Update README.md

README.md

@@ -227,6 +227,277 @@ print("Generated Translations:", results_detokenized)
 
 ## Data
 
+### Pretraining Data
+
+The training corpus consists of 70 billion tokens of Catalan- and Spanish-centric parallel data, including all of the official European languages plus Catalan, Basque, Galician, Asturian, Aragonese and Aranese. It amounts to xxxxx parallel sentence pairs. 
+
+This highly multilingual corpus is predominantly composed of data sourced from OPUS, with additional data taken from the NTEU project and Project Aina’s existing corpora. Where little parallel Catalan <-> data could be found, synthetic Catalan data was generated from the Spanish side of the collected Spanish <-> xx corpora using Project Aina’s es-> ca model. (link and correct name). The final distribution of languages was as below:
+
+
+
+Click the expand button below to see the full list of corpora included in the training data.
+
+<details>
+<summary>Data Sources</summary>
+ 
+| Dataset                                   	| Ca-xx Languages                                                                                                	|  Es-xx Langugages                             |	  Source                                                         	|
+|-----------------------------------------------|----------------------------------------------------------------|-----------------------------------------------|-----------------------------------------------------------------------------------------------------|
+|CCMatrix		|eu			|		|Opus		|
+|DGT			|			|bg,cs,da,de,el	,et,fi,fr,ga,hr,hu,lt,lv,mt,nl,pl,pt,ro,sk,sl,sv	|		|
+|ELRC-EMEA		|			|bg,cs,da,hu,lt,lv,mt,pl,ro,sk,sl		|		|
+|EMEA			|			|bg,cs,da,el,fi,hu,lt,mt,nl,pl,ro,sk,sl,sv		|		|
+|EUBookshop		|lt,pl,pt			|cs,da,de,el,fi,fr,ga,it,lv,mt,nl,pl,pt,ro,sk,sl,sv		|Opus		|
+|Europarl		|			|bg,cs,da,el,fi,fr,hu,lt,lv,nl,pl,pt	,ro,sk,sl,sv	|Opus		|
+|Europt		|			|hr		|Opus		|
+|KDE4			|bg,cs,da,de,el	,et,eu,fi,fr,ga,gl,hr,it,lt,lv,nl,pl,pt,ro,sk,sl,sv	|bg,ga,hr	|Opus		|
+|Global Voices		| bg,de,fr,it,nl,pl,pt	|bg,de,fr,pt		| Opus		|
+|GNOME		|eu,fr,ga,gl,pt		|ga		|Opus		|
+|JRC-Arquis		|			|cs,da,et,fr,lt,lv,mt,nl,pl	,ro,sv|		|
+|MultiCCAligned	|bg,cs,de,el,et,fi,fr,hr,hu,it,lt,lv,nl,pl,ro,sk,sv	|bg,fi,fr,hr,it,lv,nl,pt		|		|
+|MultiHPLT		|et,fi,ga,hr,mt		|		|Opus		|
+|MultiParaCrawl	|bg,da		|de,fr,ga,hr,hu,it,mt,pt		|	|		|
+|MultiUN		|			|fr	|	|		|
+|News Commentary 	|		|fr		|		|
+|NLLB			|bg,da,el,et,fi,fr,gl,hu,it	,lt,lv,pt,ro,sk,sl	|bg,cs,da,de,el	,et,fi,fr,hu,it,lt,lv,nl,pl,pt	,ro,sk,sl,sv|		|
+|NTEU			|			|bg,cs,da,de,el	,et,fi,fr,ga,hr,hu,it,lt,lv,mt,nl,pl,pt,ro,sk,sl,sv	|		|
+|OpenSubtitles 	|bg,cs,da,de,el	,et,eu,fi,gl,hr,hu,lt,lv,nl,pl,pt,ro,sk,sl,sv	|da,de,fi,fr,hr,hu,it,lv,nl		|Opus		|
+|Tatoeba		|de,pt			|pt		|Opus		|
+|TildeModel		|			|bg		|Opus		|
+|UNPC			|			|fr		|		|
+|WikiMatrix		|bg,cs,da,de,el	,et,eu,fi,fr,gl,hr,hu,it,lt,nl,pl,pt,ro,sk,sl,sv	|bg,fr,hr,it,pt		|Opus		|
+|XLENT		|eu,ga,gl			|ga		|Opus		|
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+<details>
+<summary>References</summary>
+
+- to be added
+
+
+</details>
+
+</details>
+
+
+We provide an extense Datasheet section following the best practices defined by [(Gebru et al., 2021)](https://arxiv.org/pdf/1803.09010).
