Upload folder using huggingface_hub

README.md

@@ -0,0 +1,243 @@
+---
+library_name: transformers
+language: en
+license: apache-2.0
+base_model: google/bert_uncased_L-4_H-256_A-4
+tags:
+- tld
+- embeddings
+- domains
+- multi-task-learning
+- bert
+pipeline_tag: feature-extraction
+widget:
+- text: "com"
+- text: "io"  
+- text: "ai"
+- text: "co.za"
+model-index:
+- name: TLD Embedding Model
+  results:
+  - task:
+      type: feature-extraction
+      name: TLD Embedding
+    metrics:
+    - type: spearman_correlation
+      value: 0.8976
+      name: Average Spearman Correlation
+---
+
+# TLD Embedding Model
+
+A state-of-the-art TLD (Top-Level Domain) embedding model that learns rich 96-dimensional representations from multiple data sources through multi-task learning. This model achieved an exceptional **0.8976 average Spearman correlation** across 63 features during training.
+
+## Model Overview
+
+This TLD embedding model creates semantic representations by jointly learning from four complementary prediction tasks:
+
+1. **Research Metrics** (18 features): Brand perception, trust scores, memorability, premium brand indices
+2. **Technical Metrics** (5 features): Registration statistics, domain rankings, usage patterns  
+3. **Economic Indicators** (21 features): Country-level GDP sector breakdowns mapped to TLD registries
+4. **Price Predictions** (18 features): Industry-specific market value scores from domain sales data
+
+The model uses a shared BERT encoder with task-specific prediction heads, enabling the embeddings to capture semantic, technical, economic, and market value aspects of each TLD.
+
+## Training Performance
+
+**Final Training Results (Epoch 25/25):**
+- **Overall Average Score**: 0.8976 (89.76% Spearman correlation)
+- **Training Loss**: 0.0034
+
+**Task-Specific Performance:**
+- **Research Task**: 0.80+ correlation on trust, adoption, and brand metrics
+- **Technical Task**: 0.93-0.99 correlation on registration and ranking metrics  
+- **Economic Task**: 0.89-0.96 correlation on GDP sector predictions
+- **Price Task**: 0.90-0.99 correlation on industry-specific price scores
+
+**Best Individual Metrics:**
+- `overall_score`: 0.990 Spearman correlation
+- `global_top_1m_share`: 0.993 Spearman correlation
+- `score_food`: 0.973 Spearman correlation
+- `three_letter_registration_percent`: 0.969 Spearman correlation
+
+## Architecture
+
+- **Base Model**: `google/bert_uncased_L-4_H-256_A-4` (Lightweight BERT)
+- **Embedding Dimension**: 96 (optimized for data size)
+- **Max Sequence Length**: 8 tokens (optimized for TLDs)
+- **MLP Hidden Size**: 192 with 15% dropout
+- **Task Weighting**: Research(0.25), Technical(0.20), Economic(0.15), Price(0.40)
+
+## Training Data Sources
+
+### Research Data (`tld_research_data.jsonl`)
+- **Coverage**: 150 TLDs with research metrics
+- **Features**: Trust scores, brand associations, memorability, adoption rates
+- **Source**: Survey data, brand perception studies, market research
+
+### Technical Data (`tld_technical_data.jsonl`) 
+- **Coverage**: 716 TLDs with technical metrics
+- **Features**: Registration patterns, domain rankings (Majestic), sales volumes
+- **Source**: Registry statistics, web crawl data, domain marketplaces
+
+### Economic Data (`country_economic_data.jsonl`)
+- **Coverage**: 126 TLDs mapped to country economies  
+- **Features**: GDP breakdowns by 21 industry sectors
+- **Source**: World Bank, IMF economic data mapped to ccTLD registries
+
+### Price Data (`tld_price_scores_by_industry_2025.csv`)
+- **Coverage**: 722 TLDs with price predictions
+- **Features**: 18 industry-specific price scores plus overall score
+- **Source**: Domain sales data processed through pairwise neural network (`compute_tld_scores_pairwise.py`)
+- **Industries**: Finance, healthcare, technology, automotive, food, gaming, etc.
