update readme
Browse files
README.md
CHANGED
|
@@ -1,3 +1,86 @@
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
license: apache-2.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- text-generation
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
tags:
|
| 8 |
+
- data-juicer
|
| 9 |
+
- pretraining
|
| 10 |
+
size_categories:
|
| 11 |
+
- 1M<n<10M
|
| 12 |
---
|
| 13 |
+
|
| 14 |
+
# The Pile -- PubMed Central (refined by Data-Juicer)
|
| 15 |
+
|
| 16 |
+
A refined version of PubMed Central dataset in The Pile by [Data-Juicer](https://github.com/alibaba/data-juicer). Removing some "bad" samples from the original dataset to make it higher-quality.
|
| 17 |
+
|
| 18 |
+
This dataset is usually used to pretrain a Large Language Model.
|
| 19 |
+
|
| 20 |
+
**Notice**: Here is a small subset for previewing. The whole dataset is available [here](https://dail-wlcb.oss-cn-wulanchabu.aliyuncs.com/LLM_data/our_refined_datasets/pretraining/the-pile-pubmed-central-refine-result.jsonl) (About 83G).
|
| 21 |
+
|
| 22 |
+
## Dataset Information
|
| 23 |
+
|
| 24 |
+
- Number of samples: 2,694,860 (Keep ~86.96% from the original dataset)
|
| 25 |
+
|
| 26 |
+
## Refining Recipe
|
| 27 |
+
```yaml
|
| 28 |
+
# global parameters
|
| 29 |
+
project_name: 'Data-Juicer-recipes-pubmed-central'
|
| 30 |
+
dataset_path: '/path/to/your/dataset' # path to your dataset directory or file
|
| 31 |
+
export_path: '/path/to/your/dataset.jsonl'
|
| 32 |
+
|
| 33 |
+
np: 50 # number of subprocess to process your dataset
|
| 34 |
+
open_tracer: true
|
| 35 |
+
|
| 36 |
+
# process schedule
|
| 37 |
+
# a list of several process operators with their arguments
|
| 38 |
+
process:
|
| 39 |
+
- clean_email_mapper:
|
| 40 |
+
- clean_links_mapper:
|
| 41 |
+
- fix_unicode_mapper:
|
| 42 |
+
- punctuation_normalization_mapper:
|
| 43 |
+
- whitespace_normalization_mapper:
|
| 44 |
+
|
| 45 |
+
- alphanumeric_filter: # 89217
|
| 46 |
+
tokenization: false
|
| 47 |
+
min_ratio: 0.2787 # 3sigma
|
| 48 |
+
- average_line_length_filter: # for code
|
| 49 |
+
max_len: 1200 # < 3sigma (1478) -- 7410
|
| 50 |
+
- character_repetition_filter:
|
| 51 |
+
rep_len: 10
|
| 52 |
+
max_ratio: 0.3741 # 3sigma -- 65849
|
| 53 |
+
- flagged_words_filter:
|
| 54 |
+
lang: en
|
| 55 |
+
tokenization: true
|
| 56 |
+
max_ratio: 0.00195 # 3sigma -- 8305
|
| 57 |
+
- language_id_score_filter: # remove language filter
|
| 58 |
+
min_score: 0.5 # 272359
|
| 59 |
+
- maximum_line_length_filter: # for code
|
| 60 |
+
max_len: 7328 # remove 23808 samples
|
| 61 |
+
- perplexity_filter:
|
| 62 |
+
lang: en
|
| 63 |
+
max_ppl: 8000 # remove 173883 samples
|
| 64 |
+
- special_characters_filter:
|
| 65 |
+
max_ratio: 0.842 # remove 87661 samples
|
| 66 |
+
- text_length_filter:
|
| 67 |
+
max_len: 136028 # 3sigma -- 15118
|
| 68 |
+
- words_num_filter:
|
| 69 |
+
lang: en
|
| 70 |
+
tokenization: true
|
| 71 |
+
min_num: 20 # remove 176537 samples
|
| 72 |
+
max_num: 23305 # remove 15016 samples
|
| 73 |
+
- word_repetition_filter:
|
| 74 |
+
lang: en
|
| 75 |
+
tokenization: true
|
| 76 |
+
rep_len: 10
|
| 77 |
+
max_ratio: 0.5981 # 3sigma -- 93843
|
| 78 |
+
|
| 79 |
+
- document_simhash_deduplicator:
|
| 80 |
+
tokenization: space
|
| 81 |
+
window_size: 6
|
| 82 |
+
lowercase: true
|
| 83 |
+
ignore_pattern: '\p{P}'
|
| 84 |
+
num_blocks: 6
|
| 85 |
+
hamming_distance: 4
|
| 86 |
+
```
|