+
+<details>
+<summary>Datasheet</summary>
+
+#### Motivation
+
+**For what purpose was the dataset created? Was there a specific task in mind? Was there a specific gap that needed to be filled? Please provide a description.**
+
+The purpose of creating this dataset is to pre-train multilingual models on parallel data in a large number of European languages, with Spanish and Catalan as the pivot languages. We have found that there is a lack of high quality parallel data in the scale necessary for training models, particularly between mid to low resource languages, and so in this dataset we have attempted to compile all publicly available resources for the included smaller languages. 
+
+**Who created the dataset (e.g., which team, research group) and on behalf of which entity (e.g., company, institution, organization)?**
+
+The dataset has been created by the Machine Translation sub-group of the Language Technologies unit (LangTech) of the Barcelona Supercomputing Center - Centro Nacional de
+Supercomputación (BSC-CNS), which aims to advance the field of natural language processing through cutting-edge research and development
+and the use of HPC. In particular, the main contributors were Audrey Mash and Francesca Fornaciari.
+
+However, the creation of the dataset would not have been possible without the collaboration of a large number of collaborators, partners,
+and public institutions, which can be found in detail in the acknowledgements.
+
+**Who funded the creation of the dataset? If there is an associated grant, please provide the name of the grantor and the grant name and number.**
+
+This work/research has been promoted and financed by the Government of Catalonia through the [Aina project](https://projecteaina.cat/).
+
+#### Composition
+
+**What do the instances that comprise the dataset represent (e.g., documents, photos, people, countries)? Are there multiple types of instances (e.g., movies, users, and ratings; people and interactions between them; nodes and edges)? Please provide a description.**
+
+The dataset consists entirely of parallel text separated at sentence level. Specifically, data was mainly sourced from the following databases and
+repositories:
+- **Opus:** Repository which aims to provide freely available parallel datasets in order to advance work in computational linguistics and automatic translation. 
+- **ELRC:** Repository used for documenting, storing, browsing and accessing Language Resources that are collected through the European Language Resource Coordination,
+
+**How many instances are there in total (of each type, if appropriate)?**
+
+The dataset contains a diverse range of sentence pairs across multiple languages. 36.02% of the data is parallel with Catalan, 27.59% is parallel with Spanish and 0.37% is parallel with English.
+
+**Does the dataset contain all possible instances or is it a sample (not necessarily random) of instances from a larger set? If the dataset is a sample, then what is the larger set? Is the sample representative of the larger set (e.g., geographic coverage)? If so, please describe how this representativeness was validated/verified. If it is not representative of the larger set, please describe why not (e.g., to cover a more diverse range of instances, because instances were withheld or unavailable).**
+
+The dataset is a sample from various sources. Language pairs which had fewer than 100 million parallel sentence pairs after filtering and cleaning were taken in their entirety. A sample of 100 million sentence pairs was taken from language pairs which had more data than this after preprocessing. All sampling was random. Where very little data existed between Catalan and the target language, synthetic Catalan data was created in order to increase the sample size. This was done using Projecte Aina’s xxx model.
+
+**What data does each instance consist of? “Raw” data (e.g., unprocessed text or images) or features? In either case, please provide a description.**
+
+Each instance consists of a parallel sentence pair processed for deduplication, language identification, and language alignment. 
+
+**Is there a label or target associated with each instance? If so, please provide a description.**
+
+Each instance is labelled with the two languages present in the sentence pair.
+
+**Is any information missing from individual instances? If so, please provide a description, explaining why this information is missing (e.g., because it was unavailable). This does not include intentionally removed information, but might include, e.g., redacted text.**
+
+No significant information is missing from the instances.
+
+**Are relationships between individual instances made explicit (e.g., users’ movie ratings, social network links)? If so, please describe how these relationships are made explicit.**
+
+Instances are related through shared language identifiers.