+
+## Installation & Usage
+
+### Loading the Model
+
+```python
+from transformers import AutoTokenizer, AutoModel
+import torch
+
+# Load model and tokenizer
+model_name = "humbleworth/tld-embedding"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModel.from_pretrained(model_name)
+model.eval()
+```
+
+### Getting TLD Embeddings
+
+```python
+def get_tld_embedding(tld, model, tokenizer):
+    """Get 96-dimensional embedding for a single TLD"""
+    # Use special token format if available, otherwise prefix with dot
+    tld_text = f"[TLD_{tld}]" if f"[TLD_{tld}]" in tokenizer.vocab else f".{tld}"
+    
+    inputs = tokenizer(
+        tld_text,
+        return_tensors="pt",
+        padding="max_length", 
+        truncation=True,
+        max_length=8
+    )
+    
+    with torch.no_grad():
+        outputs = model.encoder(**inputs)
+        cls_embedding = outputs.last_hidden_state[:, 0, :]
+        tld_embedding = model.projection(cls_embedding)
+    
+    return tld_embedding.squeeze().numpy()
+
+# Example
+com_embedding = get_tld_embedding("com", model, tokenizer)
+print(f"Embedding shape: {com_embedding.shape}")  # (96,)
+```
+
+### Batch Processing
+
+```python
+def get_tld_embeddings_batch(tlds, model, tokenizer):
+    """Get embeddings for multiple TLDs efficiently"""
+    # Use special token format if available, otherwise prefix with dot
+    tld_texts = [f"[TLD_{tld}]" if f"[TLD_{tld}]" in tokenizer.vocab else f".{tld}" for tld in tlds]
+    
+    inputs = tokenizer(
+        tld_texts,
+        return_tensors="pt",
+        padding="max_length",
+        truncation=True, 
+        max_length=8
+    )
+    
+    with torch.no_grad():
+        outputs = model.encoder(**inputs)
+        cls_embeddings = outputs.last_hidden_state[:, 0, :]
+        tld_embeddings = model.projection(cls_embeddings)
+    
+    return tld_embeddings.numpy()
+
+# Process multiple TLDs
+tlds = ["com", "io", "ai", "co.za", "tech"]
+embeddings = get_tld_embeddings_batch(tlds, model, tokenizer)
+print(f"Embeddings shape: {embeddings.shape}")  # (5, 96)
+```
+
+## Key Features
+
+### Multi-Task Learning Benefits
+- **Robust Representations**: Joint learning across diverse tasks creates more stable embeddings
+- **Transfer Learning**: Knowledge from technical metrics improves price prediction and vice versa
+- **Percentile Normalization**: All features converted to percentiles for balanced learning
+
+### Industry-Specific Intelligence  
+- **18 Industry Scores**: Specialized predictions for finance, technology, healthcare, etc.
+- **Economic Mapping**: Country-level economic data enhances ccTLD understanding
+- **Market Dynamics**: Real domain sales data captures market preferences
+
+### Technical Optimizations
+- **MPS Support**: Optimized for Apple Silicon (M1/M2) training
+- **Gradient Accumulation**: Stable training with effective batch size of 64
+- **Early Stopping**: Prevents overfitting with patience-based stopping
+- **Task Weighting**: Balanced learning prioritizing price prediction (40% weight)
+
+## Use Cases
+
+1. **Domain Valuation**: Use embeddings as features for ML-based domain appraisal
+2. **TLD Recommendation**: Find similar TLDs for branding or investment decisions  
+3. **Market Analysis**: Cluster TLDs by business characteristics or market positioning
+4. **Portfolio Optimization**: Analyze TLD portfolios using semantic similarity
+5. **Cross-Market Analysis**: Compare TLD performance across different industries
+
+## Training Configuration
+
+**Optimal Hyperparameters (Based on Data Analysis):**
+- Epochs: 25 (early stopping at patience=5)
+- Batch Size: 16 (effective 64 with accumulation) 
+- Learning Rate: 5e-4 with warmup
+- Warmup Steps: 200
+- Gradient Accumulation: 4 steps
+- Dropout: 15%
+
+**Training Command:**
+```bash
+python train_dual_task_embeddings.py \
+    --epochs 25 \
+    --batch-size 16 \
+    --learning-rate 5e-4 \
+    --warmup-steps 200 \
+    --output-dir models/tld_embedding_model
+```
+
+## Model Files
+
+```
+tld_embedding_model/
+├── config.json                    # Model configuration
+├── pytorch_model.bin              # Model weights  
+├── tokenizer.json                 # Tokenizer
+├── tokenizer_config.json          # Tokenizer config
+├── vocab.txt                      # Vocabulary
+├── special_tokens_map.json        # Special tokens
+├── training_metrics.pt            # Training metrics
+├── tld_embeddings.json           # Pre-computed embeddings
+└── README.md                      # This file
+```
+
+## Citation
+
+If you use this model in your research, please cite:
+
+```bibtex
+@software{tld_embedding_2025,
+  title = {TLD Embedding Model: Multi-Task Learning for Domain Extensions},
+  author = {HumbleWorth},
+  year = {2025},
+  note = {Achieved 0.8976 average Spearman correlation across 63 features},
+  url = {https://huggingface.co/humbleworth/tld-embedding}
+}
+```
+
+## License
+
+This model is released under the Apache 2.0 License.

