Upload all models and assets for ar (20251201)

.gitattributes

@@ -33,3 +33,9 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+visualizations/embedding_similarity.png filter=lfs diff=lfs merge=lfs -text
+visualizations/performance_dashboard.png filter=lfs diff=lfs merge=lfs -text
+visualizations/position_encoding_comparison.png filter=lfs diff=lfs merge=lfs -text
+visualizations/tsne_sentences.png filter=lfs diff=lfs merge=lfs -text
+visualizations/tsne_words.png filter=lfs diff=lfs merge=lfs -text
+visualizations/zipf_law.png filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,561 @@
+---
+language: ar
+language_name: Arabic
+language_family: arabic
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-arabic
+license: mit
+library_name: wikilangs
+pipeline_tag: feature-extraction
+datasets:
+  - omarkamali/wikipedia-monthly
+dataset_info:
+  name: wikipedia-monthly
+  description: Monthly snapshots of Wikipedia articles across 300+ languages
+metrics:
+  - name: best_compression_ratio
+    type: compression
+    value: 4.103
+  - name: best_isotropy
+    type: isotropy
+    value: 0.7155
+  - name: vocabulary_size
+    type: vocab
+    value: 1000000
+generated: 2025-12-27
+---
+
+# Arabic - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **Arabic** Wikipedia data.
+We analyze tokenizers, n-gram models, Markov chains, vocabulary statistics, and word embeddings.
+
+## 📋 Repository Contents
+
+### Models & Assets
+
+- Tokenizers (8k, 16k, 32k, 64k)
+- N-gram models (2, 3, 4-gram)
+- Markov chains (context of 1, 2, 3 and 4)
+- Subword N-gram and Markov chains
+- Embeddings in various sizes and dimensions
+- Language Vocabulary
+- Language Statistics
+![Performance Dashboard](visualizations/performance_dashboard.png)
+
+### Analysis and Evaluation
+
+- [1. Tokenizer Evaluation](#1-tokenizer-evaluation)
+- [2. N-gram Model Evaluation](#2-n-gram-model-evaluation)
+- [3. Markov Chain Evaluation](#3-markov-chain-evaluation)
+- [4. Vocabulary Analysis](#4-vocabulary-analysis)
+- [5. Word Embeddings Evaluation](#5-word-embeddings-evaluation)
+- [6. Summary & Recommendations](#6-summary--recommendations)
+- [Metrics Glossary](#appendix-metrics-glossary--interpretation-guide)
+- [Visualizations Index](#visualizations-index)
+
+---
+## 1. Tokenizer Evaluation
+
+![Tokenizer Compression](visualizations/tokenizer_compression.png)
+
+### Results
+
+| Vocab Size | Compression | Avg Token Len | UNK Rate | Total Tokens |
+|------------|-------------|---------------|----------|--------------|
+| **8k** | 3.156x | 3.13 | 0.0848% | 5,982,398 |
+| **16k** | 3.513x | 3.49 | 0.0944% | 5,374,291 |
+| **32k** | 3.837x | 3.81 | 0.1031% | 4,920,728 |
+| **64k** | 4.103x 🏆 | 4.07 | 0.1103% | 4,602,368 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `تحويل ميلفورد (كونيتيكت)`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁تحويل ▁ميل فورد ▁( ك وني تي كت )` | 9 |
+| 16k | `▁تحويل ▁ميل فورد ▁( ك وني تي كت )` | 9 |
+| 32k | `▁تحويل ▁ميل فورد ▁( كوني تيكت )` | 7 |
+| 64k | `▁تحويل ▁ميل فورد ▁( كونيتيكت )` | 6 |
+
+**Sample 2:** `قد يقصد من «الفرفار» :
+
+ الفرفار (إدا وكماض) : دوار تابع لجماعة إدا وڭماض في إقل...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁قد ▁يق صد ▁من ▁« الف رف ار » ▁: ... (+43 more)` | 53 |
+| 16k | `▁قد ▁يقصد ▁من ▁« الف رف ار » ▁: ▁الف ... (+37 more)` | 47 |
+| 32k | `▁قد ▁يقصد ▁من ▁« الف رف ار » ▁: ▁الف ... (+36 more)` | 46 |
+| 64k | `▁قد ▁يقصد ▁من ▁« الف رف ار » ▁: ▁الف ... (+34 more)` | 44 |
+
+**Sample 3:** `المراجع 
+
+تصنيف:أنهار إفريقية دولية
+تصنيف:أنهار بوروندي
+تصنيف:أنهار تنزانيا
