top word 3gram. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_frac_chars_top_4gram | The fraction of characters in the top word 4gram. | Repetitiveness | [RefinedWeb](https://arxiv.org/abs/2306.01116), [Gopher](https://arxiv.org/abs/2112.11446) | | rps_doc_ldnoobw_words | The number of sequences of words that are contained in the List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words blocklist. The blocklist is obtained from the [LDNOOBW](https://github.com/LDNOOBW/List-of-Dirty-Naughty-Obscene-and-Otherwise-Bad-Words) repo. | toxicity | [C4](https://arxiv.org/abs/1910.10683) | | rps_doc_ut1_blacklist | A categorical id corresponding to the list of categories of the domain of the document. Categories are obtained from the UT1 blacklist. The list is obtained from [UT-Capitole](https://dsi.ut-capitole.fr/blacklists/). | toxicictiy | [RefinedWeb](https://arxiv.org/abs/2306.01116) | | minhash_signature_0.7 | Banded minhash signature of the document, for fuzzy deduplication at Jaccard similarity 0.7. The signature is based on 128 hash functions and grouped into 14 bands and 9 rows for LSH. | Deduplication | | minhash_signature_0.8 | Banded minhash signature of the document, for fuzzy deduplication at Jaccard similarity 0.8. The signature is based on 128 hash functions and grouped into 9 bands and 13 rows for LSH. | Deduplication | | minhash_signature_0.9 | Banded minhash signature of the document, for fuzzy deduplication at Jaccard similarity 0.9. The signature is based on 128 hash functions and grouped into 5 bands and 25 rows for LSH.. | Deduplication | | minhash_signature_1.0 | Banded minhash signature of the document, for fuzzy deduplication at Jaccard similarity 1.0. The signature is based on 128 hash functions and grouped into 1 band and 128 rows for LSH. | Deduplication | #### Raw Document and Token Counts (`head_middle`) | | # Documents (deduped) | Estimated Token count (deduped) | |-------|-----------------------|---------------------------------| | en | 24.5B | 37.0T | | de | 2.7B | 4.1T | | fr | 2.2B | 3.7T | | es | 2.3B | 3.9T | | it | 1.2B | 1.9T | | Total | 32.9B | 50.6T | #### Deduplicated Document and Token Counts (`head_middle`) | | # Documents (total) | Estimated Token count (total) | |-------|---------------------|-------------------------------| | en | 14.5B | 20.5T | | de | 1.9B | 3.0T | | fr | 1.6B | 2.7T | | es | 1.8B | 2.8T | | it | 0.9B | 1.5T | | Total | 20.8B | 30.4T | ### Languages English, German, French, Italian, Spanish ## Dataset Structure The dataset is structured into four components, each following the same key structure: ``` ├── documents ├── 2018-43 ├── 0000 ├── en_head.json.gz ├── ... ├── it_middle.json.gz ├── quality_signals ├── 2018-43 ├── 0000 ├── en_head.signals.json.gz ├── ... ├── it_middle.json.gz ├── duplicates ├── 2018-43 ├── 0000 ├── en_head.duplicates.parquet ├── ... ├── it_middle.duplicates.parquet ├── minhash ├── 2018-43 ├── 0000 ├── en_head.minhash.parquet ├── ... ├── it_middle.minhash.parquet ``` Documents files, which contain the text, folow the schema defined by CCNet: ```json { "url": "...", "date_download": "2014-08-20T06:48:26Z", "digest": "sha1:46OPKWZ7MAG5624VYYA3U3YH2MJ727B6", "length": 1095, "nlines": 8, "source_domain": "...", "title": "...", "raw_content": "Dear ...", "cc_segment": "crawl-data/CC-MAIN-2014-35/...", "original_nlines": 11, "original_length": 1174, "line_ids": [ 0, 1, 3, 4, 6, 7, 8, 9 ], "language": "en", "language_score": 0.92, "perplexity": 217.2, "bucket": "head" } ``` The quality signals follow the schema ```json { "id": "2018-43/0000/en_head.json.gz/0", "id_int": 7972430436813205988, "metadata": { "cc_segment": "crawl-data/...", "cc_net_source": "2018-43/0000/en_head.json.gz", "url": "...", "source_domain": "...", "language": "en", "snapshot_id": "2018-43" }, "quality_signals": { "ccnet_original_length": [ [ 0, 7033, 8711.0 ] ], ..., "rps_doc_stop_word_fraction": [ [ 0, 7033, 0.45121107 ] ], "rps_lines_num_words": [ [ 0, 25, 2 ], ..., [ 6980, 7033, 10 ] ] } } ``` where signal scores are encoded as a list of tuples `(start, end, score)`, where `start` and `end` are the locations in the `raw_content` string where the `score` applies. ## Dataset Creation The dataset is based on 84 snapshots provided by Common Crawl. Each snapshot was processed using the CCNet pipeline and split into `head` `middle` `tail` buckets, depending on the perplexity score. In a second step, the documents in the `head` and `middle` buckets were annotated with the quality signals described above. Finally, the documents were deduplicated based on the text, using a Bloomfilter. The duplicates were kept in the dataset, but are marked in the `duplicates` component. ## Citation To cite RedPajama, please use: ``` @software{together2023redpajama, author = {Together Computer}, title = {RedPajama: an Open Dataset for Training Large Language Models}, month = October, year = 2023, url = {https://github.com/togethercomputer/RedPajama-Data} } ``` ## Acknowledgements We are appreciative to so many partners and collaborators that together are pushing forward the frontier of open LLM models. - Thank you to the OLMo team at AI2 and friends at OpenGPT-X for the insightful discussions about datasets and data quality! Also for everyone who builds on the RedPajama dataset, including Cerebras for their SlimPajama efforts, and the over 500 models built on RedPajam to date by the open-source AI community. - We are grateful to the great team at EleutherAI for paving the path on open training datasets with The Pile and for open-sourcing code we use in training some of the RedPajama models. - Thank you to our partners of RedPajama-v1, including Ontocord.ai, MILA Québec AI Institute, ETH DS3Lab, Université de Montréal, Stanford Center for Research on Foundation Models (CRFM), Stanford Hazy Research research group and LAION. ## License Please refer to the [Common Crawl Foundation Terms of Use](https://commoncrawl.org/terms-of-use) for the data. The code used to load and process the dataset is licensed under the Apache 2.0 license.