{ "name": "neural-weights-en-us", "version": "0.6.0", "phase": "Stage 3 — street decomposition + PO box + intersection", "license": "AGPL-3.0-only", "locale": "en-us", "training": { "corpus_version": "0.4.0+stage3", "tokenizer_version": "0.6.0-a0", "steps": 100000, "best_step": 100000, "hardware": "NVIDIA A100-SXM4-40GB (Modal cloud)", "recipe": "v0.5.1 base + STAGE3 (33 BIO labels) + synth-po-box source @ 1.5x. CE-only (crf_loss_weight=0.0 after two NaN attempts with crf>0; the 33x33 transition table + bf16 was numerically unstable). lr=1.5e-4 constant, warmup=1000." }, "architecture": { "hidden_size": 384, "num_hidden_layers": 6, "num_attention_heads": 6, "intermediate_size": 1536, "max_position_embeddings": 128, "vocab_size": 48000, "num_labels": 33, "params": "29.3M (29M encoder + 9M embedding from 48K vocab)", "crf_at_training": false, "crf_at_inference": true, "phrase_priors": true }, "labels": [ "O", "B-country", "I-country", "B-region", "I-region", "B-locality", "I-locality", "B-dependent_locality", "I-dependent_locality", "B-postcode", "I-postcode", "B-subregion", "I-subregion", "B-cedex", "I-cedex", "B-venue", "I-venue", "B-street", "I-street", "B-house_number", "I-house_number", "B-street_prefix", "I-street_prefix", "B-street_suffix", "I-street_suffix", "B-unit", "I-unit", "B-po_box", "I-po_box", "B-intersection_a", "I-intersection_a", "B-intersection_b", "I-intersection_b" ], "components_supported": [ "country", "region", "locality", "dependent_locality", "postcode", "subregion", "cedex", "venue", "street", "house_number", "street_prefix", "street_suffix", "unit", "po_box", "intersection_a", "intersection_b" ], "notes": "v0.6.0 — Stage 3 ships. Schema expanded from 10 to 16 tags / 21 to 33 BIO labels. STAGE2 label IDs preserved exactly (new tags at IDs 21-32). TIGER, NAD, and BAN adapters now emit street_prefix/street/street_suffix/unit from existing structured input. New synth-po-box corpus source provides 50K PO box training examples across en-US/CA/GB/AU, fr-FR/CA, es-ES/MX/AR.", "format": { "model": "ONNX int8 dynamic (quantized from fp32)", "tokenizer": "SentencePiece unigram, byte_fallback=true, vocab_size=48000", "max_sequence_length": 128, "opset": 17, "fp32_size_mb": 111.8, "int8_size_mb": 28.1 }, "files": { "model": "model.onnx", "tokenizer": "tokenizer.model", "model_card": "model-card.json" }, "base_relpath": "/data/output-v060/checkpoints/step-100000" }