+تصني...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁المراجع ▁تصنيف : أن هار ▁إ فريقية ▁دولية ▁تصنيف : ... (+16 more)` | 26 |
+| 16k | `▁المراجع ▁تصنيف : أنهار ▁إفريقية ▁دولية ▁تصنيف : أنهار ▁بور ... (+12 more)` | 22 |
+| 32k | `▁المراجع ▁تصنيف : أنهار ▁إفريقية ▁دولية ▁تصنيف : أنهار ▁بور ... (+9 more)` | 19 |
+| 64k | `▁المراجع ▁تصنيف : أنهار ▁إفريقية ▁دولية ▁تصنيف : أنهار ▁بوروندي ... (+8 more)` | 18 |
+
+
+### Key Findings
+
+- **Best Compression:** 64k achieves 4.103x compression
+- **Lowest UNK Rate:** 8k with 0.0848% unknown tokens
+- **Trade-off:** Larger vocabularies improve compression but increase model size
+- **Recommendation:** 32k vocabulary provides optimal balance for production use
+
+---
+## 2. N-gram Model Evaluation
+
+![N-gram Perplexity](visualizations/ngram_perplexity.png)
+
+![N-gram Coverage](visualizations/ngram_coverage.png)
+
+### Results
+
+| N-gram | Perplexity | Entropy | Unique N-grams | Top-100 Coverage | Top-1000 Coverage |
+|--------|------------|---------|----------------|------------------|-------------------|
+| **2-gram** | 224,018 🏆 | 17.77 | 6,245,473 | 10.5% | 22.3% |
+| **2-gram** | 514 🏆 | 9.00 | 52,884 | 52.6% | 94.6% |
+| **3-gram** | 831,530 | 19.67 | 14,344,223 | 6.6% | 16.3% |
+| **3-gram** | 4,885 | 12.25 | 487,957 | 23.0% | 53.9% |
+| **4-gram** | 1,784,666 | 20.77 | 25,822,600 | 4.6% | 13.6% |
+| **4-gram** | 29,916 | 14.87 | 3,376,435 | 13.3% | 31.6% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `تصنيف :` | 9,397,729 |
+| 2 | `ً ا` | 2,647,403 |
+| 3 | `: لاعبو` | 1,539,560 |
+| 4 | `| |` | 1,324,145 |
+| 5 | `كرة قدم` | 758,315 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `تصنيف : لاعبو` | 1,539,552 |
+| 2 | `تصنيف : مواليد` | 617,808 |
+| 3 | `: لاعبو كرة` | 498,223 |
+| 4 | `| | |` | 459,400 |
+| 5 | `تصنيف : أشخاص` | 441,938 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `تصنيف : لاعبو كرة` | 498,220 |
+| 2 | `: لاعبو كرة قدم` | 381,016 |
+| 3 | `القرن 20 تصنيف :` | 278,900 |
+| 4 | `في القرن 20 تصنيف` | 266,135 |
+| 5 | `| | | |` | 255,908 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 514
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~32% of corpus
+- **Recommendation:** 4-gram or 5-gram for best predictive performance
+
+---
+## 3. Markov Chain Evaluation
+
+![Markov Entropy](visualizations/markov_entropy.png)
+
+![Markov Branching](visualizations/markov_branching.png)
+
+### Results
+
+| Context | Avg Entropy | Perplexity | Branching Factor | Unique Contexts | Predictability |
+|---------|-------------|------------|------------------|-----------------|----------------|
+| **1** | 0.7411 | 1.671 | 12.81 | 5,367,543 | 25.9% |
+| **1** | 1.8418 | 3.585 | 17.44 | 13,038 | 0.0% |
+| **2** | 0.4074 | 1.326 | 2.70 | 68,744,585 | 59.3% |
+| **2** | 0.7015 | 1.626 | 5.08 | 227,339 | 29.9% |
+| **3** | 0.1748 | 1.129 | 1.44 | 185,531,332 | 82.5% |
+| **3** | 0.8426 | 1.793 | 5.22 | 1,153,787 | 15.7% |
+| **4** | 0.0757 🏆 | 1.054 | 1.16 | 267,713,644 | 92.4% |
+| **4** | 0.7645 🏆 | 1.699 | 3.88 | 6,025,353 | 23.6% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `. اكت ُ ص م َ ّ ة . أخته ، وافتتاح مشروع مرصد أونديجوف |`
+2. `في المدار في جمهورية ألمانيا تصنيف : خلافات في . أنظر : تشغيل الحواسيب . 1`
+3. `، واستحوذت أيضا التحريفية للبلغاريين والأجانب أو تمليح اللحوم والأتواب والملابس والبطانيات الى القاه...`
+
+**Context Size 2:**
+
+1. `تصنيف : كتاب ومؤلفو قصص مصورة تصنيف : فائزون بميداليات برونزية في ألعاب الكومنولث في إنجلترا تصنيف`
+2. `ً ا للاغتسال . وقال القرطبي في تفسيره على أنه آمن خلال الرضاعة الطبيعية يسبب زيادة الكوليسترول`
