{"bundle_note":"# Functional Syntax Publication Bundle\n\nThis document captures the current paper-ready state of the project.\n\n## Reference Layer\n\n- Book-derived functional syntax analyzer is stable on the grouped book corpus.\n- Final grouped book evaluation is exact.\n- Weak queue is empty.\n- The live unseen bank is frozen at 179 sentences and exact on category match.\n\n## Generated Corpus\n\n- Generated reference corpus size: 9,200 sentences\n- Final-category rate: 0.885\n- Keep / augmentation: 8,401 sentences\n- Review / cleanup: 399 sentences\n- Discard: 400 sentences\n- Train-ready export: `functional_syntax/data/generated_reference_train_ready.tsv`\n- Train/dev/test split:\n  - `functional_syntax/data/generated_reference_train.tsv`\n  - `functional_syntax/data/generated_reference_dev.tsv`\n  - `functional_syntax/data/generated_reference_test.tsv`\n\n## Sentence-Level Baseline\n\n- Model: `functional_syntax/models/generated_reference_sentence_classifier.pt`\n- Train accuracy: 0.9999\n- Dev accuracy: 0.9226\n- Test accuracy: 0.9155\n- Test macro F1: 0.916\n\n## Token-Level Baseline\n\n- Token corpus:\n  - `functional_syntax/data/generated_reference_token_train.tsv`\n  - `functional_syntax/data/generated_reference_token_dev.tsv`\n  - `functional_syntax/data/generated_reference_token_test.tsv`\n- Model: `functional_syntax/models/generated_reference_token_bilstm_crf.pt`\n- Train semantic accuracy: 0.9971\n- Dev semantic accuracy: 0.9192\n- Test semantic accuracy: 0.9214\n- Test syntactic accuracy: 0.9208\n- Test pragmatic accuracy: 0.9469\n- Test joint sequence accuracy: 0.7738\n\n## Narrative for Paper\n\nThe current paper story is:\n\n1. A book-grounded reference analyzer provides stable 4-layer outputs.\n2. The generated corpus is validated and split into training-ready data.\n3. A sentence-family baseline shows the weak labels are learnable.\n4. A token-level BiLSTM-CRF baseline learns semantic, syntactic, and pragmatic labels with strong token accuracy.\n5. Structural placement remains rule-derived from the book template.\n\n## Publication Summary\n\n- `functional_syntax/docs/paper_publication_summary.md`\n- `functional_syntax/data/paper_publication_summary.json`\n- `functional_syntax/docs/paper_methods_results_draft.md`\n- `functional_syntax/docs/related_research_map.md`\n- `functional_syntax/docs/external_data_use_matrix.md`\n- `functional_syntax/docs/external_data_download_plan.md`\n- `functional_syntax/docs/external_data_fetch_checklist.md`\n- `functional_syntax/docs/external_data_import_guide.md`\n- `functional_syntax/docs/external_syntax_corpus.md`\n- `functional_syntax/docs/paper_publication_package.zip`\n\n## Recommended Claim Scope\n\n- Safe claim: a hybrid Arabic functional syntax system combining rule-based structure with learned semantic/syntactic/pragmatic labeling.\n- Avoid claiming the token model is gold-supervised; it is trained on analyzer-derived weak labels.\n- Keep the book corpus as the reference layer and the generated corpus as augmentation.\n","checkpoint_excerpt":"# Functional Syntax Checkpoint\n\nDate: 2026-06-19\n\n## Current State\n\nThe book-derived functional syntax analyzer is stable on the grouped book corpus.\nThe generated corpus audit treats helper negation rows as review, not bad:\n\n- good: 8401\n- suspicious: 799\n- bad: 0\n\nThe live inference bank is also frozen as the current stable unseen regression set:\n\n- bank size: 179 sentences\n- exact category match: 1.0\n- high-confidence sentences: 179\n- medium-confidence sentences: 0\n- low-confidence sentences: 0\n\n## Built Components\n\n- Shared analyzer in `functional_syntax/analyzer.py`\n- CLI, API, visualizer, table view, report page, weak-category queue, and category drill-down routes\n- Book-derived gold scaffolds and review artifacts from `f1` to `f16`\n\n## Final Corpus Evaluation\n\nGrouped evaluation on the book corpus:\n\n- sentences: 360 unique surfaces\n- sentence category accuracy: 1.0\n- token accuracy: 1.0\n- slot accuracy: 1.0\n- semantic accuracy: 1.0\n- syntactic accuracy: 1.0\n- pragmatic accuracy: 1.0\n- structural accuracy: 1.0\n\n## Queue State\n\n- No weak categories remain in the filtered weak queue.\n- `book_weak_category_queue.tsv` is empty.\n\n## Analyzer Notes\n\n- Exact deduped override loaders are used for the book-driven categories.\n- Duplicate surface sentences are evaluated ambiguity-aware by surface form.