+
+**Are there recommended data splits (e.g., training, development/validation, testing)? If so, please provide a description of these splits, explaining the rationale behind them.**
+
+The dataset is split randomly into training, validation, and test sets.
+
+**Are there any errors, sources of noise, or redundancies in the dataset? If so, please provide a description.**
+
+Despite filtering for alignment and language identification, a small number of misaligned sentence pairs and incorrectly labelled languages may remain present in the data. The thresholds chosen for this task aim to achieve an optimal balance, prioritising higher accuracy.
+
+**Is the dataset self-contained, or does it link to or otherwise rely on external resources (e.g., websites, tweets, other datasets)? If it links to or relies on external resources, a) are there guarantees that they will exist, and remain constant, over time; b) are there official archival versions of the complete dataset (i.e., including the external resources as they existed at the time the dataset was created); c) are there any restrictions (e.g., licenses, fees) associated with any of the external resources that might apply to a dataset consumer? Please provide descriptions of all external resources and any restrictions associated with them, as well as links or other access points, as appropriate.**
+
+The dataset is self-contained and does not rely on external resources.
+
+**Does the dataset contain data that might be considered confidential (e.g., data that is protected by legal privilege or by doctor–patient confidentiality, data that includes the content of individuals’ non-public communications)? If so, please provide a description.**
+
+The dataset does not contain confidential data.
+
+**Does the dataset contain data that, if viewed directly, might be offensive, insulting, threatening, or might otherwise cause anxiety? If so, please describe why. If the dataset does not relate to people, you may skip the remaining questions in this section.**
+
+The dataset includes web-crawled content, which may overrepresent pornographic material across languages (Kreutzer et al., 2022). We have performed no filtering for toxic material.
+
+**Does the dataset identify any subpopulations (e.g., by age, gender)? If so, please describe how these subpopulations are identified and provide a description of their respective distributions within the dataset.**
+
+The dataset does not explicitly identify any subpopulations.
+
+**Is it possible to identify individuals (i.e., one or more natural persons), either directly or indirectly (i.e., in combination with other data) from the dataset? If so, please describe how.**
+
+Web-sourced instances in the dataset may contain personally identifiable information (PII) that is publicly available on the Web, such as
+names, IP addresses, email addresses, and phone numbers. While it would be possible to indirectly identify individuals through the
+combination of multiple data points, the nature and scale of web data makes it difficult to parse such information. 
+
+**Does the dataset contain data that might be considered sensitive in any way? If so, please provide a description.**
+
+Given that the dataset includes web-sourced content and other publicly available documents, instances may inadvertently reveal financial
+information, health-related details, or forms of government identification, such as social security numbers (Subramani et al., 2023),
+especially if the content originates from less-regulated sources or user-generated platforms.
+
+#### Collection Process
+
+**How was the data collected?**
+
+This dataset is constituted by combining several sources, all of which take the form of web-sourced datasets with some preprocessing available under permissive license (p.e. Common Crawl).
+
+**What mechanisms or procedures were used to collect the data? How were these mechanisms or procedures validated?**
+
+All datasets were acquired through open direct download and validated with data integrity tests. 
+
+**If the dataset is a sample from a larger set, what was the sampling strategy?**
+
+The sampling strategy was to use the whole dataset resulting from the filtering explained in the ‘preprocessing/cleaning/labelling’ section,
+with the particularity that language pairs consisting of over 100 million sentence pairs were randomly sampled down to 100 million.
+
+**Who was involved in the data collection process and how were they compensated?**
+
+This data is generally extracted, filtered and sampled by automated processes. The code required to run these processes has been developed
+entirely by members of the LangTech data team, or otherwise obtained from open-source software. Furthermore, there has been no monetary
+consideration for acquiring data from suppliers.
+
+**Over what timeframe was the data collected? Does this timeframe match the creation timeframe of the data associated with the instances? If not, please describe the timeframe in which the data associated with the instances was created.**
+
+Data were acquired and processed from April 2023 to August 2024. However, as mentioned, much data has been obtained from open projects such
+as Common Crawl, which contains data from 2014, so it is the end date (04/2024) rather than the start date that is important.