+3. `: لاعبو كرة قدم صرب مغتربون في روسيا تصنيف : أفلام دراما باللغة الإنجليزية تصنيف : سائقو`
+
+**Context Size 3:**
+
+1. `تصنيف : لاعبو بوتكيت ريد سوكس تصنيف : مواليد 1955 تصنيف : مؤيدون لتنظيم ملكية الأسلحة تصنيف :`
+2. `تصنيف : مواليد 1986 تصنيف : لاعبو وسط كرة قدم رجالية تصنيف : مواليد 1390 هـ تصنيف :`
+3. `: لاعبو كرة قدم مغاربة تصنيف : عداؤو مسافات متوسطة نيوزيلنديون تصنيف : مواليد 1981 تصنيف : مواليد`
+
+**Context Size 4:**
+
+1. `تصنيف : لاعبو كرة قدم مغتربون في المجر تصنيف : لاعبو كرة اليد في الألعاب الأولمبية الصيفية 1956 تصني...`
+2. `: لاعبو كرة قدم مغتربون في إنجلترا تصنيف : لاعبو كرة قدم مغتربون في إيطاليا تصنيف : أماكن مأهولة`
+3. `القرن 20 تصنيف : كاتبات أمريكيات في القرن 20 تصنيف : شعراء بالعربية في القرن 21 تصنيف : لاعبو`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 92.4% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (6,025,353 contexts)
+- **Recommendation:** Context-3 or Context-4 for text generation
+
+---
+## 4. Vocabulary Analysis
+
+![Zipf's Law](visualizations/zipf_law.png)
+
+![Top Words](visualizations/top20_words.png)
+
+![Coverage Curve](visualizations/vocab_coverage.png)
+
+### Statistics
+
+| Metric | Value |
+|--------|-------|
+| Vocabulary Size | 1,000,000 |
+| Total Tokens | 366,842,150 |
+| Mean Frequency | 366.84 |
+| Median Frequency | 12 |
+| Frequency Std Dev | 20900.79 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | في | 14,346,570 |
+| 2 | تصنيف | 9,437,038 |
+| 3 | من | 8,350,052 |
+| 4 | على | 3,295,037 |
+| 5 | ا | 2,755,855 |
+| 6 | إلى | 2,451,934 |
+| 7 | عام | 1,684,151 |
+| 8 | لاعبو | 1,540,822 |
+| 9 | أن | 1,441,897 |
+| 10 | مع | 1,171,753 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | твоим | 4 |
+| 2 | своему | 4 |
+| 3 | вашей | 4 |
+| 4 | нашу | 4 |
+| 5 | кого | 4 |
+| 6 | чьей | 4 |
+| 7 | работать | 4 |
+| 8 | говорит | 4 |
+| 9 | говорят | 4 |
+| 10 | идёт | 4 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 0.9655 |
+| R² (Goodness of Fit) | 0.990109 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 24.9% |
+| Top 1,000 | 48.1% |
+| Top 5,000 | 68.6% |
+| Top 10,000 | 76.6% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9901 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 24.9% of corpus
+- **Long Tail:** 990,000 words needed for remaining 23.4% coverage
+
+---
+## 5. Word Embeddings Evaluation
+
+![Embedding Isotropy](visualizations/embedding_isotropy.png)
+
+![Similarity Matrix](visualizations/embedding_similarity.png)
+
+![t-SNE Words](visualizations/tsne_words.png)
+
+![t-SNE Sentences](visualizations/tsne_sentences.png)
+
+### Model Comparison
+
+| Model | Vocab Size | Dimension | Avg Norm | Std Norm | Isotropy |
+|-------|------------|-----------|----------|----------|----------|
+| **mono_32d** | 1,505,991 | 32 | 3.562 | 1.491 | 0.7155 🏆 |
+| **mono_64d** | 1,505,991 | 64 | 3.899 | 1.405 | 0.7134 |
+| **mono_128d** | 1,505,991 | 128 | 4.337 | 1.358 | 0.6849 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.7155 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 1,505,991 words
+- **Recommendation:** 100d for balanced semantic capture and efficiency
+
+---
+## 6. Summary & Recommendations
+
+![Performance Dashboard](visualizations/performance_dashboard.png)
+
+### Production Recommendations
+
+| Component | Recommended | Rationale |
+|-----------|-------------|-----------|
+| Tokenizer | **32k BPE** | Best compression (4.10x) with low UNK rate |
+| N-gram | **5-gram** | Lowest perplexity (514) |
+| Markov | **Context-4** | Highest predictability (92.4%) |
+| Embeddings | **100d** | Balanced semantic capture and isotropy |
+
+---