\n- The remaining book patterns are covered by exact category paths and slot overrides.\n- Initial and mid-sentence hamzated transitive verbs now recover slots correctly:\n  - `\u0623\u0643\u0645\u0644 \u0627\u0644\u0648\u0644\u062f \u062f\u0631\u0648\u0633\u0647 \u0641\u064a \u0627\u0644\u0645\u0646\u0632\u0644` -> `\u0641 / \u0641\u0627 / \u0645\u0641 / \u0635 / \u0635`\n  - `\u0632\u064a\u062f \u0623\u0643\u0644 \u0627\u0644\u062a\u0641\u0627\u062d\u0629` -> `\u06452 / \u0641 / \u0641\u0627`\n- Attached prepositional phrases like `\u0628\u0627\u0644\u0643\u0631\u0629` now stay in `\u0635` instead of being misread as a core argument.\n- Yes/no questions ending in `\u0623\u0645 \u0644\u0627` now stay in `interrog_yes_no` instead of falling through to the negation helper path:\n  - `\u0647\u0644 \u0641\u0647\u0645\u062a \u0627\u0644\u062f\u0631\u0633 \u0623\u0645 \u0644\u0627`\n  - `\u0647\u0644 \u0623\u0643\u0644\u062a \u0627\u0644\u0641\u0627\u0643\u0647\u0629 \u0623\u0645 \u0644\u0627`\n  - `\u0647\u0644 \u0623\u062d\u0628\u0628\u062a \u0627\u0644\u0645\u0648\u0633\u064a\u0642\u0649 \u0623\u0645 \u0644\u0627`\n- Closed-class words now bypass lexical root/wazn inference and surface as function-word classes:\n  - pronouns like `\u0647\u0648`\n  - demonstratives like `\u0647\u0630\u0627`\n  - relatives like `\u0627\u0644\u0630\u064a`\n  - interrogatives like `\u0645\u0646` and `\u0645\u0627`\n  - composites like `\u0644\u0647` as `\u062d\u0631\u0641 \u062c\u0631 + \u0636\u0645\u064a\u0631`\n\n## Representative Covered Categories\n\n- `coordination`\n- `embedded_complement`\n- `object_types`\n- `causative`\n- `interrog_yes_no`\n- `topic_resumptive`\n- `restrict_illa`\n- `relative_clause`\n- `interrog_hamza_choice`\n- `vso_transitive`\n- `external_theme_amma`\n- `nominal_predicate`\n- `connected_filler`\n- `interrog_wh_place_time`\n- `vso_multi_argument`\n- `interrog_wh_subject`\n- `vso_ditransitive`\n- `restrict_innama`\n- `fronted_object_focus`\n- `interrog_wh_object`\n- `copula_kana`\n- `vocative`\n- `vso_intransitive`\n- `fronted_adjunct_focus`\n- `neg_ma`\n\n## Next Work\n\nAny remaining work is product-facing, not analyzer correctness. The live bank can be extended later, but it is not blocking.\n\n## Generated Corpus Split\n\n- The generated reference corpus has been split into keep / review / discard buckets.\n- `neg_lam` is discarded as helper-only noise.\n- `neg_la` is routed to manual review.\n- `interrog_hal` and `topic_sentence` keep their final rows, with helper-edge rows marked for review.\n- Current generated split:\n  - keep: 8401\n  - review: 399\n  - discard: 400\n\n## Generated Workflow Outputs\n\n- Augmentation input: `functional_syntax/data/generated_reference_augmentation.tsv`\n- Manual cleanup queue: `functional_syntax/data/generated_reference_cleanup_queue.tsv`\n- Train-ready export: `functional_syntax/data/generated_reference_train_ready.tsv`\n- Train/dev/test split: `functional_syntax/data/generated_reference_train.tsv`, `functional_syntax/data/generated_reference_dev.tsv`, `functional_syntax/data/generated_reference_test.tsv`\n- Baseline model: `functional_syntax/models/generated_reference_sentence_classifier.pt`\n- Training report: `functional_syntax/docs/generated_reference_sentence_train.md`\n- Token corpus: `functional_syntax/data/generated_reference_token_train.tsv`, `functional_syntax/data/generated_reference_token_dev.tsv`, `functional_syntax/data/generated_reference_token_test.tsv`\n- Token baseline model: `functional_syntax/models/generated_reference_token_bilstm_crf.pt`\n- Token training report: `functional_syntax/docs/generated_reference_token_bilstm_crf.md`\n- Paper publication bundle: `functional_syntax/docs/paper_publication_bundle.md`\n- Paper publication summary: `functional_syntax/docs/paper_publication_summary.md`\n- Methods/results draft: `functional_syntax/docs/paper_methods_results_draft.md`\n- Related research map: `functional_syntax/docs/related_research_map.md`\n- External data use matrix: `functional_syntax/docs/external_data_use_matrix.md`\n- External data download plan: `functional_syntax/docs/external_data_download_plan.md`\n- External data fetch checklist: `functional_syntax/docs/external_data_fetch_checklist.md`\n- External data import guide: `functional_syntax/docs/external_data_import_guide.md`\n- External syntax corpus: `functional_syntax/docs/external_syntax_corpus.md`\n- Imported