added_tokens.json

@@ -0,0 +1,684 @@
+{
+  "[TLD_ac.at]": 30523,
+  "[TLD_ac.cn]": 30524,
+  "[TLD_ac.id]": 30525,
+  "[TLD_ac.il]": 30526,
+  "[TLD_ac.in]": 30527,
+  "[TLD_ac.ir]": 30528,
+  "[TLD_ac.jp]": 30529,
+  "[TLD_ac.ke]": 30530,
+  "[TLD_ac.kr]": 30531,
+  "[TLD_ac.nz]": 30532,
+  "[TLD_ac.th]": 30533,
+  "[TLD_ac.uk]": 30534,
+  "[TLD_ac.za]": 30535,
+  "[TLD_ac]": 30522,
+  "[TLD_academy]": 30536,
+  "[TLD_accountants]": 30537,
+  "[TLD_ad]": 30538,
+  "[TLD_ae]": 30539,
+  "[TLD_aero]": 30540,
+  "[TLD_africa]": 30541,
+  "[TLD_ag]": 30542,
+  "[TLD_agency]": 30543,
+  "[TLD_ai]": 30544,
+  "[TLD_al]": 30545,
+  "[TLD_am]": 30546,
+  "[TLD_apartments]": 30547,
+  "[TLD_app]": 30548,
+  "[TLD_ar]": 30549,
+  "[TLD_archi]": 30550,
+  "[TLD_art]": 30551,
+  "[TLD_as]": 30552,
+  "[TLD_asia]": 30553,
+  "[TLD_asn.au]": 30554,
+  "[TLD_associates]": 30555,
+  "[TLD_at]": 30556,
+  "[TLD_au]": 30557,
+  "[TLD_auction]": 30558,
+  "[TLD_audio]": 30559,
+  "[TLD_autos]": 30560,
+  "[TLD_az]": 30561,
+  "[TLD_ba]": 30562,
+  "[TLD_baby]": 30563,
+  "[TLD_band]": 30564,
+  "[TLD_bank]": 30565,
+  "[TLD_bar]": 30566,
+  "[TLD_bargains]": 30567,
+  "[TLD_be]": 30568,
+  "[TLD_beauty]": 30569,
+  "[TLD_beer]": 30570,
+  "[TLD_bel.tr]": 30571,
+  "[TLD_berlin]": 30572,
+  "[TLD_best]": 30573,
+  "[TLD_bet]": 30574,
+  "[TLD_bg]": 30575,
+  "[TLD_bid]": 30576,
+  "[TLD_bike]": 30577,
+  "[TLD_bingo]": 30578,
+  "[TLD_bio]": 30579,
+  "[TLD_biz.pl]": 30581,
+  "[TLD_biz]": 30580,
+  "[TLD_black]": 30582,
+  "[TLD_blog]": 30583,
+  "[TLD_blue]": 30584,
+  "[TLD_bo]": 30585,
+  "[TLD_boats]": 30586,
+  "[TLD_bond]": 30587,
+  "[TLD_boston]": 30588,
+  "[TLD_boutique]": 30589,
+  "[TLD_br]": 30590,
+  "[TLD_brussels]": 30591,
+  "[TLD_builders]": 30592,
+  "[TLD_business]": 30593,
+  "[TLD_buzz]": 30594,
+  "[TLD_by]": 30595,
+  "[TLD_bz]": 30596,
+  "[TLD_bzh]": 30597,
+  "[TLD_ca]": 30598,
+  "[TLD_cab]": 30599,
+  "[TLD_cafe]": 30600,
+  "[TLD_cam]": 30601,
+  "[TLD_camera]": 30602,
+  "[TLD_camp]": 30603,
+  "[TLD_capital]": 30604,
+  "[TLD_cards]": 30605,
+  "[TLD_care]": 30606,
+  "[TLD_casa]": 30607,
+  "[TLD_cash]": 30608,
+  "[TLD_casino]": 30609,
+  "[TLD_cat]": 30610,
+  "[TLD_cc]": 30611,
+  "[TLD_cd]": 30612,
+  "[TLD_center]": 30613,
+  "[TLD_ceo]": 30614,
+  "[TLD_cf]": 30615,
+  "[TLD_cfd]": 30616,
+  "[TLD_ch]": 30617,
+  "[TLD_charity]": 30618,
+  "[TLD_chat]": 30619,
+  "[TLD_cheap]": 30620,
+  "[TLD_christmas]": 30621,
+  "[TLD_church]": 30622,
+  "[TLD_ci]": 30623,
+  "[TLD_city]": 30624,
+  "[TLD_cl]": 30625,
+  "[TLD_claims]": 30626,
+  "[TLD_click]": 30627,
+  "[TLD_clinic]": 30628,
+  "[TLD_clothing]": 30629,
+  "[TLD_cloud]": 30630,
+  "[TLD_club]": 30631,
+  "[TLD_cm]": 30632,
+  "[TLD_cn]": 30633,
+  "[TLD_co.at]": 30635,
+  "[TLD_co.id]": 30636,
+  "[TLD_co.il]": 30637,
+  "[TLD_co.in]": 30638,
+  "[TLD_co.jp]": 30639,
+  "[TLD_co.ke]": 30640,
+  "[TLD_co.kr]": 30641,
+  "[TLD_co.nz]": 30642,
+  "[TLD_co.th]": 30643,
+  "[TLD_co.tz]": 30644,
+  "[TLD_co.uk]": 30645,
+  "[TLD_co.za]": 30646,
+  "[TLD_co.zw]": 30647,
+  "[TLD_co]": 30634,
+  "[TLD_coach]": 30648,
+  "[TLD_codes]": 30649,
+  "[TLD_coffee]": 30650,
+  "[TLD_com.ar]": 30652,
+  "[TLD_com.au]": 30653,
+  "[TLD_com.az]": 30654,
+  "[TLD_com.bd]": 30655,
+  "[TLD_com.br]": 30656,
+  "[TLD_com.bz]": 30657,
+  "[TLD_com.cn]": 30658,