+## Appendix: Metrics Glossary & Interpretation Guide
+
+This section provides definitions, intuitions, and guidance for interpreting the metrics used throughout this report.
+
+### Tokenizer Metrics
+
+**Compression Ratio**
+> *Definition:* The ratio of characters to tokens (chars/token). Measures how efficiently the tokenizer represents text.
+>
+> *Intuition:* Higher compression means fewer tokens needed to represent the same text, reducing sequence lengths for downstream models. A 3x compression means ~3 characters per token on average.
+>
+> *What to seek:* Higher is generally better for efficiency, but extremely high compression may indicate overly aggressive merging that loses morphological information.
+
+**Average Token Length (Fertility)**
+> *Definition:* Mean number of characters per token produced by the tokenizer.
+>
+> *Intuition:* Reflects the granularity of tokenization. Longer tokens capture more context but may struggle with rare words; shorter tokens are more flexible but increase sequence length.
+>
+> *What to seek:* Balance between 2-5 characters for most languages. Arabic/morphologically-rich languages may benefit from slightly longer tokens.
+
+**Unknown Token Rate (OOV Rate)**
+> *Definition:* Percentage of tokens that map to the unknown/UNK token, indicating words the tokenizer cannot represent.
+>
+> *Intuition:* Lower OOV means better vocabulary coverage. High OOV indicates the tokenizer encounters many unseen character sequences.
+>
+> *What to seek:* Below 1% is excellent; below 5% is acceptable. BPE tokenizers typically achieve very low OOV due to subword fallback.
+
+### N-gram Model Metrics
+
+**Perplexity**
+> *Definition:* Measures how "surprised" the model is by test data. Mathematically: 2^(cross-entropy). Lower values indicate better prediction.
+>
+> *Intuition:* If perplexity is 100, the model is as uncertain as if choosing uniformly among 100 options at each step. A perplexity of 10 means effectively choosing among 10 equally likely options.
+>
+> *What to seek:* Lower is better. Perplexity decreases with larger n-grams (more context). Values vary widely by language and corpus size.
+
+**Entropy**
+> *Definition:* Average information content (in bits) needed to encode the next token given the context. Related to perplexity: perplexity = 2^entropy.
+>
+> *Intuition:* High entropy means high uncertainty/randomness; low entropy means predictable patterns. Natural language typically has entropy between 1-4 bits per character.
+>
+> *What to seek:* Lower entropy indicates more predictable text patterns. Entropy should decrease as n-gram size increases.
+
+**Coverage (Top-K)**
+> *Definition:* Percentage of corpus occurrences explained by the top K most frequent n-grams.
+>
+> *Intuition:* High coverage with few patterns indicates repetitive/formulaic text; low coverage suggests diverse vocabulary usage.
+>
+> *What to seek:* Depends on use case. For language modeling, moderate coverage (40-60% with top-1000) is typical for natural text.
+
+### Markov Chain Metrics
+
+**Average Entropy**
+> *Definition:* Mean entropy across all contexts, measuring average uncertainty in next-word prediction.
+>
+> *Intuition:* Lower entropy means the model is more confident about what comes next. Context-1 has high entropy (many possible next words); Context-4 has low entropy (few likely continuations).
+>
+> *What to seek:* Decreasing entropy with larger context sizes. Very low entropy (<0.1) indicates highly deterministic transitions.
+
+**Branching Factor**
+> *Definition:* Average number of unique next tokens observed for each context.