external datasets:\n  - `functional_syntax/external_data/ud_arabic_padt/`\n  - `functional_syntax/external_data/ud_arabic_pud/`\n- `\u0643\u0645 \u0639\u062f\u062f \u0623\u0644\u0648\u0627\u0646 \u0642\u0648\u0633 \u0642\u0632\u062d` now routes as a nominal quantitative interrogative:\n  - `\u0643\u0645` \u2192 `\u0645\u00d8`\n  - `\u0639\u062f\u062f` \u2192 `\u0635`\n  - `\u0623\u0644\u0648\u0627\u0646` \u2192 `\u0645\u0641`\n- lexical morphology fixed for the same phrase:\n  - `\u0643\u0645` \u2192 `\u0627\u0633\u0645 \u0627\u0633\u062a\u0641\u0647\u0627\u0645`\n  - `\u0639\u062f\u062f` \u2192 `\u0627\u0633\u0645`\n  - `\u0642\u0648\u0633` / `\u0642\u0632\u062d` \u2192 `\u0627\u0633\u0645`\n- Publication package zip: `functional_syntax/docs/paper_publication_package.zip`\n- Workflow summary: `functional_syntax/docs/generated_reference_workflow.md`\n\n## Sentence-Family Baseline\n\n- Train accuracy: 0.9999\n- Dev accuracy: 0.9226\n- Test accuracy: 0.9155\n- Test macro F1: 0.916\n- Model type: BiLSTM sentence classifier over the weak generated labels\n- This is a sentence-family baseline, not the final token-level BiLSTM-CRF model\n\n## Token-Level Baseline\n\n- Train semantic accuracy: 0.9971\n- Dev semantic accuracy: 0.9192\n- Test semantic accuracy: 0.9214\n- Test syntactic accuracy: 0.9208\n- Test pragmatic accuracy: 0.9469\n- Test joint sequence accuracy: 0.7738\n- Model type: shared BiLSTM with three CRF heads over semantic, syntactic, pragmatic labels\n- Structural layer remains rule-derived\n","generated_corpus":{"discard":400,"final_category_rate":0.885,"keep":8401,"patterns":23,"review":399,"sentences":9200},"reference_layer":{"book_final_metrics":{"pragmatic_accuracy":1.0,"semantic_accuracy":1.0,"sentence_category_accuracy":1.0,"slot_accuracy":1.0,"structural_accuracy":1.0,"syntactic_accuracy":1.0,"token_accuracy":1.0},"book_grouped_exact":true},"sentence_baseline":{"dev_metrics":{"accuracy":0.9226,"confusion_matrix":[[124,0,0,14,0,0,0,1,0,0,0,0,0,0,0,8,3,0,0,0,1],[0,83,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0],[1,0,56,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[17,0,0,32,0,0,0,0,0,0,0,0,0,0,0,6,0,0,0,0,0],[0,0,0,0,41,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,1,0,0,0,40,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,41,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,38,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,36,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,36,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,24,0,0,0,0,0,0],[2,0,1,0,0,0,0,0,0,0,0,0,0,0,0,21,0,0,0,0,0],[4,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,9,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,13,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,8,0,0],[0,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,5,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,8]],"loss":0.3017,"macro_f1":0.924,"per_label":[{"f1":0.8294,"label_id":0,"precision":0.8378,"recall":0.8212,"support":151},{"f1":0.9881,"label_id":1,"precision":0.9881,"recall":0.9881,"support":84},{"f1":0.9573,"label_id":2,"precision":0.9333,"recall":0.9825,"support":57},{"f1":0.6337,"label_id":3,"precision":0.6957,"recall":0.5818,"support":55},{"f1":1.0,"label_id":4,"precision":1.0,"recall":1.0,"support":41},{"f1":0.9877,"label_id":5,"precision":1.0,"recall":0.9756,"support":41},{"f1":0.9877,"label_id":6,"precision":0.9756,"recall":1.0,"support":40},{"f1":0.988,"label_id":7,"precision":0.9762,"recall":1.0,"support":41},{"f1":1.0,"label_id":8,"precision":1.0,"recall":1.0,"support":40},{"f1":1.0,"label_id":9,"precision":1.0,"recall":1.0,"support":40},{"f1":1.0,"label_id":10,"precision":1.0,"recall":1.0,"support":40},{"f1":1.0,"label_id":11,"precision":1.0,"recall":1.0,"support":38},{"f1":1.0,"label_id":12,"precision":1.0,"recall":1.0,"support":36},{"f1":1.0,"label_id":13,"precision":1.0,"recall":1.0,"support":36},{"f1":1.0,"label_id":14,"precision":1.0,"recall":1.0,"support":24},{"f1":0.6885,"label_id":15,"precision":0.5676,"recall":0.875,"support":24},{"f1":0.6923,"label_id":16,"precision":0.75,"recall":0.6429,"support":14},{"f1":1.0,"label_id":17,"precision":1.0,"recall":1.0,"support":13},{"f1":0.9412,"label_id":18,"precision":1.0,"recall":0.8889,"support":9},{"f1":0.7692,"label_id":19,"precision":1.0,"recall":0.625,"support":8},{"f1":0.9412,"label_id":20,"precision":0.8889,"recall":1.0,"support":8}]