+  "[TLD_com.co]": 30659,
+  "[TLD_com.cy]": 30660,
+  "[TLD_com.do]": 30661,
+  "[TLD_com.ec]": 30662,
+  "[TLD_com.eg]": 30663,
+  "[TLD_com.es]": 30664,
+  "[TLD_com.gh]": 30665,
+  "[TLD_com.hk]": 30666,
+  "[TLD_com.in]": 30667,
+  "[TLD_com.kg]": 30668,
+  "[TLD_com.mt]": 30669,
+  "[TLD_com.mx]": 30670,
+  "[TLD_com.my]": 30671,
+  "[TLD_com.ng]": 30672,
+  "[TLD_com.np]": 30673,
+  "[TLD_com.pe]": 30674,
+  "[TLD_com.ph]": 30675,
+  "[TLD_com.pk]": 30676,
+  "[TLD_com.pl]": 30677,
+  "[TLD_com.py]": 30678,
+  "[TLD_com.sa]": 30679,
+  "[TLD_com.sg]": 30680,
+  "[TLD_com.tr]": 30681,
+  "[TLD_com.tw]": 30682,
+  "[TLD_com.ua]": 30683,
+  "[TLD_com.uy]": 30684,
+  "[TLD_com.vc]": 30685,
+  "[TLD_com.ve]": 30686,
+  "[TLD_com.vn]": 30687,
+  "[TLD_com]": 30651,
+  "[TLD_community]": 30688,
+  "[TLD_company]": 30689,
+  "[TLD_computer]": 30690,
+  "[TLD_construction]": 30691,
+  "[TLD_consulting]": 30692,
+  "[TLD_contact]": 30693,
+  "[TLD_contractors]": 30694,
+  "[TLD_cooking]": 30695,
+  "[TLD_cool]": 30696,
+  "[TLD_coop]": 30697,
+  "[TLD_country]": 30698,
+  "[TLD_coupons]": 30699,
+  "[TLD_credit]": 30700,
+  "[TLD_cruises]": 30701,
+  "[TLD_cu]": 30702,
+  "[TLD_cx]": 30703,
+  "[TLD_cyou]": 30704,
+  "[TLD_cz]": 30705,
+  "[TLD_dance]": 30706,
+  "[TLD_date]": 30707,
+  "[TLD_dating]": 30708,
+  "[TLD_de]": 30709,
+  "[TLD_deals]": 30710,
+  "[TLD_delivery]": 30711,
+  "[TLD_dental]": 30712,
+  "[TLD_desi]": 30713,
+  "[TLD_design]": 30714,
+  "[TLD_dev]": 30715,
+  "[TLD_diamonds]": 30716,
+  "[TLD_diet]": 30717,
+  "[TLD_digital]": 30718,
+  "[TLD_direct]": 30719,
+  "[TLD_directory]": 30720,
+  "[TLD_discount]": 30721,
+  "[TLD_dj]": 30722,
+  "[TLD_dk]": 30723,
+  "[TLD_do]": 30724,
+  "[TLD_doctor]": 30725,
+  "[TLD_dog]": 30726,
+  "[TLD_domains]": 30727,
+  "[TLD_download]": 30728,
+  "[TLD_dz]": 30729,
+  "[TLD_earth]": 30730,
+  "[TLD_ec]": 30731,
+  "[TLD_eco]": 30732,
+  "[TLD_ed.jp]": 30733,
+  "[TLD_edu.ar]": 30735,
+  "[TLD_edu.au]": 30736,
+  "[TLD_edu.br]": 30737,
+  "[TLD_edu.cn]": 30738,
+  "[TLD_edu.co]": 30739,
+  "[TLD_edu.ec]": 30740,
+  "[TLD_edu.eg]": 30741,
+  "[TLD_edu.hk]": 30742,
+  "[TLD_edu.in]": 30743,
+  "[TLD_edu.mx]": 30744,
+  "[TLD_edu.my]": 30745,
+  "[TLD_edu.ng]": 30746,
+  "[TLD_edu.pe]": 30747,
+  "[TLD_edu.ph]": 30748,
+  "[TLD_edu.pk]": 30749,
+  "[TLD_edu.pl]": 30750,
+  "[TLD_edu.sa]": 30751,
+  "[TLD_edu.sg]": 30752,
+  "[TLD_edu.tr]": 30753,
+  "[TLD_edu.tw]": 30754,
+  "[TLD_edu.ua]": 30755,
+  "[TLD_edu.uy]": 30756,
+  "[TLD_edu.vn]": 30757,
+  "[TLD_edu]": 30734,
+  "[TLD_education]": 30758,
+  "[TLD_ee]": 30759,
+  "[TLD_email]": 30760,
+  "[TLD_energy]": 30761,
+  "[TLD_engineering]": 30762,
+  "[TLD_enterprises]": 30763,
+  "[TLD_equipment]": 30764,
+  "[TLD_es]": 30765,
+  "[TLD_estate]": 30766,
+  "[TLD_et]": 30767,
+  "[TLD_eu]": 30768,
+  "[TLD_eus]": 30769,
+  "[TLD_events]": 30770,
+  "[TLD_exchange]": 30771,
+  "[TLD_expert]": 30772,
+  "[TLD_exposed]": 30773,
+  "[TLD_express]": 30774,
+  "[TLD_fail]": 30775,
+  "[TLD_faith]": 30776,
+  "[TLD_family]": 30777,
+  "[TLD_fan]": 30778,
+  "[TLD_fans]": 30779,
+  "[TLD_farm]": 30780,
+  "[TLD_fashion]": 30781,
+  "[TLD_fi]": 30782,
+  "[TLD_finance]": 30783,
+  "[TLD_financial]": 30784,
+  "[TLD_fish]": 30785,
+  "[TLD_fit]": 30786,
+  "[TLD_fitness]": 30787,
+  "[TLD_flowers]": 30788,
+  "[TLD_fm]": 30789,
