Download metadata/7_3.stats.json from TurkuNLP/HPLT3-PerRef: direct link, hf CLI and curl.
- Browser
- Download file 918 Bytes
-
https://huggingface.co/datasets/TurkuNLP/HPLT3-PerRef/resolve/main/metadata/7_3.stats.json
- Command line
-
hf download hf://datasets/TurkuNLP/HPLT3-PerRef/metadata/7_3.stats.json
-
curl -L -o 7_3.stats.json https://huggingface.co/datasets/TurkuNLP/HPLT3-PerRef/resolve/main/metadata/7_3.stats.json
918 Bytes
| { | |
| "source_shard": "https://data.hplt-project.org/three/sorted/pes_Arab/7_3.jsonl.zst", | |
| "source_filename": "7_3.jsonl", | |
| "model": "alrazz-n/bge-m3_spamham_best_model", | |
| "model_path": "/scratch/project_2005092/nima/saved_models/bge-m3-retromae_spamham", | |
| "threshold": 0.3, | |
| "threshold_definition": "remove documents where junk_probability >= threshold", | |
| "threshold_source": "dev set precision-recall analysis", | |
| "dev_metrics": { | |
| "junk_precision": 0.9484, | |
| "junk_recall": 0.9439, | |
| "good_recall": 0.9725 | |
| }, | |
| "class_mapping": { | |
| "0": "junk", | |
| "1": "good" | |
| }, | |
| "batch_size": 128, | |
| "max_length": 1024, | |
| "max_chars_for_inference": 20000, | |
| "kept_documents": 5759684, | |
| "junk_documents": 2863295, | |
| "junk_storage": "id + junk_probability only", | |
| "total_documents": 8622979, | |
| "kept_ratio": 0.667945961598654, | |
| "junk_ratio": 0.332054038401346, | |
| "processed_at": "2026-07-26T04:30:16.534839" | |
| } |