Create README.md

README.md

@@ -0,0 +1,51 @@
+---
+configs:
+- config_name: FrDoc2BotGeneration
+  data_files:
+    - split: train
+      path: "train.csv"
+language:
+- fr
+- vi
+---
+
+> [!NOTE]
+> Dataset origin: https://modelscope.cn/datasets/DAMO_ConvAI/FrViDoc2Bot
+
+
+
+
+
+# Document-grounded dialogue 
+Goal-oriented document-grounded dialogue systems enable end users to interactively query about
+domain-specific information based on the given documents. The tasks of querying document knowledge via conversational systems continue to attract a lot of attention from both research and industrial communities for various applications. The previous works addressed the task of English and Chinese document-grounded dialogue systems, leaving other languages less well explored. Thus, large communities of users are prevented access to automated services and information. We aim to extend the effort by introducing the Third  ACL DialDoc Workshop shared task involving documents and dialogues in diverse languages. For the first time, we provide training and evaluation data for document-grounded dialogue in Vietnamese and French, and organize the currently available Chinese and English document-grounded dialogue data. We hope that participants can leverage the linguistic similarities, for example, a large number of Vietnamese words are derived from Chinese, and English and French both belong to the Indo-European language family, to improve their models' performance in Vietnamese and French.
+
+
+### Description
+
+
+**This Project contains our annotated Vietnamese and French document-grounded dialogue training data, the development data that the participants are required to provide the model predicts, as well as the passage corpus that the training and development data depend on.**
+
+
+
+### Dataset Format
+
+Each piece of data in the training set contains three attributes: query, passages, and response.
+
+The query is a concatenation of the conversation history in reverse order, with the last turn marked as "<last_turn>", and the rest marked with "<user>" for user input and "<agent>" for system output. For example：
+'<last_turn> Quelle est la situation dans laquelle l'établissement scolaire se trouve en tant que juge et partie ? <agent> déclaration d’accident scolaire » et destiné au recteur d’académie dont il dépend. <user> Quel est le rapport que le médecin scolaire doit rédiger après un accident scolaire? <agent> Les assurances « individuelle accident » ont souvent plusieurs niveaux de garantie. <user> Quelle est la caractéristique des assurances "individuelle accident"?'
+
+
+
+The 'passages' attribute contains the passage arranged according to reply dependencies, followed by a reverse-ordered chain of titles concatenated with "//" as the delimiter. Example:
+
+'En effet, dans ce cas, l’établissement scolaire est à la fois juge (rédacteur du rapport d’accident) et partie ! // Responsabilité[modifier | modifier le code] // En cas d'accident[modifier | modifier le code] // Assurance scolaire — Wikipédia // fr-Insurance'
+
+The response attribute is the desired output, beginning with "<response>". Here's an example: '<response> En effet, dans ce cas, l’établissement scolaire est à la fois juge (rédacteur du rapport d’accident) et partie !'
+
+
+
+We have provided the [passage corpus](https://huggingface.co/datasets/de-francophones/FrViDoc2Bot/blob/main/passages.csv) that all dialogues in the training, validation, and test sets rely on in passages.csv.
+
+
+The [validation.json](https://huggingface.co/datasets/de-francophones/FrViDoc2Bot/blob/main/validation.json) file contains the development data that participants need to provide model predictions for.