+>
+> *Intuition:* High branching = many possible continuations (flexible but uncertain); low branching = few options (predictable but potentially repetitive).
+>
+> *What to seek:* Branching factor should decrease with context size. Values near 1.0 indicate nearly deterministic chains.
+
+**Predictability**
+> *Definition:* Derived metric: (1 - normalized_entropy) × 100%. Indicates how deterministic the model's predictions are.
+>
+> *Intuition:* 100% predictability means the next word is always certain; 0% means completely random. Real text falls between these extremes.
+>
+> *What to seek:* Higher predictability for text generation quality, but too high (>98%) may produce repetitive output.
+
+### Vocabulary & Zipf's Law Metrics
+
+**Zipf's Coefficient**
+> *Definition:* The slope of the log-log plot of word frequency vs. rank. Zipf's law predicts this should be approximately -1.
+>
+> *Intuition:* A coefficient near -1 indicates the corpus follows natural language patterns where a few words are very common and most words are rare.
+>
+> *What to seek:* Values between -0.8 and -1.2 indicate healthy natural language distribution. Deviations may suggest domain-specific or artificial text.
+
+**R² (Coefficient of Determination)**
+> *Definition:* Measures how well the linear fit explains the frequency-rank relationship. Ranges from 0 to 1.
+>
+> *Intuition:* R² near 1.0 means the data closely follows Zipf's law; lower values indicate deviation from expected word frequency patterns.
+>
+> *What to seek:* R² > 0.95 is excellent; > 0.99 indicates near-perfect Zipf adherence typical of large natural corpora.
+
+**Vocabulary Coverage**
+> *Definition:* Cumulative percentage of corpus tokens accounted for by the top N words.
+>
+> *Intuition:* Shows how concentrated word usage is. If top-100 words cover 50% of text, the corpus relies heavily on common words.
+>
+> *What to seek:* Top-100 covering 30-50% is typical. Higher coverage indicates more repetitive text; lower suggests richer vocabulary.
+
+### Word Embedding Metrics
+
+**Isotropy**
+> *Definition:* Measures how uniformly distributed vectors are in the embedding space. Computed as the ratio of minimum to maximum singular values.
+>
+> *Intuition:* High isotropy (near 1.0) means vectors spread evenly in all directions; low isotropy means vectors cluster in certain directions, reducing expressiveness.
+>
+> *What to seek:* Higher isotropy generally indicates better-quality embeddings. Values > 0.1 are reasonable; > 0.3 is good. Lower-dimensional embeddings tend to have higher isotropy.
+
+**Average Norm**
+> *Definition:* Mean magnitude (L2 norm) of word vectors in the embedding space.
+>
+> *Intuition:* Indicates the typical "length" of vectors. Consistent norms suggest stable training; high variance may indicate some words are undertrained.
+>
+> *What to seek:* Relatively consistent norms across models. The absolute value matters less than consistency (low std deviation).
+
+**Cosine Similarity**
+> *Definition:* Measures angular similarity between vectors, ranging from -1 (opposite) to 1 (identical direction).
+>
+> *Intuition:* Words with similar meanings should have high cosine similarity. This is the standard metric for semantic relatedness in embeddings.
+>
+> *What to seek:* Semantically related words should score > 0.5; unrelated words should be near 0. Synonyms often score > 0.7.
+
+**t-SNE Visualization**
+> *Definition:* t-Distributed Stochastic Neighbor Embedding - a dimensionality reduction technique that preserves local structure for visualization.
+>
