Skip to main content
SimilarExport ZIP
100%

using-pretrained-abstract-to-tweet-model.ipynb

✓ Published0🌍 Public
CCliffordAnderson
Last edited Jun 30, 2025
Created on Jun 30, 2025

This example shows how a pre-trained transformer model converts academic abstracts into concise, tweet-sized summaries, demonstrating a text-to-text generation workflow. It loads a Hugging Face `AutoModelForSeq2SeqLM` and `AutoTokenizer` from the `andersoncliffb/abstracts_to_tweet_model` repository, then creates a `text2text-generation` pipeline to produce the output. The notebook’s code uses the Transformers library’s pipeline API, with the model and tokenizer specified directly. It does not include a data visualization, instead focusing on the model’s generative output from input text.

AI-generated description

Similar vizzes

Loading thumbnail…

Spark Notes

This example shows a text-analytics pipeline that annotates a CSV-loaded corpus of academic articles. It renames raw columns to meaningful fields such as `article`, `journal`, `date`, and `text`, then uses Spark NLP’s `PretrainedPipeline("explain_document_ml")` to transform the data, adding tokens and other linguistic annotations. The code demonstrates schema inspection via `printSchema` and previews the resulting token column with `select("token").show()`, highlighting how Spark DataFrames integrate with a pretrained NLP model for scalable document processing.

CCliffordAnderson
60% match
Loading thumbnail…

synthetic-data-generation-of-wikipedia-infoboxes.ipynb

This example demonstrates how to generate synthetic Wikipedia infoboxes from stub articles and fine-tune a T5 model to reproduce that transformation. It shows a complete pipeline that starts with a Hugging Face dataset of women-in-religion stubs, uses GPT-4o-mini via the OpenAI API to produce infoboxes, and then trains a T5 model using the DataDreamer framework. The notebook also publishes the resulting dataset and model to the Hugging Face Hub, with experiment tracking through Weights & Biases. The visible code focuses on data loading, API key configuration, and session initialization rather than interactive chart rendering.

CCliffordAnderson
60% match
Loading thumbnail…

download-hf-dataset-as-jsonl.ipynb

This example demonstrates an ETL pipeline that downloads a HuggingFace dataset of Wikipedia stubs about women in religion and converts it into JSONL files formatted for OpenAI fine-tuning. It shows how to transform raw dataset splits into the chat-based structure with system, user, and assistant roles. The notebook uses the `datasets` library’s `load_dataset` and `to_json` methods to handle training and validation splits separately, saving the output to `/content/`. A custom `create_openai_format` function then wraps each example in the conversational structure required by OpenAI’s API.

CCliffordAnderson
55% match
Loading thumbnail…

Gist e049bb6d7696d1b2d4ff242e68fcd062

This example analyzes text data by performing topic modeling and visualization. It shows the resulting topics and their associated terms, revealing latent themes in the dataset. The code uses NLTK for tokenization and stopword removal, TextBlob for sentiment analysis, and Gensim’s LdaModel and CoherenceModel for topic extraction. Plotly’s `scatter` and `bar` traces render the interactive visualizations, with data processed from a pandas DataFrame. The rendering approach relies on Plotly’s offline mode, initiated via `pyo.init_notebook_mode()`.

AArizonaTay
55% match
Loading thumbnail…

Gist e71cb443307994e24e63465bac636d3e

This example demonstrates how to extract and summarize the content of a New York Times opinion article about regulating artificial intelligence. The notebook uses the `newspaper` library to download and parse the article, returning its title, publication date, and initial text. It then calls the `nlp()` method to generate a list of keywords and a multi-sentence summary, showing a straightforward pipeline from a URL to condensed article insights. The code relies on `Article` from `newspaper` and standard pandas imports.

GGeorgeMcIntire
54% match
Loading thumbnail…

Gist 55fc864ec9cea1107845d3ed7a3dab8b

This example demonstrates a text-cleaning pipeline for caption data, showing how raw captions are converted into tokenized, filtered lists for downstream analysis. The visualization itself is not rendered; instead, the focus is on the data preprocessing logic. The code uses `pandas` with `apply` to run natural language processing via `spaCy` (`nlp`), extracting tokens from the `Caption` column. It then applies a custom `tokens` function that removes stop words and non-alphanumeric characters using `re.match`, while preserving tokens containing "@" symbols. The result is a new `clean_caption` column, highlighting a common step in text mining workflows.

AArizonaTay
53% match
Loading thumbnail…

Gist d2f2bc6be1716820160ed5183ea07eaa

This example shows a browser-based text analytics pipeline that scrapes and processes web content before visualizing topic models. The code first uses Selenium with Chrome WebDriver to scrape data, then applies NLP libraries including NLTK, spaCy, TextBlob, and gensim’s LdaModel for topic modeling and coherence scoring. The final visualization employs Plotly’s `plotly.express` and `plotly.graph_objs` to render interactive charts, with `pyo.init_notebook_mode()` enabling notebook display. The rendering approach is evident from the use of `go` traces and `px` figures, though the specific chart type is not shown in the code snippet. The data source is dynamic web pages, as indicated by the Selenium scraping logic.

AArizonaTay
53% match
Loading thumbnail…

Minimal character-level language model with a Vanilla Recurrent Neural Network, in Python/numpy

This example shows a minimal character-level language model built with a vanilla recurrent neural network, trained on a plain text file to learn and generate sequences of characters. The visualization highlights the network’s hidden state dynamics and sampling process as it predicts subsequent characters. Implemented in Python using NumPy, the code defines a single-layer RNN with a tanh activation, applies backpropagation through time for gradient computation, and uses Adagrad for parameter updates. The training loop periodically samples from the model to display generated text, with loss values and sampled outputs printed for monitoring.

CCameron-Grams
52% match