Publish embed-statute-en v1.0.0 (internal run statute-en-e2): English central-statute retriever, +48% R@1, +131% low-overlap slice, disjoint CIs
a530a08 verified | { | |
| "format_version": 2, | |
| "checkpoint": "intfloat/multilingual-e5-small", | |
| "dimension": 384, | |
| "pooling": "mean", | |
| "max_length": 256, | |
| "normalize": true, | |
| "data_provenance": "licensed", | |
| "query_prefix": "", | |
| "passage_prefix": "", | |
| "lora": { | |
| "rank": 32, | |
| "alpha": 64, | |
| "dropout": 0.0, | |
| "targets": [ | |
| "query", | |
| "value" | |
| ] | |
| }, | |
| "adapter_parameters": 589824, | |
| "notes": { | |
| "experiment": "statute-en-e2", | |
| "data_provenance": "licensed", | |
| "trained_on": "/home/arnab/quanfire-ai-data/embedding/statute-en/corpus/pairs-neg-m05.jsonl.gz", | |
| "scored_against": "/home/arnab/quanfire-ai-data/embedding/statute-en/corpus/pairs-neg-m05.jsonl.gz", | |
| "train_pairs": 32432, | |
| "adaptation": "in-distribution", | |
| "train_kinds": [ | |
| "adjacent", | |
| "heading_section", | |
| "title_lead" | |
| ], | |
| "train_languages": [ | |
| "en" | |
| ], | |
| "held_out_languages": [ | |
| "en" | |
| ], | |
| "recall_at_1_before": 0.1815, | |
| "recall_at_1_after": 0.269, | |
| "epochs": 2, | |
| "learning_rate": 0.0001, | |
| "batch_size": 256 | |
| } | |
| } | |