+> *Intuition:* Clusters in t-SNE plots indicate groups of semantically related words. Spread indicates vocabulary diversity; tight clusters suggest semantic coherence.
+>
+> *What to seek:* Meaningful clusters (e.g., numbers together, verbs together). Avoid over-interpreting distances - t-SNE preserves local, not global, structure.
+
+### General Interpretation Guidelines
+
+1. **Compare within model families:** Metrics are most meaningful when comparing models of the same type (e.g., 8k vs 64k tokenizer).
+2. **Consider trade-offs:** Better performance on one metric often comes at the cost of another (e.g., compression vs. OOV rate).
+3. **Context matters:** Optimal values depend on downstream tasks. Text generation may prioritize different metrics than classification.
+4. **Corpus influence:** All metrics are influenced by corpus characteristics. Wikipedia text differs from social media or literature.
+5. **Language-specific patterns:** Morphologically rich languages (like Arabic) may show different optimal ranges than analytic languages.
+
+
+### Visualizations Index
+
+| Visualization | Description |
+|---------------|-------------|
+| Tokenizer Compression | Compression ratios by vocabulary size |
+| Tokenizer Fertility | Average token length by vocabulary |
+| Tokenizer OOV | Unknown token rates |
+| Tokenizer Total Tokens | Total tokens by vocabulary |
+| N-gram Perplexity | Perplexity by n-gram size |
+| N-gram Entropy | Entropy by n-gram size |
+| N-gram Coverage | Top pattern coverage |
+| N-gram Unique | Unique n-gram counts |
+| Markov Entropy | Entropy by context size |
+| Markov Branching | Branching factor by context |
+| Markov Contexts | Unique context counts |
+| Zipf's Law | Frequency-rank distribution with fit |
+| Vocab Frequency | Word frequency distribution |
+| Top 20 Words | Most frequent words |
+| Vocab Coverage | Cumulative coverage curve |
+| Embedding Isotropy | Vector space uniformity |
+| Embedding Norms | Vector magnitude distribution |
+| Embedding Similarity | Word similarity heatmap |
+| Nearest Neighbors | Similar words for key terms |
+| t-SNE Words | 2D word embedding visualization |
+| t-SNE Sentences | 2D sentence embedding visualization |
+| Position Encoding | Encoding method comparison |
+| Model Sizes | Storage requirements |
+| Performance Dashboard | Comprehensive performance overview |
+
+---
+## About This Project
+
+### Data Source
+
+Models trained on [wikipedia-monthly](https://huggingface.co/datasets/omarkamali/wikipedia-monthly) - a monthly snapshot of Wikipedia articles across 300+ languages.
+
+### Project
+
+A project by **[Wikilangs](https://wikilangs.org)** - Open-source NLP models for every Wikipedia language.
+
+### Maintainer
+
+[Omar Kamali](https://omarkamali.com) - [Omneity Labs](https://omneitylabs.com)
+
+### Citation
+
+If you use these models in your research, please cite:
+
+```bibtex
+@misc{wikilangs2025,
+  author = {Kamali, Omar},
+  title = {Wikilangs: Open NLP Models for Wikipedia Languages},
+  year = {2025},
+  publisher = {HuggingFace},
+  url = {https://huggingface.co/wikilangs}
+  institution = {Omneity Labs}
+}
+```
+
+### License
+
+MIT License - Free for academic and commercial use.
+
+### Links
+
+- 🌐 Website: [wikilangs.org](https://wikilangs.org)
+- 🤗 Models: [huggingface.co/wikilangs](https://huggingface.co/wikilangs)
+- 📊 Data: [wikipedia-monthly](https://huggingface.co/datasets/omarkamali/wikipedia-monthly)
+- 👤 Author: [Omar Kamali](https://huggingface.co/omarkamali)
+---
+*Generated by Wikilangs Models Pipeline*
+
+*Report Date: 2025-12-27 16:32:09*