},"history":[{"dev_accuracy":0.8524,"dev_loss":0.4107,"dev_macro_f1":0.8564,"epoch":1,"train_accuracy":0.7254,"train_loss":0.9376},{"dev_accuracy":0.8655,"dev_loss":0.3516,"dev_macro_f1":0.863,"epoch":2,"train_accuracy":0.8954,"train_loss":0.2192},{"dev_accuracy":0.8964,"dev_loss":0.3035,"dev_macro_f1":0.9066,"epoch":3,"train_accuracy":0.9495,"train_loss":0.0876},{"dev_accuracy":0.9226,"dev_loss":0.2801,"dev_macro_f1":0.9209,"epoch":4,"train_accuracy":0.9772,"train_loss":0.0337},{"dev_accuracy":0.9262,"dev_loss":0.2878,"dev_macro_f1":0.9267,"epoch":5,"train_accuracy":0.9905,"train_loss":0.0145},{"dev_accuracy":0.925,"dev_loss":0.3064,"dev_macro_f1":0.9238,"epoch":6,"train_accuracy":0.9957,"train_loss":0.0079},{"dev_accuracy":0.9238,"dev_loss":0.3034,"dev_macro_f1":0.9232,"epoch":7,"train_accuracy":0.997,"train_loss":0.0066},{"dev_accuracy":0.9226,"dev_loss":0.3017,"dev_macro_f1":0.924,"epoch":8,"train_accuracy":0.9993,"train_loss":0.0029}],"label_map":{"0":"vso_transitive","1":"copula_kana","10":"vocative","11":"vso_ditransitive","12":"interrog_yes_no","13":"interrog_wh_subject","14":"fronted_adjunct_focus","15":"vso_multi_argument","16":"fronted_object_focus","17":"connected_filler","18":"nominal_predicate","19":"interrog_hamza_choice","2":"coordination","20":"object_types","3":"vso_intransitive","4":"interrog_wh_place_time","5":"interrog_wh_object","6":"relative_clause","7":"restrict_illa","8":"external_theme_amma","9":"restrict_innama"},"model_path":"/Users/al-hmouz/Documents/Arabic_Tasreef/functional_syntax/models/generated_reference_sentence_classifier.pt","summary":{"dev_rows":840,"test_rows":840,"train_rows":6719},"test_metrics":{"accuracy":0.9155,"confusion_matrix":[[126,0,1,15,0,0,0,0,0,0,0,0,0,0,0,5,2,0,2,0,0],[0,84,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[2,0,49,1,0,0,0,0,0,0,0,1,2,0,0,1,0,0,0,1,0],[10,0,0,37,0,0,0,0,0,0,0,0,0,0,0,7,1,0,0,0,0],[0,0,0,0,40,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0],[0,0,0,0,0,41,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,41,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,40,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,37,0,0,0,0,0,0,0,1,0],[0,0,0,1,0,0,0,0,0,0,0,0,35,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,36,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,24,0,0,0,0,0,0],[6,0,0,4,0,0,0,0,0,0,0,0,1,0,0,13,0,0,0,0,0],[2,0,0,0,0,0,0,0,0,0,0,0,1,1,0,1,9,0,0,0,0],[1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,12,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,9,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,8,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,8]],"loss":0.3291,"macro_f1":0.916,"per_label":[{"f1":0.8456,"label_id":0,"precision":0.8571,"recall":0.8344,"support":151},{"f1":1.0,"label_id":1,"precision":1.0,"recall":1.0,"support":84},{"f1":0.9159,"label_id":2,"precision":0.98,"recall":0.8596,"support":57},{"f1":0.6549,"label_id":3,"precision":0.6379,"recall":0.6727,"support":55},{"f1":0.9877,"label_id":4,"precision":1.0,"recall":0.9756,"support":41},{"f1":1.0,"label_id":5,"precision":1.0,"recall":1.0,"support":41},{"f1":1.0,"label_id":6,"precision":1.0,"recall":1.0,"support":40},{"f1":1.0,"label_id":7,"precision":1.0,"recall":1.0,"support":41},{"f1":1.0,"label_id":8,"precision":1.0,"recall":1.0,"support":40},{"f1":1.0,"label_id":9,"precision":1.0,"recall":1.0,"support":40},{"f1":1.0,"label_id":10,"precision":1.0,"recall":1.0,"support":40},{"f1":0.9737,"label_id":11,"precision":0.9737,"recall":0.9737,"support":38},{"f1":0.9333,"label_id":12,"precision":0.8974,"recall":0.9722,"support":36},{"f1":0.973,"label_id":13,"precision":0.9474,"recall":1.0,"support":36},{"f1":1.0,"label_id":14,"precision":1.0,"recall":1.0,"support":24},{"f1":0.5098,"label_id":15,"precision":0.4815,"recall":0.5417,"support":24},{"f1":0.6923,"label_id":16,"precision":0.75,"recall":0.6429,"support":14},{"f1":0.96,"label_id":17,"precision":1.0,"recall":0.9231,"support":13},{"f1":0.9,"label_id":18,"precision":0.8182,"recall":1.0,"support":9},{"f1":0.8889,"label_id":19,"precision":0.8,"recall":1.0,"support":8},{"f1":1.0,"label_id":20,"precision":1.0,"recall":1.0,"support":8}]},"train_metrics":{"accuracy":0.9999,"confusion_matrix":[[1205,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,668,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,456,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,439,