+
+**Were any ethical review processes conducted? If so, please provide a description of these review processes, including the outcomes, as well as a link or other access point to any supporting documentation.**
+
+No particular ethical review process has been carried out as the data is mostly open and not particularly sensitive. However, we have an
+internal evaluation team and a bias team to monitor ethical issues. In addition, we work closely with ‘Observatori d'Ètica en Intel·ligència
+Artificial’ (OEIAC) and  ‘Agencia Española de Supervisión de la Inteligencia Artificial’ (AESIA) to audit the processes we carry out from an
+ethical and legal point of view, respectively.
+
+#### Preprocessing
+
+**Was any preprocessing/cleaning/labeling of the data done? If so, please provide a description. If not, you may skip the remaining questions in this section.**
+
+All data was filtered according to two specific criteria:
+- Alignment - sentence level alignments were calculated using LaBSE and sentence pairs with a score below 0.75 were discarded.
+- Language identification - The probability of being the target language was calculated using either IdiomaCognitor or Lingua.py and sentences identified as unlikely to be the correct language were filtered out. Thresholds varied by language. 
+
+**Was the “raw” data saved in addition to the preprocessed/cleaned/labeled data? If so, please provide a link or other access point to the “raw” data.**
+
+The original raw data was kept on the BSC servers but is not publicly available.
+
+**Is the software that was used to preprocess/clean/label the data available? If so, please provide a link or other access point.**
+
+No, our internal cleaning pipeline for parallel data has not been made publicly available.
+
+#### Uses
+
+**Has the dataset been used for any tasks already? If so, please provide a description.**
+
+Pre-train the SalamandraTA model family.
+
+**What (other) tasks could the dataset be used for?**
+
+The data can be used primarily to pre-train other Machine Translation models.
+
+**Is there anything about the composition of the dataset or the way it was collected and preprocessed/cleaned/labeled that might impact future uses? Is there anything a dataset consumer could do to mitigate these risks or harms?**
+ 
+Web-crawled content is over-represented with standard language varieties, impacting language model performance for minority languages.
+Language diversity in data is crucial to avoid bias, especially in encoding non-standard dialects, preventing the exclusion of demographic
+groups. Moreover, despite legal uncertainties in web-scraped data, we prioritize permissive licenses and privacy protection measures,
+acknowledging the challenges posed by personally identifiable information (PII) within large-scale datasets. Our ongoing efforts aim to
+address privacy concerns and contribute to a more inclusive linguistic dataset.
+
+**Are there tasks for which the dataset should not be used?**
+
+-
+
+#### Distribution
+
+**Will the dataset be distributed to third parties outside of the entity on behalf of which the dataset was created? If so, please provide a description.**
+
+The dataset will not be released or distributed to third parties. Any related question to distribution is omitted in this section.
+
+#### Maintenance
+
+**Who will be supporting/hosting/maintaining the dataset?**
+
+The dataset will be hosted by the Language Technologies unit (LangTech) of the Barcelona Supercomputing Center (BSC). The team will ensure
+regular updates and monitor the dataset for any issues related to content integrity, legal compliance, and bias for the sources they are
+responsible for.
+
+**How can the owner/curator/manager of the dataset be contacted?**
+
+The data owner may be contacted with the email address [email protected].
+
+**Will the dataset be updated?**
+
+The dataset will not be updated.
+
+**If the dataset relates to people, are there applicable limits on the retention of the data associated with the instances? If so, please describe these limits and explain how they will be enforced.**
+
+The dataset does not keep sensitive data that could allow direct identification of individuals, apart from the data that is publicly
+available in web-sourced content. Due to the sheer volume and diversity of web data, it is not feasible to notify individuals or manage data
+retention on an individual basis. However, efforts are made to mitigate the risks associated with sensitive information through
+pre-processing and filtering to remove identifiable or harmful content. Despite these measures, vigilance is maintained to address potential
+privacy and ethical issues.
+
+**Will older versions of the dataset continue to be supported/hosted/maintained? If so, please describe how. If not, please describe how its obsolescence will be communicated to dataset consumers.**
+
+Since the dataset will not be updated, only the final version will be kept.
+
+**If others want to extend/augment/build on/contribute to the dataset, is there a mechanism for them to do so?**
+
+The dataset does not allow for external contributions.
+
+</details>
+
 
 
 ## Evaluation