Update README.md
Browse files
README.md
CHANGED
|
@@ -108,9 +108,17 @@ Recently, IBM has introduced GneissWeb; a large dataset yielding around 10 trill
|
|
| 108 |
|
| 109 |
1. fastText models used in the curation of GneissWeb
|
| 110 |
1. [Quality Classifier](https://huggingface.co/ibm-granite/GneissWeb.Quality_annotator)
|
|
|
|
| 111 |
The fastText model takes as input text and classifies whether the text is ''high-quality'' (labeled as __label__hq) or ''low-quality'' (labeled as __label__cc). The model can be used with python (please refer to fasttext documentation for details on using fasttext classifiers) or with IBM Data Prep Kit (DPK) (please refer to the example notebook for using a fastText model with DPK).
|
| 112 |
The GneissWeb ensemble filter uses the confidence score given to __label__hq for filtering documents based on an appropriately chosen threshold. The fastText model is used along with [DCLM-fastText] (https://huggingface.co/mlfoundations/fasttext-oh-eli5) and other quality annotators.
|
| 113 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 114 |
|
| 115 |
2. [Bloom filter](https://huggingface.co/ibm-granite/GneissWeb.bloom) built on the document ids of GneissWeb documents. This can be used to recreat GneissWeb using the document ids from FineWeb 1.1.0 or any version of Common Crawl
|
| 116 |
|
|
|
|
| 108 |
|
| 109 |
1. fastText models used in the curation of GneissWeb
|
| 110 |
1. [Quality Classifier](https://huggingface.co/ibm-granite/GneissWeb.Quality_annotator)
|
| 111 |
+
|
| 112 |
The fastText model takes as input text and classifies whether the text is ''high-quality'' (labeled as __label__hq) or ''low-quality'' (labeled as __label__cc). The model can be used with python (please refer to fasttext documentation for details on using fasttext classifiers) or with IBM Data Prep Kit (DPK) (please refer to the example notebook for using a fastText model with DPK).
|
| 113 |
The GneissWeb ensemble filter uses the confidence score given to __label__hq for filtering documents based on an appropriately chosen threshold. The fastText model is used along with [DCLM-fastText] (https://huggingface.co/mlfoundations/fasttext-oh-eli5) and other quality annotators.
|
| 114 |
+
2. Classifiers for [Science](https://huggingface.co/ibm-granite/GneissWeb.Sci_classifier), [Technology](https://huggingface.co/ibm-granite/GneissWeb.Tech_classifier), [Medical](https://huggingface.co/ibm-granite/GneissWeb.Med_classifier) and [Education](https://huggingface.co/ibm-granite/GneissWeb.Edu_classifier)
|
| 115 |
+
|
| 116 |
+
The fastText model takes as input text and classifies whether the text categorized as ''medical'' (labeled as `__label__hq`) or other categories''cc'' (labeled as `__label__cc`).
|
| 117 |
+
The model can be used with python (please refer to [fasttext documentation](https://fasttext.cc/docs/en/python-module.html) for details on using fasttext classifiers)
|
| 118 |
+
or with [IBM Data Prep Kit](https://github.com/IBM/data-prep-kit/) (DPK) (please refer to the [example notebook](https://github.com/IBM/data-prep-kit/blob/dev/transforms/language/gneissweb_classification/gneissweb_classification.ipynb) for using a fastText model with DPK).
|
| 119 |
+
|
| 120 |
+
The GneissWeb ensemble filter uses the confidence score given to `__label__hq` for filtering documents based on an appropriately chosen threshold.
|
| 121 |
+
The fastText model is used along with [GneissWeb.Edu_classifier](https://huggingface.co/ibm-granite/GneissWeb.Edu_classifier), [GneissWeb.Tech_classifier](https://huggingface.co/ibm-granite/GneissWeb.Tech_classifier), and [GneissWeb.Sci_classifier](https://huggingface.co/ibm-granite/GneissWeb.Sci_classifier) and other quality annotators.
|
| 122 |
|
| 123 |
2. [Bloom filter](https://huggingface.co/ibm-granite/GneissWeb.bloom) built on the document ids of GneissWeb documents. This can be used to recreat GneissWeb using the document ids from FineWeb 1.1.0 or any version of Common Crawl
|
| 124 |
|