Readme Updates v2

ALIA-biomedical-datatrove.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e93ff274b7d6a93e7296fed36c2e7680615066bfd86177cfc43315e06229ffa1
-size 18547678707
+oid sha256:ac7f3024b9fbff2512e94e94ede32951db72b540007d05c8618c87aecc7fd702
+size 7550802834

README.md

@@ -105,30 +105,31 @@ Each instance in the corpus has the following structure:
 ### Data Splits
 
 The complete dataset contains the following main sources with their statistics:
+
 | Source Dataset | Num Tokens | Num Instances | Tokens Percentage | Link |
 | :--- | :---: | :---: | :---: | :--- |
 | SECOMCYC | 428,621 | 60 | 0.0077% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/SECOMCYC) |
 | SER | 441,098 | 12 | 0.0079% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/SER) |
-| SANGVA | 504,259 | 7 | 0.0090% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/SANGVA) |
+| SANGVA | 504,259 | 7 | 0.0091% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/SANGVA) |
 | Ministerio_Sanidad_Medic_Trans | 519,871 | 36 | 0.0093% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Ministerio_Sanidad_Medic_Trans) |
 | CARMEN_I | 742,437 | 1,310 | 0.0133% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/CARMEN_I) |
 | Tox_Habits | 1,061,706 | 1,040 | 0.0191% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Tox_Habits) |
 | SPA_Junta_De_Andalucia | 1,206,355 | 27 | 0.0217% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/SPA_Junta_De_Andalucia) |
 | AEPCP | 1,507,891 | 40 | 0.0271% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/AEPCP) |
-| RECCMI | 1,827,907 | 30 | 0.0328% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/RECCMI) |
 | AEPED | 1,946,814 | 788 | 0.0350% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/AEPED) |
 | Guia_Salud | 2,015,004 | 21 | 0.0362% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Guia_Salud) |
-| BARR_2 | 2,108,752 | 2,858 | 0.0378% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/BARR_2) |
+| BARR_2 | 2,108,752 | 2,858 | 0.0379% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/BARR_2) |
 | Ministerio_Sanidad_Estrategias | 6,038,859 | 172 | 0.1084% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Ministerio_Sanidad_Estrategias) |
-| Prod_Cient_AETSA | 11,169,765 | 303 | 0.2005% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Prod_Cient_AETSA) |
+| Prod_Cient_AETSA | 11,169,765 | 303 | 0.2006% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Prod_Cient_AETSA) |
 | MedlinePlus | 11,259,425 | 5,531 | 0.2021% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/MedlinePlus) |
-| Multi_Clin_Sum | 38,208,297 | 53,691 | 0.6858% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Multi_Clin_Sum) |
-| MESINESP_2 | 60,639,295 | 135,286 | 1.0884% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/MESINESP_2) |
-| Wikipedia_Biomedical | 64,881,930 | 39,601 | 1.1646% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Wikipedia_Biomedical) |
-| CIMA_AEMPS | 123,472,173 | 16,392 | 2.2162% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/CIMA_AEMPS) |
-| Miscelanea_Roberta | 1,554,077,398 | 11,776 | 27.8939% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Miscelanea_Roberta) |
-| Translated_Pubmed | 3,687,325,552 | 10,033,666 | 66.1833% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Translated_Pubmed) |
-| **TOTAL** | **5,571,383,409** | **10,302,647** | **100.00%** | - |
+| Multi_Clin_Sum | 38,208,297 | 53,691 | 0.6860% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Multi_Clin_Sum) |
+| MESINESP_2 | 60,639,295 | 135,286 | 1.0888% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/MESINESP_2) |
+| Wikipedia_Biomedical | 64,881,930 | 39,601 | 1.1650% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Wikipedia_Biomedical) |
+| CIMA_AEMPS | 123,472,173 | 16,392 | 2.2169% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/CIMA_AEMPS) |
+| Miscelanea_Roberta | 1,554,077,398 | 11,776 | 27.9031% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Miscelanea_Roberta) |
+| Translated_Pubmed | 3,687,325,552 | 10,033,666 | 66.2050% | [GitHub](https://github.com/sinai-uja/ALIA-UJA/tree/dev/data/llms/datasets/biomedical/Translated_Pubmed) |
+| **TOTAL** | **5,569,555,502** | **10,302,617** | **100.00%** | - |
+
 
 
 ### Example Usage
@@ -197,7 +198,7 @@ All data come from official and publicly accessible sources.
 
 #### Preprocessing system
 
-The corpus is based on a previous version of nearly 20 billion tokens that was processed with an advanced cleaning methodology based on [datatrove](https://github.com/huggingface/datatrove). This system automates the cleaning and preparation of large volumes of text in Spanish, eliminating duplicate and low-quality content.
+The corpus is based on a previous version of nearly 6 billion tokens that was processed with an advanced cleaning methodology based on [datatrove](https://github.com/huggingface/datatrove). This system automates the cleaning and preparation of large volumes of text in Spanish, eliminating duplicate and low-quality content.
 
 **Step 1: Configuration and paths loading**
 - Loading YAML configuration files with parameters (language threshold, filters, etc.)
@@ -218,7 +219,7 @@ The corpus is based on a previous version of nearly 20 billion tokens that was p
 
 Token counting was performed using [tiktoken](https://github.com/openai/tiktoken).
 
-The final result is a corpus of **5571383409 tokens** distributed across **10302647**, optimized for language model training.
+The final result is a corpus of **5,569,555,502 tokens** distributed across **10302617**, optimized for language model training.
 
 ### Annotations