+  "[TLD_football]": 30790,
+  "[TLD_forsale]": 30791,
+  "[TLD_foundation]": 30792,
+  "[TLD_fr]": 30793,
+  "[TLD_fun]": 30794,
+  "[TLD_fund]": 30795,
+  "[TLD_furniture]": 30796,
+  "[TLD_fyi]": 30797,
+  "[TLD_gal]": 30798,
+  "[TLD_gallery]": 30799,
+  "[TLD_game]": 30800,
+  "[TLD_games]": 30801,
+  "[TLD_garden]": 30802,
+  "[TLD_gd]": 30803,
+  "[TLD_gdn]": 30804,
+  "[TLD_ge]": 30805,
+  "[TLD_gen.tr]": 30806,
+  "[TLD_gg]": 30807,
+  "[TLD_gift]": 30808,
+  "[TLD_gifts]": 30809,
+  "[TLD_gives]": 30810,
+  "[TLD_gl]": 30811,
+  "[TLD_glass]": 30812,
+  "[TLD_global]": 30813,
+  "[TLD_go.jp]": 30814,
+  "[TLD_gold]": 30815,
+  "[TLD_golf]": 30816,
+  "[TLD_google]": 30817,
+  "[TLD_gov.ae]": 30819,
+  "[TLD_gov.ar]": 30820,
+  "[TLD_gov.au]": 30821,
+  "[TLD_gov.bd]": 30822,
+  "[TLD_gov.br]": 30823,
+  "[TLD_gov.by]": 30824,
+  "[TLD_gov.cn]": 30825,
+  "[TLD_gov.co]": 30826,
+  "[TLD_gov.eg]": 30827,
+  "[TLD_gov.gr]": 30828,
+  "[TLD_gov.hk]": 30829,
+  "[TLD_gov.il]": 30830,
+  "[TLD_gov.in]": 30831,
+  "[TLD_gov.it]": 30832,
+  "[TLD_gov.kh]": 30833,
+  "[TLD_gov.lk]": 30834,
+  "[TLD_gov.lv]": 30835,
+  "[TLD_gov.mo]": 30836,
+  "[TLD_gov.my]": 30837,
+  "[TLD_gov.ng]": 30838,
+  "[TLD_gov.np]": 30839,
+  "[TLD_gov.ph]": 30840,
+  "[TLD_gov.pk]": 30841,
+  "[TLD_gov.pl]": 30842,
+  "[TLD_gov.pt]": 30843,
+  "[TLD_gov.rs]": 30844,
+  "[TLD_gov.sa]": 30845,
+  "[TLD_gov.sg]": 30846,
+  "[TLD_gov.tr]": 30847,
+  "[TLD_gov.tw]": 30848,
+  "[TLD_gov.ua]": 30849,
+  "[TLD_gov.uk]": 30850,
+  "[TLD_gov.vn]": 30851,
+  "[TLD_gov.za]": 30852,
+  "[TLD_gov]": 30818,
+  "[TLD_gr.jp]": 30854,
+  "[TLD_gr]": 30853,
+  "[TLD_graphics]": 30855,
+  "[TLD_gratis]": 30856,
+  "[TLD_green]": 30857,
+  "[TLD_group]": 30858,
+  "[TLD_gs]": 30859,
+  "[TLD_guide]": 30860,
+  "[TLD_guru]": 30861,
+  "[TLD_gy]": 30862,
+  "[TLD_hair]": 30863,
+  "[TLD_haus]": 30864,
+  "[TLD_health]": 30865,
+  "[TLD_healthcare]": 30866,
+  "[TLD_help]": 30867,
+  "[TLD_hk]": 30868,
+  "[TLD_hn]": 30869,
+  "[TLD_hockey]": 30870,
+  "[TLD_holdings]": 30871,
+  "[TLD_holiday]": 30872,
+  "[TLD_homes]": 30873,
+  "[TLD_horse]": 30874,
+  "[TLD_host]": 30875,
+  "[TLD_hosting]": 30876,
+  "[TLD_house]": 30877,
+  "[TLD_hr]": 30878,
+  "[TLD_ht]": 30879,
+  "[TLD_hu]": 30880,
+  "[TLD_icu]": 30881,
+  "[TLD_id]": 30882,
+  "[TLD_ie]": 30883,
+  "[TLD_im]": 30884,
+  "[TLD_immobilien]": 30885,
+  "[TLD_in]": 30886,
+  "[TLD_inc]": 30887,
+  "[TLD_info.pl]": 30889,
+  "[TLD_info]": 30888,
+  "[TLD_ink]": 30890,
+  "[TLD_institute]": 30891,
+  "[TLD_insure]": 30892,
+  "[TLD_int]": 30893,
+  "[TLD_international]": 30894,
+  "[TLD_investments]": 30895,
+  "[TLD_io]": 30896,
+  "[TLD_ir]": 30897,
+  "[TLD_irish]": 30898,
+  "[TLD_is]": 30899,
+  "[TLD_it]": 30900,
+  "[TLD_je]": 30901,
+  "[TLD_jewelry]": 30902,
+  "[TLD_jobs]": 30903,
+  "[TLD_jp]": 30904,
+  "[TLD_kaufen]": 30905,
+  "[TLD_kg]": 30906,
+  "[TLD_kim]": 30907,
+  "[TLD_kitchen]": 30908,
+  "[TLD_kr]": 30909,
+  "[TLD_kz]": 30910,
+  "[TLD_la]": 30911,
+  "[TLD_land]": 30912,
+  "[TLD_lat]": 30913,
+  "[TLD_law]": 30914,
+  "[TLD_lawyer]": 30915,
+  "[TLD_lc]": 30916,
+  "[TLD_lease]": 30917,
+  "[TLD_legal]": 30918,
+  "[TLD_lg.jp]": 30919,
+  "[TLD_lgbt]": 30920,
+  "[TLD_li]": 30921,
+  "[TLD_life]": 30922,
+  "[TLD_lighting]": 30923,
+  "[TLD_limited]": 30924,
+  "[TLD_limo]": 30925,
+  "[TLD_link]": 30926,