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,332,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,327,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,324,0,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,324,0,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,320,0,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,320,0,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,320,0,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,300,0,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,291,0,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,285,0,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,194,0,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,193,0,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,114,0,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,100,0,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,72,0,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,67,0],[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,67]],"loss":0.0011,"macro_f1":0.9999,"per_label":[{"f1":0.9996,"label_id":0,"precision":1.0,"recall":0.9992,"support":1206},{"f1":1.0,"label_id":1,"precision":1.0,"recall":1.0,"support":668},{"f1":1.0,"label_id":2,"precision":1.0,"recall":1.0,"support":456},{"f1":0.9989,"label_id":3,"precision":0.9977,"recall":1.0,"support":439},{"f1":1.0,"label_id":4,"precision":1.0,"recall":1.0,"support":332},{"f1":1.0,"label_id":5,"precision":1.0,"recall":1.0,"support":327},{"f1":1.0,"label_id":6,"precision":1.0,"recall":1.0,"support":324},{"f1":1.0,"label_id":7,"precision":1.0,"recall":1.0,"support":324},{"f1":1.0,"label_id":8,"precision":1.0,"recall":1.0,"support":320},{"f1":1.0,"label_id":9,"precision":1.0,"recall":1.0,"support":320},{"f1":1.0,"label_id":10,"precision":1.0,"recall":1.0,"support":320},{"f1":1.0,"label_id":11,"precision":1.0,"recall":1.0,"support":300},{"f1":1.0,"label_id":12,"precision":1.0,"recall":1.0,"support":291},{"f1":1.0,"label_id":13,"precision":1.0,"recall":1.0,"support":285},{"f1":1.0,"label_id":14,"precision":1.0,"recall":1.0,"support":194},{"f1":1.0,"label_id":15,"precision":1.0,"recall":1.0,"support":193},{"f1":1.0,"label_id":16,"precision":1.0,"recall":1.0,"support":114},{"f1":1.0,"label_id":17,"precision":1.0,"recall":1.0,"support":100},{"f1":1.0,"label_id":18,"precision":1.0,"recall":1.0,"support":72},{"f1":1.0,"label_id":19,"precision":1.0,"recall":1.0,"support":67},{"f1":1.0,"label_id":20,"precision":1.0,"recall":1.0,"support":67}]}},"token_baseline":{"best_state":{"best_dev_score":0.9271333333333334,"epoch":6,"history":[{"dev_joint_sequence_accuracy":0.575,"dev_loss":1.3413,"dev_pragmatic_accuracy":0.9045,"dev_semantic_accuracy":0.8699,"dev_syntactic_accuracy":0.8699,"epoch":1,"train_loss":2.759},{"dev_joint_sequence_accuracy":0.6738,"dev_loss":1.0195,"dev_pragmatic_accuracy":0.9266,"dev_semantic_accuracy":0.8965,"dev_syntactic_accuracy":0.8974,"epoch":2,"train_loss":1.1177},{"dev_joint_sequence_accuracy":0.7131,"dev_loss":0.9254,"dev_pragmatic_accuracy":0.9382,"dev_semantic_accuracy":0.9081,"dev_syntactic_accuracy":0.9078,"epoch":3,"train_loss":0.7277},{"dev_joint_sequence_accuracy":0.7238,"dev_loss":0.9633,"dev_pragmatic_accuracy":0.9382,"dev_semantic_accuracy":0.9126,"dev_syntactic_accuracy":0.9132,"epoch":4,"train_loss":0.4847},{"dev_joint_sequence_accuracy":0.7429,"dev_loss":0.9584,"dev_pragmatic_accuracy":0.9394,"dev_semantic_accuracy":0.9147,"dev_syntactic_accuracy":0.915,"epoch":5,"train_loss":0.3237},{"dev_joint_sequence_accuracy":0.7512,"dev_loss":1.0079,"dev_pragmatic_accuracy":0.9436,"dev_semantic_accuracy":0.9192,"dev_syntactic_accuracy":0.9186,"epoch":6,"train_loss":0.2203}]},"dev_metrics":{"joint_sequence_accuracy":0.7512,"loss":1.0079,"pragmatic_accuracy":0.9436,"sample_predictions":[{"gold_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"gold_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u2014"],"gold_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u2014"],"pred_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"pred_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u2014"],"pred_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u2014"],"sentence":""},{"gold_pragmatic":["\u062a\u0645\u0647\u064a\u062f","\u2014","\u2014","\u2014","\u2014"],"gold_semantic":["\u2014","\u2014","\u2014","\u0645\u0643\u0627\u0646","\u2014"],"gold_syntactic":["\u2014","\u2014","\u2014","\u0638\u0631\u0641","\u2014"],"pred_