YAML Metadata Warning: The pipeline tag "data-filtering" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, text2text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, other

This fastText model is a filter for selecting high-quality pretraining data, as described in Improving Pretraining Data Using Perplexity Correlations. It targets the LAMBADA IT task.

The model uses perplexity correlations to identify text segments highly correlated with strong performance on downstream benchmarks. It doesn't perform text classification directly; instead, it outputs a score indicating the suitability of a text segment for pretraining.

For complete usage instructions and the theoretical background, please refer to the project's GitHub repository.

Downloads last month: 19

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support