+  "[TLD_live]": 30927,
+  "[TLD_lk]": 30928,
+  "[TLD_llc]": 30929,
+  "[TLD_loan]": 30930,
+  "[TLD_loans]": 30931,
+  "[TLD_lol]": 30932,
+  "[TLD_london]": 30933,
+  "[TLD_love]": 30934,
+  "[TLD_lt]": 30935,
+  "[TLD_ltd]": 30936,
+  "[TLD_lu]": 30937,
+  "[TLD_luxe]": 30938,
+  "[TLD_luxury]": 30939,
+  "[TLD_lv]": 30940,
+  "[TLD_ly]": 30941,
+  "[TLD_ma]": 30942,
+  "[TLD_makeup]": 30943,
+  "[TLD_management]": 30944,
+  "[TLD_market]": 30945,
+  "[TLD_marketing]": 30946,
+  "[TLD_mba]": 30947,
+  "[TLD_md]": 30948,
+  "[TLD_me.uk]": 30950,
+  "[TLD_me]": 30949,
+  "[TLD_media.pl]": 30952,
+  "[TLD_media]": 30951,
+  "[TLD_men]": 30953,
+  "[TLD_miami]": 30954,
+  "[TLD_mil]": 30955,
+  "[TLD_mk]": 30956,
+  "[TLD_ml]": 30957,
+  "[TLD_mn]": 30958,
+  "[TLD_mobi]": 30959,
+  "[TLD_moda]": 30960,
+  "[TLD_moe]": 30961,
+  "[TLD_mom]": 30962,
+  "[TLD_money]": 30963,
+  "[TLD_monster]": 30964,
+  "[TLD_mortgage]": 30965,
+  "[TLD_motorcycles]": 30966,
+  "[TLD_movie]": 30967,
+  "[TLD_ms]": 30968,
+  "[TLD_mt]": 30969,
+  "[TLD_mu]": 30970,
+  "[TLD_museum]": 30971,
+  "[TLD_mx]": 30972,
+  "[TLD_my]": 30973,
+  "[TLD_name]": 30974,
+  "[TLD_ne.jp]": 30975,
+  "[TLD_net.au]": 30977,
+  "[TLD_net.br]": 30978,
+  "[TLD_net.cn]": 30979,
+  "[TLD_net.co]": 30980,
+  "[TLD_net.nz]": 30981,
+  "[TLD_net.ph]": 30982,
+  "[TLD_net.pl]": 30983,
+  "[TLD_net.ua]": 30984,
+  "[TLD_net]": 30976,
+  "[TLD_network]": 30985,
+  "[TLD_news]": 30986,
+  "[TLD_nf]": 30987,
+  "[TLD_ng]": 30988,
+  "[TLD_ngo]": 30989,
+  "[TLD_nhs.uk]": 30990,
+  "[TLD_ninja]": 30991,
+  "[TLD_nl]": 30992,
+  "[TLD_no]": 30993,
+  "[TLD_nrw]": 30994,
+  "[TLD_nu]": 30995,
+  "[TLD_nyc]": 30996,
+  "[TLD_nz]": 30997,
+  "[TLD_olsztyn.pl]": 30998,
+  "[TLD_one]": 30999,
+  "[TLD_onl]": 31000,
+  "[TLD_online]": 31001,
+  "[TLD_ooo]": 31002,
+  "[TLD_opole.pl]": 31003,
+  "[TLD_or.at]": 31004,
+  "[TLD_or.jp]": 31005,
+  "[TLD_org.ar]": 31007,
+  "[TLD_org.au]": 31008,
+  "[TLD_org.br]": 31009,
+  "[TLD_org.cn]": 31010,
+  "[TLD_org.co]": 31011,
+  "[TLD_org.es]": 31012,
+  "[TLD_org.hk]": 31013,
+  "[TLD_org.il]": 31014,
+  "[TLD_org.in]": 31015,
+  "[TLD_org.mx]": 31016,
+  "[TLD_org.my]": 31017,
+  "[TLD_org.nz]": 31018,
+  "[TLD_org.pe]": 31019,
+  "[TLD_org.ph]": 31020,
+  "[TLD_org.pk]": 31021,
+  "[TLD_org.pl]": 31022,
+  "[TLD_org.sg]": 31023,
+  "[TLD_org.tr]": 31024,
+  "[TLD_org.tw]": 31025,
+  "[TLD_org.ua]": 31026,
+  "[TLD_org.uk]": 31027,
+  "[TLD_org.za]": 31028,
+  "[TLD_org]": 31006,
+  "[TLD_organic]": 31029,
+  "[TLD_page]": 31030,
+  "[TLD_paris]": 31031,
+  "[TLD_partners]": 31032,
+  "[TLD_parts]": 31033,
+  "[TLD_party]": 31034,
+  "[TLD_pe]": 31035,
+  "[TLD_pet]": 31036,
+  "[TLD_ph]": 31037,
+  "[TLD_photo]": 31038,
+  "[TLD_photography]": 31039,
+  "[TLD_photos]": 31040,
+  "[TLD_pics]": 31041,
+  "[TLD_pictures]": 31042,
+  "[TLD_pink]": 31043,
+  "[TLD_pizza]": 31044,
+  "[TLD_pk]": 31045,
+  "[TLD_pl]": 31046,
+  "[TLD_place]": 31047,
+  "[TLD_plus]": 31048,
+  "[TLD_pm]": 31049,
+  "[TLD_poker]": 31050,
+  "[TLD_police.uk]": 31051,
+  "[TLD_porn]": 31052,
+  "[TLD_press]": 31053,
+  "[TLD_pro]": 31054,
+  "[TLD_promo]": 31055,
+  "[TLD_properties]": 31056,
+  "[TLD_property]": 31057,
+  "[TLD_ps]": 31058,
+  "[TLD_pt]": 31059,
+  "[TLD_pub]": 31060,
+  "[TLD_pw]": 31061,
+  "[TLD_qa]": 31062,
+  "[TLD_quest]": 31063,
