Update README.md

README.md

@@ -102,19 +102,16 @@ For each pdf document, we store statistics on the file size, number of words (as
 
 File size and page rendering time are used to set thresholds in the final dataset: the goal is to remove files that are larger than 100 MB, or that take more than 500ms to render on a modern machine, to optimize dataloading at scale. Having "too large" or "too slow" files would add a burden to large-scale training pipelines and we choose to alleviate this in the current release. Finally, a full pass over the dataset is done, trying to open and decode a bytestream from each raw object and discarding any object (pair pdf/json) that fails to be opened, to remove corrupted data.
 
-We get to 48 million pages kept as valid samples.
-
 As a last step, we use XLM-Roberta to restrict the dataset to an english subset, specifically `papluca/xlm-roberta-base-language-detection` , on the first 512 words of the first page of each document. 
 Be aware that some documents may have several languages embedded in them, or that some predictions might be inaccurate. A majority of documents from the original corpus are in English language.
-
- 
+  
 <center>
     <img src="https://huggingface.co/datasets/pixparse/pdfa-english-train/resolve/main/doc_images/languages_pdfa_xlmroberta.png" alt="A histogram of languages count in the PDFA dataset." width="600" height="300">
     <p><em>A histogram of language distribution taken on a fraction of the original -non-filtered on language- PDFA dataset. </em></p>
 </center>
 
 At the end, each document exists as a pairing of a pdf and a json file containing extensive OCR annotation as well as metadata information about rendering times. The filterings and packaging in 
-webdataset format are tailored towards multimodal machine learning at scale, specifically image-to-text tasks.
+webdataset format are tailored towards multimodal machine learning at scale, specifically image-to-text tasks. 
 
 ### Data, metadata and statistics.
 
@@ -203,14 +200,18 @@ The top-level key, `pages`, is a list of every page in the document. The above e
 
 For each page, 
 `images_bbox` gives the bounding boxes of the images embedded in the page. 
-`images_bbox_no_text_overlap` gives a reduced list of bounding boxes that have no overlap with text found in the pdf - that does not mean
+`images_bbox_no_text_overlap` gives a reduced list of bounding boxes that have no overlap with text found in the pdf. Text might be present as a drawing or another representation, however.
 ``
 
-
 `score` is a placeholder of value 1.0 for the entire dataset. 
 
+Such a formatting follows the multimodal dataset from the Industry Document Library, `https://huggingface.co/datasets/pixparse/IDL-wds`. 
+Estimating the number of tokens is done using a `LlamaTokenizer` from `tokenizers`. There is a clear power law distribution with respect to data length.
 
-Such a formatting follows the multimodal dataset from the Industry Document Library,  `https://huggingface.co/datasets/pixparse/IDL-wds`. 
+<center>
+    <img src="https://huggingface.co/datasets/pixparse/pdfa-english-train/resolve/main/doc_images/token_count_distribution.png" alt="A histogram of token count distribution per page" width="600" height="300">
+    <p><em>A histogram of token count distribution per page, taken from a subset of the dataset. There is a visible power law. </em></p>
+</center>
 
 ### Data Splits
 
@@ -227,7 +228,7 @@ Pablo Montalvo, Ross Wightman
 
 ### Disclaimer and note to researchers
 
-This dataset is intended as an OCR-heavy pretraining basis for vision-language models. As a corpus, does not represent the intent and purpose from  CC-MAIN-2021-31-PDF-UNTRUNCATED. The original is made to represent extant pdf data in its diversity and complexity. In particular, common issues related to misuse of pdfs such as mojibake (garbled text due to decoding erros) are yet to be addressed systematically, and this dataset present simplifications that can hide such issues found in the wild. In order to address this biases, we recommend to examine carefully both the simplified annotation and the original `pdf` data, beyond a simple rendering. 
+This dataset is intended as an OCR-heavy pretraining basis for vision-language models. As a corpus, it does not represent the intent and purpose from  CC-MAIN-2021-31-PDF-UNTRUNCATED. The original is made to represent extant pdf data in its diversity and complexity. In particular, common issues related to misuse of pdfs such as mojibake (garbled text due to decoding erros) are yet to be addressed systematically, and this dataset present simplifications that can hide such issues found in the wild. In order to address this biases, we recommend to examine carefully both the simplified annotation and the original `pdf` data, beyond a simple rendering. 
 
 Further, the annotation is limited to what can be extracted and is readily available - text drawn in images and only present as a bitmap rendition might be missed entirely by said annotation.