--- license: gfdl task_categories: - sentence-similarity language: - en size_categories: - 100K<n<1M configs: - config_name: raw data_files: - split: train path: raw/train-* - split: test path: raw/test-* - config_name: pair-score data_files: - split: train path: pair-score/train-* - split: test path: pair-score/test-* - config_name: pair-score-hard data_files: - split: train path: pair-score-hard/train-* - split: test path: pair-score-hard/test-* - config_name: triplet data_files: - split: train path: triplet/train-* - split: test path: triplet/test-* - config_name: triplet-hard data_files: - split: train path: triplet-hard/train-* - split: test path: triplet-hard/test-* --- # Wiki Sim ## Overview This new semi-synthetic dataset is derived from `wikimedia/wikipedia`. Each row contains 1-3 references sentences extracted from the original dataset. For each reference sentence, we use an optimized DSPy program to generate 4 similar sentences: - *Synonym* (Replace words with synonyms to maintain the same meaning.) - *Paraphrase* (Rephrase the sentence using a different structure while keeping the same idea.) - *Conceptual Overlap* (Express a related concept differently without changing the core meaning.) - *Contextual Meaning* (Modify the sentence to derive meaning from context, preserving the original intent.) Additionally, we score each result using `cross-encoder/stsb-roberta-large`. We use this to mine hard negatives from different contiguous sentences in the original passage, retaining the most similar result. ## Purpose We aim to expand training for small models like [WordLlama](https://github.com/dleemiller/WordLlama), general embedding models, and targeting benchmarks like stsb and similarity tasks differing from NLI or QnA. ## Dataset The colums of the dataset include: `synonym` `paraphrase` `conceptual_overlap` `contextual_meaning` `reference` `negative` `negative_score` `model_id` `cross_encoder` `synonym_score` `paraphrase_score` `conceptual_overlap_score` `contextual_meaning_score` where `reference` and `negative` are derived from `wikimedia/wikipedia`, and the similarity text columns are synthetically derived. We filter all rows where negative scores exceed any of the similarity scores. ## Results The 4 instruction types produce results of varying similarity scores, with the most similar being `synonym` and least similar `contextual meaning`. <img src="cdf_plot_scores.png" alt="CDF Plot" width="600"/> ## Subsets * `pair-score` - random choice weighted to a target of 0.9 * `pair-score-hard` random choice weighted to a target of 0.85 * `triplet` - random choice weighted to a target of 0.9 * `triplet-hard` - random choice weighted to a target of 0.85 * `raw` - full dataset