+  "[TLD_re]": 31064,
+  "[TLD_recipes]": 31065,
+  "[TLD_red]": 31066,
+  "[TLD_rent]": 31067,
+  "[TLD_rentals]": 31068,
+  "[TLD_repair]": 31069,
+  "[TLD_report]": 31070,
+  "[TLD_rest]": 31071,
+  "[TLD_restaurant]": 31072,
+  "[TLD_review]": 31073,
+  "[TLD_reviews]": 31074,
+  "[TLD_rip]": 31075,
+  "[TLD_ro]": 31076,
+  "[TLD_rocks]": 31077,
+  "[TLD_rs]": 31078,
+  "[TLD_ru]": 31079,
+  "[TLD_run]": 31080,
+  "[TLD_rzeszow.pl]": 31081,
+  "[TLD_sa]": 31082,
+  "[TLD_sale]": 31083,
+  "[TLD_salon]": 31084,
+  "[TLD_sbs]": 31085,
+  "[TLD_sc]": 31086,
+  "[TLD_school]": 31087,
+  "[TLD_science]": 31088,
+  "[TLD_scot]": 31089,
+  "[TLD_se]": 31090,
+  "[TLD_services]": 31091,
+  "[TLD_sex]": 31092,
+  "[TLD_sexy]": 31093,
+  "[TLD_sg]": 31094,
+  "[TLD_sh]": 31095,
+  "[TLD_shoes]": 31096,
+  "[TLD_shop]": 31097,
+  "[TLD_shopping]": 31098,
+  "[TLD_show]": 31099,
+  "[TLD_si]": 31100,
+  "[TLD_singles]": 31101,
+  "[TLD_site]": 31102,
+  "[TLD_sk]": 31103,
+  "[TLD_ski]": 31104,
+  "[TLD_skin]": 31105,
+  "[TLD_sklep.pl]": 31106,
+  "[TLD_sn]": 31107,
+  "[TLD_so]": 31108,
+  "[TLD_soccer]": 31109,
+  "[TLD_social]": 31110,
+  "[TLD_software]": 31111,
+  "[TLD_solar]": 31112,
+  "[TLD_solutions]": 31113,
+  "[TLD_space]": 31114,
+  "[TLD_st]": 31115,
+  "[TLD_store]": 31116,
+  "[TLD_stream]": 31117,
+  "[TLD_studio]": 31118,
+  "[TLD_style]": 31119,
+  "[TLD_su]": 31120,
+  "[TLD_supplies]": 31121,
+  "[TLD_supply]": 31122,
+  "[TLD_support]": 31123,
+  "[TLD_surf]": 31124,
+  "[TLD_surgery]": 31125,
+  "[TLD_swiss]": 31126,
+  "[TLD_sx]": 31127,
+  "[TLD_systems]": 31128,
+  "[TLD_tax]": 31129,
+  "[TLD_taxi]": 31130,
+  "[TLD_tc]": 31131,
+  "[TLD_team]": 31132,
+  "[TLD_tech]": 31133,
+  "[TLD_technology]": 31134,
+  "[TLD_tel]": 31135,
+  "[TLD_tips]": 31136,
+  "[TLD_tires]": 31137,
+  "[TLD_tj]": 31138,
+  "[TLD_tk]": 31139,
+  "[TLD_tl]": 31140,
+  "[TLD_tm]": 31141,
+  "[TLD_tn]": 31142,
+  "[TLD_to]": 31143,
+  "[TLD_today]": 31144,
+  "[TLD_tokyo]": 31145,
+  "[TLD_tools]": 31146,
+  "[TLD_top]": 31147,
+  "[TLD_tours]": 31148,
+  "[TLD_town]": 31149,
+  "[TLD_toys]": 31150,
+  "[TLD_trade]": 31151,
+  "[TLD_training]": 31152,
+  "[TLD_travel]": 31153,
+  "[TLD_tube]": 31154,
+  "[TLD_tv]": 31155,
+  "[TLD_tw]": 31156,
+  "[TLD_ua]": 31157,
+  "[TLD_ug]": 31158,
+  "[TLD_uk]": 31159,
+  "[TLD_university]": 31160,
+  "[TLD_uno]": 31161,
+  "[TLD_us]": 31162,
+  "[TLD_uz]": 31163,
+  "[TLD_va]": 31164,
+  "[TLD_vacations]": 31165,
+  "[TLD_vc]": 31166,
+  "[TLD_vegas]": 31167,
+  "[TLD_ventures]": 31168,
+  "[TLD_vet]": 31169,
+  "[TLD_vg]": 31170,
+  "[TLD_video]": 31171,
+  "[TLD_vin]": 31172,
+  "[TLD_vip]": 31173,
+  "[TLD_vision]": 31174,
+  "[TLD_vn]": 31175,
+  "[TLD_voyage]": 31176,
+  "[TLD_vu]": 31177,
+  "[TLD_wales]": 31178,
+  "[TLD_wang]": 31179,
+  "[TLD_warszawa.pl]": 31180,
+  "[TLD_watch]": 31181,
+  "[TLD_waw.pl]": 31182,
+  "[TLD_website]": 31183,
+  "[TLD_wedding]": 31184,
+  "[TLD_wiki]": 31185,
+  "[TLD_win]": 31186,
+  "[TLD_wine]": 31187,
+  "[TLD_work]": 31188,
+  "[TLD_works]": 31189,
+  "[TLD_world]": 31190,
+  "[TLD_wroclaw.pl]": 31191,
+  "[TLD_ws]": 31192,
+  "[TLD_wtf]": 31193,
+  "[TLD_xn--3ds443g]": 31194,
+  "[TLD_xn--90ais]": 31195,
+  "[TLD_xn--c1avg]": 31196,
+  "[TLD_xn--p1ai]": 31197,
+  "[TLD_xn--tckwe]": 31198,
+  "[TLD_xxx]": 31199,
+  "[TLD_xyz]": 31200,
+  "[TLD_yachts]": 31201,
