Update README.md

README.md

@@ -3,10 +3,31 @@ title: GradioFlanT5BloomAndTaskSource
 emoji: 📊
 colorFrom: green
 colorTo: gray
-sdk: docker
+sdk: gradio
+sdk_version: 3.18.0
 app_file: app.py
 pinned: false
 license: mit
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+
+The Flan-T5 model is a variant of the T5 (Text-to-Text Transfer Transformer) model, which is a large-scale neural network architecture designed for a wide range of natural language processing tasks, including language translation, question-answering, and summarization, among others.
+
+The Flan-T5 model was trained on a combination of several large-scale datasets, including:
+
+C4: The Colossal Clean Crawled Corpus - a large, multilingual dataset of web pages collected by crawling the internet, containing over 700GB of text in more than 100 languages.
+
+Wikipedia: A dataset of text extracted from Wikipedia articles in various languages.
+
+Common Crawl News: A dataset of news articles collected from various news sources.
+
+BooksCorpus: A large dataset of text passages extracted from over 11,000 books, containing over 800 million words.
+
+OpenWebText: A dataset of text scraped from web pages, similar to C4.
+
+WebText: A smaller dataset of web pages containing around 40GB of text.
+
+English-language books and articles from the JSTOR database.
+
+These datasets were preprocessed and used to train the Flan-T5 model on a range of natural language processing tasks, including text classification, question-answering, and summarization, among others. The Flan-T5 model has been fine-tuned on various downstream tasks, including sentiment analysis, summarization, and language translation.