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014","\u2014"],"pred_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u0645\u0643\u0627\u0646","\u2014"],"pred_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u0638\u0631\u0641","\u2014"],"sentence":""},{"gold_pragmatic":["\u2014","\u2014","\u0645\u062d\u0648\u0631"],"gold_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u062a\u0642\u0628\u0644","\u0645\u0646\u0641\u0630"],"gold_syntactic":["\u0641\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u0641\u0627\u0639\u0644"],"pred_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014"],"pred_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u2014"],"pred_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u2014"],"sentence":""},{"gold_pragmatic":["\u2014","\u2014","\u2014"],"gold_semantic":["\u0645\u062d\u0645\u0648\u0644","\u2014","\u2014"],"gold_syntactic":["\u0641\u0639\u0644","\u2014","\u2014"],"pred_pragmatic":["\u0645\u062d\u0648\u0631","\u2014","\u0645\u062d\u0648\u0631"],"pred_semantic":["\u0645\u0646\u0641\u0630","\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630"],"pred_syntactic":["\u0641\u0627\u0639\u0644","\u0641\u0639\u0644","\u0641\u0627\u0639\u0644"],"sentence":""},{"gold_pragmatic":["\u062a\u0645\u0647\u064a\u062f","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"gold_semantic":["\u2014","\u0645\u0646\u0641\u0630","\u0645\u0643\u0627\u0646","\u2014"],"gold_syntactic":["\u2014","\u0641\u0627\u0639\u0644","\u0638\u0631\u0641","\u2014"],"pred_pragmatic":["\u062a\u0645\u0647\u064a\u062f","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"pred_semantic":["\u2014","\u0645\u0646\u0641\u0630","\u0645\u0643\u0627\u0646","\u2014"],"pred_syntactic":["\u2014","\u0641\u0627\u0639\u0644","\u0638\u0631\u0641","\u2014"],"sentence":""}],"semantic_accuracy":0.9192,"syntactic_accuracy":0.9186},"model_path":"/Users/al-hmouz/Documents/Arabic_Tasreef/functional_syntax/models/generated_reference_token_bilstm_crf.pt","summary":{"dev_sentences":840,"dev_tokens":3352,"test_sentences":840,"test_tokens":3332,"train_sentences":6719,"train_tokens":26590,"vocab_size":3328},"test_metrics":{"joint_sequence_accuracy":0.7738,"loss":0.9513,"pragmatic_accuracy":0.9469,"sample_predictions":[{"gold_pragmatic":["\u0645\u062d\u0648\u0631","\u2014","\u2014"],"gold_semantic":["\u0645\u0646\u0641\u0630","\u0645\u062d\u0645\u0648\u0644","\u2014"],"gold_syntactic":["\u0641\u0627\u0639\u0644","\u0641\u0639\u0644","\u2014"],"pred_pragmatic":["\u0645\u062d\u0648\u0631","\u2014","\u2014"],"pred_semantic":["\u0645\u0646\u0641\u0630","\u0645\u062d\u0645\u0648\u0644","\u2014"],"pred_syntactic":["\u0641\u0627\u0639\u0644","\u0641\u0639\u0644","\u2014"],"sentence":""},{"gold_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"gold_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u2014"],"gold_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u2014"],"pred_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"pred_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u2014"],"pred_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u2014"],"sentence":""},{"gold_pragmatic":["\u2014","\u2014","\u0645\u062d\u0648\u0631"],"gold_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u062a\u0642\u0628\u0644","\u0645\u0646\u0641\u0630"],"gold_syntactic":["\u0641\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u0641\u0627\u0639\u0644"],"pred_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014"],"pred_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644"],"pred_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647"],"sentence":""},{"gold_pragmatic":["\u062a\u0645\u0647\u064a\u062f","\u0645\u062d\u0648\u0631","\u2014"],"gold_semantic":["\u2014","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644"],"gold_syntactic":["\u2014","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647"],"pred_pragmatic":["\u062a\u0645\u0647\u064a\u062f","\u0645\u062d\u0648\u0631","\u2014"],"pred_semantic":["\u2014","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644"],"pred_syntactic":["\u2014","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647"],"sentence":""},{"gold_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014","\u2014"],"gold_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u0645\u0643\u0627\u0646","\u2014"],"gold_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u0638\u0631\u0641","\u2014"],"pred_