+  "[TLD_yoga]": 31202,
+  "[TLD_zone]": 31203
+}

config.json

@@ -0,0 +1,116 @@
+{
+  "architectures": [
+    "DualTaskTLDModel"
+  ],
+  "base_model_name": "google/bert_uncased_L-4_H-256_A-4",
+  "categorical_features": [],
+  "categorical_mappings": {},
+  "continuous_features": [
+    "abuse_rate_percent",
+    "avg_transfer_days",
+    "brand_association_strength",
+    "community_depth_score",
+    "dnssec_adoption_percent",
+    "four_letter_registration_percent",
+    "global_top_1m_share",
+    "hack_usage_popularity",
+    "influencer_adoption_rate",
+    "innovation_perception_score",
+    "iso_country_code",
+    "majestic_top_1m_count",
+    "market_momentum_score",
+    "media_sentiment_score",
+    "memorability_score",
+    "premium_brand_index",
+    "professional_usage_rate",
+    "registration_restrictions",
+    "registry_marketing_activity",
+    "reputation_trust_score",
+    "sales_10y_above_10_count",
+    "tech_startup_adoption_index",
+    "three_letter_registration_percent",
+    "tld_class"
+  ],
+  "economic_features": [
+    "gdp_total",
+    "healthcare_pharmaceuticals",
+    "banking_capital_markets",
+    "insurance",
+    "investment_wealth_management",
+    "education_edtech",
+    "retail_ecommerce",
+    "consumer_packaged_goods",
+    "food_beverage_restaurants",
+    "travel_tourism_hospitality",
+    "real_estate_proptech",
+    "automotive_mobility",
+    "technology_software",
+    "telecommunications_isps",
+    "energy_utilities",
+    "industrial_manufacturing_engineering",
+    "construction_infrastructure",
+    "logistics_shipping_transportation",
+    "media_entertainment_streaming",
+    "gaming_igaming",
+    "professional_legal_services"
+  ],
+  "embedding_dim": 96,
+  "feature_stats": {},
+  "mlp_dropout": 0.15,
+  "mlp_hidden_size": 192,
+  "model_type": "dual_task_tld",
+  "ordinal_features": [
+    "tld_class",
+    "registration_restrictions"
+  ],
+  "price_features": [
+    "overall_score",
+    "score_automotive",
+    "score_construction",
+    "score_education",
+    "score_energy",
+    "score_engineering",
+    "score_fashion",
+    "score_finance",
+    "score_food",
+    "score_gaming",
+    "score_healthcare",
+    "score_insurance",
+    "score_legal",
+    "score_media",
+    "score_music",
+    "score_pets",
+    "score_sports",
+    "score_technology"
+  ],
+  "research_features": [
+    "abuse_rate_percent",
+    "avg_transfer_days",
+    "brand_association_strength",
+    "community_depth_score",
+    "dnssec_adoption_percent",
+    "hack_usage_popularity",
+    "influencer_adoption_rate",
+    "innovation_perception_score",
+    "market_momentum_score",
+    "media_sentiment_score",
+    "memorability_score",
+    "premium_brand_index",
+    "professional_usage_rate",
+    "registration_restrictions",
+    "registry_marketing_activity",
+    "reputation_trust_score",
+    "tech_startup_adoption_index",
+    "tld_class"
+  ],
+  "technical_features": [
+    "four_letter_registration_percent",
+    "global_top_1m_share",
+    "majestic_top_1m_count",
+    "sales_10y_above_10_count",
+    "three_letter_registration_percent"
+  ],
+  "torch_dtype": "float32",
+  "transformers_version": "4.44.2",
+  "vocab_size": 31204
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd185cd13924e5a365b7f37644b32fc306f3fed20a8e14912f575e8790b16f70
+size 51066352