Model Summarization Predictions on CNN DailyMail Test Set

Aggregating model predictions for a summarization task on the CNN DailyMail dataset, generated by the nbroad/longt5-base-global-mediasum model. The data is from the test split using dataset config 3.0.0. The number of rows, columns, and specific data fields are unknown.

Use Cases

• Compare predicted summaries against reference texts to evaluate model performance on the test split.
• Analyze the characteristics of summaries generated by the nbroad/longt5-base-global-mediasum model for the CNN DailyMail dataset.
• Use the model predictions as a baseline for benchmarking new summarization models on the same task and dataset.

Strengths

• Predictions are generated for a standard benchmark dataset (CNN DailyMail) using a specific model version.
• The data is associated with a defined task (Summarization) and dataset config (3.0.0).
• The source model (nbroad/longt5-base-global-mediasum) is explicitly named.

Limitations

• The dataset's structure is undefined, with columns, sample data, and row count all unknown.
• The data is a derived output (model predictions) and not the original source articles or reference summaries.
• The data was last updated in July 2022, which may not reflect the performance of current model versions.

Provenance

Source

AutoTrain Evaluator (Hugging Face)

Collection Method

Model predictions generated by AutoTrain for a summarization evaluation job.

Freshness

Last updated 2022-07-28.