models/embeddings/monolingual/ar_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a15e8d3f91894ef733d19288cf8f9905570f8b2cb6c96d55ef1706a098238e3
+size 2599525002

models/embeddings/monolingual/ar_128d.meta.json

@@ -0,0 +1 @@
+{"lang": "ar", "dim": 128, "max_seq_len": 512, "is_aligned": false}

models/embeddings/monolingual/ar_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ar",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 1505991
+}

models/embeddings/monolingual/ar_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c69754e00f4a7215bd624c3eada95e3f38a60dfc1f0e7cb54a049f1e7c7dfe5b
+size 674923914

models/embeddings/monolingual/ar_32d.meta.json

@@ -0,0 +1 @@
+{"lang": "ar", "dim": 32, "max_seq_len": 512, "is_aligned": false}

models/embeddings/monolingual/ar_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ar",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 1505991
+}

models/embeddings/monolingual/ar_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b216bcc9d5802fb5f2765d412ed8e269961ce1d314758d5c5986a12d8d067325
+size 1316457610

models/embeddings/monolingual/ar_64d.meta.json

@@ -0,0 +1 @@
+{"lang": "ar", "dim": 64, "max_seq_len": 512, "is_aligned": false}

models/embeddings/monolingual/ar_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ar",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 1505991
+}

models/subword_markov/ar_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f138ceea502bc716af3c32b8672e5560c86fb05241a4c6df25882580b5a54208
+size 1240058

models/subword_markov/ar_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "ar",
+  "unique_contexts": 13038,
+  "total_transitions": 2227602784
+}

models/subword_markov/ar_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3927d88cac7fc454b48e23919fdaacb6c000902069154a165316b426b04cb9d2
+size 9037901

models/subword_markov/ar_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "ar",
+  "unique_contexts": 227339,
+  "total_transitions": 2226268866
+}

models/subword_markov/ar_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:19911a0827041ed4348c506978baaae745ceac594b0d61c980d2bc3777aa6a97
+size 44692812

models/subword_markov/ar_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "ar",
+  "unique_contexts": 1153787,
+  "total_transitions": 2224934948
+}

models/subword_markov/ar_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67f755180f4df9e71b72cbe5cce476a0f75aa2d3bb2be4d74074ae69f6f89b19
+size 190825986

models/subword_markov/ar_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "ar",
+  "unique_contexts": 6025353,
+  "total_transitions": 2223601030
+}

models/subword_ngram/ar_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:71ed96bee2deb0bad70f5449c7c15403d96685f635321bae37b67c5fac7ce756
+size 750740