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014","\u2014"],"pred_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u0645\u0643\u0627\u0646","\u2014"],"pred_syntactic":["\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u0638\u0631\u0641","\u2014"],"sentence":""}],"semantic_accuracy":0.9214,"syntactic_accuracy":0.9208},"train_metrics":{"joint_sequence_accuracy":0.975,"loss":0.0703,"pragmatic_accuracy":0.9953,"sample_predictions":[{"gold_pragmatic":["\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014","\u0645\u062d\u0648\u0631"],"gold_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630"],"gold_syntactic":["\u2014","\u0641\u0639\u0644","\u0641\u0627\u0639\u0644"],"pred_pragmatic":["\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014","\u0645\u062d\u0648\u0631"],"pred_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630"],"pred_syntactic":["\u2014","\u0641\u0639\u0644","\u0641\u0627\u0639\u0644"],"sentence":""},{"gold_pragmatic":["\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"gold_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u0643\u0627\u0646","\u2014"],"gold_syntactic":["\u2014","\u0641\u0639\u0644_\u0646\u0627\u0642\u0635","\u0627\u0633\u0645_\u0643\u0627\u0646","\u0638\u0631\u0641","\u2014"],"pred_pragmatic":["\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"pred_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u0643\u0627\u0646","\u2014"],"pred_syntactic":["\u2014","\u0641\u0639\u0644_\u0646\u0627\u0642\u0635","\u0627\u0633\u0645_\u0643\u0627\u0646","\u0638\u0631\u0641","\u2014"],"sentence":""},{"gold_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"gold_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u0643\u0627\u0646","\u2014"],"gold_syntactic":["\u0641\u0639\u0644_\u0646\u0627\u0642\u0635","\u0627\u0633\u0645_\u0643\u0627\u0646","\u0638\u0631\u0641","\u2014"],"pred_pragmatic":["\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"pred_semantic":["\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u0643\u0627\u0646","\u2014"],"pred_syntactic":["\u0641\u0639\u0644_\u0646\u0627\u0642\u0635","\u0627\u0633\u0645_\u0643\u0627\u0646","\u0638\u0631\u0641","\u2014"],"sentence":""},{"gold_pragmatic":["\u0645\u0628\u062a\u062f\u0623","\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"gold_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u0645\u062a\u0642\u0628\u0644"],"gold_syntactic":["\u2014","\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647"],"pred_pragmatic":["\u0645\u0628\u062a\u062f\u0623","\u2014","\u0645\u062d\u0648\u0631","\u2014","\u2014"],"pred_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u0645\u0646\u0641\u0630","\u0645\u062a\u0642\u0628\u0644","\u0645\u062a\u0642\u0628\u0644"],"pred_syntactic":["\u2014","\u0641\u0639\u0644","\u0641\u0627\u0639\u0644","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647"],"sentence":""},{"gold_pragmatic":["\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014","\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014"],"gold_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u2014","\u0645\u062a\u0642\u0628\u0644"],"gold_syntactic":["\u2014","\u0641\u0639\u0644","\u2014","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647"],"pred_pragmatic":["\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014","\u0628\u0624\u0631\u0629_\u062c\u062f\u064a\u062f","\u2014"],"pred_semantic":["\u2014","\u0645\u062d\u0645\u0648\u0644","\u2014","\u0645\u062a\u0642\u0628\u0644"],"pred_syntactic":["\u2014","\u0641\u0639\u0644","\u2014","\u0645\u0641\u0639\u0648\u0644_\u0628\u0647"],"sentence":""}],"semantic_accuracy":0.9971,"syntactic_accuracy":0.997}},"token_corpus":{"label_mix":{"dev":{"connected_filler":81,"coordination":219,"copula_kana":303,"external_theme_amma":185,"fronted_adjunct_focus":96,"fronted_object_focus":48,"interrog_hamza_choice":30,"interrog_wh_object":165,"interrog_wh_place_time":135,"interrog_wh_subject":133,"interrog_yes_no":152,"nominal_predicate":32,"object_types":29,"relative_clause":172,"restrict_illa":193,"restrict_innama":134,"vocative":174,"vso_ditransitive":163,"vso_intransitive":205,"vso_multi_argument":93,"vso_transitive":610},"test":{"connected_filler":73,"coordination":212,"copula_kana":298,"external_theme_amma":189,"fronted_adjunct_focus":96,"fronted_object_focus":49,"interrog_hamza_choice":29,"interrog_wh_object":162,"interrog_wh_place_time":140,"interrog_wh_subject":134,"interrog_yes_no":156,"nominal_predicate":33,"object_types":32,"relative_clause":174,"restrict_illa":193,"restrict_innama":137,"vocative":177,"vso_ditransitive":154,"vso_intransitive":199,"vso_multi_argument":91,"vso_transitive":604},"train":{"connected_filler":579,"coordination":1706,"copula_kana":2347,"external_theme_amma":1508,"fronted_adjunct_focus":790,"fronted_object_focus":413,"interrog_hamza_choice":257,"interrog_wh_object":1266,"interrog_wh_place_time":1114,"interrog_wh_subject":1095,"interrog_yes_no":1224,"nominal_predicate":276,"object_types":261,"relative_clause":1367,"restrict_illa":1550,"restrict_innama":1124,"vocative":1377,"vso_ditransitive":1260,"vso_intransitive":1601,"vso_multi_argument":728,"vso_transitive":4747}},"outputs":{"dev":"/Users/al-hmouz/Documents/Arabic_Tasreef/functional_syntax/data/generated_reference_token_dev.tsv","test":"/Users/al-hmouz/Documents/Arabic_Tasreef/functional_syntax/data/generated_reference_token_test.tsv","train":"/Users/al-hmouz/Documents/Arabic_Tasreef/functional_syntax/data/generated_reference_token_train.tsv"},"sentence_counts":{"dev":840,"test":840,"train":6719},"split_category_counts":{"dev":{"connected_filler":13,"coordination":57,"copula_kana":84,"external_theme_amma":40,"fronted_adjunct_focus":24,"fronted_object_focus":13,"interrog_hamza_choice":8,"interrog_wh_object":41,"interrog_wh_place_time":41,"interrog_wh_subject":26,"interrog_yes_no":36,"nominal_predicate":6,"object_types":8,"relative_clause":40,"restrict_illa":41,"restrict_innama":40,"vocative":40,"vso_ditransitive":38,"vso_intransitive":69,"vso_multi_argument":26,"vso_transitive":149},"test":{"connected_filler":13,"coordination":57,"copula_kana":84,"external_theme_amma":40,"fronted_adjunct_focus":24,"fronted_object_focus":13,"interrog_hamza_choice":8,"interrog_wh_object":41,"interrog_wh_place_time":41,"interrog_wh_subject":26,"interrog_yes_no":36,"nominal_predicate":4,"object_types":8,"relative_clause":40,"restrict_illa":41,"restrict_innama":40,"vocative":40,"vso_ditransitive":38,"vso_intransitive":68,"vso_multi_argument":28,"vso_transitive":150},"train":{"connected_filler":100,"coordination":457,"copula_kana":668,"external_theme_amma":320,"fronted_adjunct_focus":194,"fronted_object_focus":100,"interrog_hamza_choice":67,"interrog_wh_object":330,"interrog_wh_place_time":332,"interrog_wh_subject":178,"interrog_yes_no":291,"nominal_predicate":37,"object_types":67,"relative_clause":321,"restrict_illa":324,"restrict_innama":320,"vocative":320,"vso_ditransitive":300,"vso_intransitive":590,"vso_multi_argument":226,"vso_transitive":1177}},"split_confidence_counts":{"dev":{"high":299,"medium":541},"test":{"high":306,"medium":534},"train":{"high":2405,"medium":4314}},"token_counts":{"dev":3352,"test":3332,"train":26590}},"train_split":{"dev_distribution":{"connected_filler":1,"coordination":56,"copula_kana":80,"external_theme_amma":40,"fronted_adjunct_focus":24,"fronted_object_focus":10,"interrog_hamza_choice":8,"interrog_wh_object":41,"interrog_wh_place_time":41,"interrog_wh_subject":38,"interrog_yes_no":36,"neg_ma":35,"nominal_predicate":5,"object_types":8,"relative_clause":40,"restrict_illa":40,"restrict_innama":40,"vocative":40,"vso_ditransitive":38,"vso_intransitive":46,"vso_multi_argument":21,"vso_transitive":126},"dev_rows":814,"input_rows":8142,"ratios":{"dev":0.1,"test":0.1,"train":0.8},"seed":20260618,"test_distribution":{"connected_filler":1,"coordination":56,"copula_kana":80,"external_theme_amma":40,"fronted_adjunct_focus":24,"fronted_object_focus":10,"interrog_hamza_choice":8,"interrog_wh_object":41,"interrog_wh_place_time":41,"interrog_wh_subject":38,"interrog_yes_no":36,"neg_ma":35,"nominal_predicate":5,"object_types":8,"relative_clause":40,"restrict_illa":40,"restrict_innama":40,"vocative":40,"vso_ditransitive":38,"vso_intransitive":46,"vso_multi_argument":21,"vso_transitive":126},"test_rows":814,"train_distribution":{"connected_filler":7,"coordination":445,"copula_kana":640,"external_theme_amma":320,"fronted_adjunct_focus":194,"fronted_object_focus":77,"interrog_hamza_choice":67,"interrog_wh_object":326,"interrog_wh_place_time":332,"interrog_wh_subject":307,"interrog_yes_no":285,"neg_ma":281,"nominal_predicate":36,"object_types":67,"relative_clause":321,"restrict_illa":324,"restrict_innama":320,"vocative":320,"vso_ditransitive":300,"vso_intransitive":370,"vso_multi_argument":168,"vso_transitive":1007},"train_rows":6514}}