models/subword_ngram/ar_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "ar",
+  "unique_ngrams": 52884,
+  "total_ngrams": 2227602784
+}

models/subword_ngram/ar_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3bbdad1db2897a969c03e3f47d560f9d1dea5e8a3448c3f90f1c693656c1a2ff
+size 6329144

models/subword_ngram/ar_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "ar",
+  "unique_ngrams": 487957,
+  "total_ngrams": 2226268866
+}

models/subword_ngram/ar_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1d75bfcaa4ba765d4a698064ad06b8d77a93f04f7e84e5b96dc323009209b04
+size 43956365

models/subword_ngram/ar_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "ar",
+  "unique_ngrams": 3376435,
+  "total_ngrams": 2224934948
+}

models/tokenizer/ar_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7edaf2dbdc3ada01d30b8717ebf8add3d42cafa3d77bcdd80f720a97d9746d1
+size 559100

models/tokenizer/ar_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:442456b5ec25cb0683e0fe13294df9823fad435a42a56c7742c1cae081aa137a
+size 896676

models/tokenizer/ar_tokenizer_64k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:95c3e0d8a7a2ddcdb77379df9ab674034abf4e5f512785ebba09ecfc8b354f66
+size 1589031

models/tokenizer/ar_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f348b2c21a24d3a2ea09cf2eb93040bd9b203a25a891e426a51af0742693d90
+size 395404

models/vocabulary/ar_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a53a54e5478b4d7db731e85c634c1c05678e6956d42d16a8b8195a001195afbf
+size 15296577

models/vocabulary/ar_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "ar",
+  "vocabulary_size": 1000000,
+  "statistics": {
+    "type_token_ratio": 0.014399599085234752,
+    "coverage": {
+      "top_100": 0.24468957694891222,
+      "top_1000": 0.473562596402342,
+      "top_5000": 0.6748063946199147,
+      "top_10000": 0.7536829840670893
+    },
+    "hapax_count": 3360857,
+    "hapax_ratio": 0.6262799827295155,
+    "total_documents": 1333918
+  }
+}

models/word_markov/ar_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:595835c3288deb49535347878aaafaab9a18550487e54619d262b4d69c8215b8
+size 680761022

models/word_markov/ar_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "ar",
+  "unique_contexts": 5367543,
+  "total_transitions": 454840210
+}

models/word_markov/ar_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b3829aa63360c23d9ab12cd8bf7e141a172af6760dd613f90c50bc82f224475a
+size 2909790923

models/word_markov/ar_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "ar",
+  "unique_contexts": 68744585,
+  "total_transitions": 453506293
+}

models/word_markov/ar_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f834d8afb681b1fd6421bf196a28f7c9d985554a444560361588d8ffd31a584
+size 5786767054

models/word_markov/ar_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "ar",
+  "unique_contexts": 185531332,
+  "total_transitions": 452178131
+}

models/word_markov/ar_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:44384cee303cadec635f144b3cb696268a3650cc07f961bd6bcabe86f0407477
+size 7923726330

models/word_markov/ar_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "ar",
+  "unique_contexts": 267713644,
+  "total_transitions": 450864610
+}

models/word_ngram/ar_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bc9e09762e567e19160e345cf37f8110cad4c879ea76deb67d4bb98f71a0515a
+size 133530956

models/word_ngram/ar_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "ar",
+  "unique_ngrams": 6245473,
+  "total_ngrams": 454840210
+}

models/word_ngram/ar_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9784de16dd5fcc1ecb1cc4c96feff66ec0f2805b234fafb54d9732cfebafa123
+size 349991627

models/word_ngram/ar_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "ar",
+  "unique_ngrams": 14344223,
+  "total_ngrams": 453506293
+}

models/word_ngram/ar_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6a6e8a5896a9a380e2250a09454a912c564de9e3c16d392896346a8f97b61663
+size 674654663

models/word_ngram/ar_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "ar",
+  "unique_ngrams": 25822600,
+  "total_ngrams": 452178131
+}