Upload all models and assets for azb (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 +1,554 @@
+---
+language: azb
+language_name: AZB
+language_family: turkic_oghuz
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-turkic_oghuz
+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: 4.198
+  - name: best_isotropy
+    type: isotropy
+    value: 0.8242
+  - name: vocabulary_size
+    type: vocab
+    value: 317640
+generated: 2025-12-27
+---
+
+# AZB - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **AZB** 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** | 3.179x | 3.12 | 0.3165% | 385,493 |
+| **16k** | 3.558x | 3.49 | 0.3542% | 344,465 |
+| **32k** | 3.895x | 3.82 | 0.3877% | 314,660 |
+| **64k** | 4.198x 🏆 | 4.12 | 0.4179% | 291,946 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `ماقنتیتی ( ) روسیه اؤلکه‌سینده یئر آلان بیر کند دیر و مورمانسک اوبلاستیندا یئرلش...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁م اق ن تی تی ▁( ▁) ▁روسیه ▁اؤلکه ▁سینده ... (+19 more)` | 29 |
+| 16k | `▁ماق ن تی تی ▁( ▁) ▁روسیه ▁اؤلکه ▁سینده ▁یئر ... (+16 more)` | 26 |
+| 32k | `▁ماق نتی تی ▁( ▁) ▁روسیه ▁اؤلکه ▁سینده ▁یئر ▁آلان ... (+13 more)` | 23 |
+| 64k | `▁ماق نتی تی ▁( ▁) ▁روسیه ▁اؤلکه ▁سینده ▁یئر ▁آلان ... (+13 more)` | 23 |
+
+**Sample 2:** `سید ضیاء الدین طباطبائی یزدی (دوغوم:۱۲۸۶ شیراز -اؤلوم:۷ شهریور ۱۳۴۸ تهران) احمدش...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁سید ▁ض یا ء ▁الدین ▁ط با ط با ئی ... (+34 more)` | 44 |
+| 16k | `▁سید ▁ضیا ء ▁الدین ▁ط با ط با ئی ▁ی ... (+29 more)` | 39 |
+| 32k | `▁سید ▁ضیا ء ▁الدین ▁ط باط با ئی ▁یز دی ... (+26 more)` | 36 |
+| 64k | `▁سید ▁ضیاء ▁الدین ▁طباطبا ئی ▁یزدی ▁( دوغوم : ۱۲ ... (+21 more)` | 31 |
+
+**Sample 3:** `ویکسن (اینگیلیسجه: Vixen, Louisiana) آمریکانین لوئیزیانا ایالتینده یئرلشن بیر یا...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ویک سن ▁( اینگیلیسجه : ▁v ix en , ▁louisiana ... (+15 more)` | 25 |
+| 16k | `▁ویک سن ▁( اینگیلیسجه : ▁v ix en , ▁louisiana ... (+15 more)` | 25 |
+| 32k | `▁ویک سن ▁( اینگیلیسجه : ▁v ix en , ▁louisiana ... (+15 more)` | 25 |
+| 64k | `▁ویک سن ▁( اینگیلیسجه : ▁vixen , ▁louisiana ) ▁آمریکانین ... (+13 more)` | 23 |
+
+
+### Key Findings
+
+- **Best Compression:** 64k achieves 4.198x compression
+- **Lowest UNK Rate:** 8k with 0.3165% 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** | 7,645 🏆 | 12.90 | 268,529 | 30.2% | 57.7% |
+| **2-gram** | 655 🏆 | 9.35 | 14,131 | 46.5% | 93.7% |
+| **3-gram** | 13,711 | 13.74 | 512,828 | 24.9% | 52.4% |
+| **3-gram** | 4,808 | 12.23 | 141,425 | 21.1% | 58.1% |
+| **4-gram** | 23,522 | 14.52 | 979,944 | 21.9% | 47.2% |
+| **4-gram** | 19,536 | 14.25 | 804,296 | 13.9% | 41.7% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `بؤلمه :` | 391,479 |
+| 2 | `او ْ` | 139,043 |
+| 3 | `‌ لار` | 137,076 |
+| 4 | `قایناقلار بؤلمه` | 125,059 |
+| 5 | `‌ نین` | 118,471 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `قایناقلار بؤلمه :` | 125,058 |
+| 2 | `( اینگیلیسجه :` | 112,489 |
+| 3 | `. قایناقلار بؤلمه` | 97,901 |
+| 4 | `قایناق ‌ لار` | 92,555 |
+| 5 | `ایشلدنلری طرفیندن یارانمیش` | 76,848 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `. قایناقلار بؤلمه :` | 97,901 |
+| 2 | `ایشلدنلری طرفیندن یارانمیش «` | 76,839 |
+| 3 | `ْ خلانیلیبدیر ) .` | 76,827 |
+| 4 | `» ، مقاله ‌` | 76,824 |
+| 5 | `، مقاله ‌ سیندن` | 76,824 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 655
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~42% 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** | 0.4591 | 1.375 | 4.28 | 1,005,885 | 54.1% |
+| **1** | 1.2059 | 2.307 | 10.74 | 2,934 | 0.0% |
+| **2** | 0.2308 | 1.173 | 1.71 | 4,302,287 | 76.9% |
+| **2** | 1.1061 | 2.153 | 8.32 | 31,498 | 0.0% |
+| **3** | 0.1101 | 1.079 | 1.28 | 7,369,809 | 89.0% |
+| **3** | 0.9393 | 1.918 | 5.09 | 262,151 | 6.1% |
+| **4** | 0.0637 🏆 | 1.045 | 1.14 | 9,387,722 | 93.6% |
+| **4** | 0.7222 🏆 | 1.650 | 3.21 | 1,333,406 | 27.8% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `‌ ایللیین اووللرین ده یاشایان اینسانلار بؤلمه : bhumibol adulyadej crown , footballer , alchimie du`
+2. `. قان معائناتی و زنگین ‌ نین ایشلدنلری طرفیندن یارانمیش « panama 2 episodes ایلآدی رو`
+3. `: sergey ryzhikov ( ۱۹۴۸ ایلینده گونئی کارولینا ( تحصیلات ) رومنی ( لهیستانجا : 2`
+
+**Context Size 2:**
+
+1. `بؤلمه : وورت بؤلگه ‌ سیاراک بؤلگه ‌ سی ‌ دیر گؤرونتولر قایناق ‌ لار اینگیلیسجه ویکی`
+2. `او ْ یونچو . ۲۶ ژوئن ۱۹۵۷ میلادی تاریخیندا وفات ائدیب . قایناق ‌ لار اینگیلیسجه ویکی`
+3. `‌ لار بؤلمه : آمریکا شهرلری بؤلمه : گرینزبورو دوغوملولار بؤلمه : ۲۰۱۹ - جو ایلینده افغانیستاندا`
+
+**Context Size 3:**
+
+1. `قایناقلار بؤلمه : هیندوستان کندلری en : kahra`
+2. `( اینگیلیسجه : louis prima ) آمریکالی موغنی ، یازیچی و او ْ ندان ایستیفاده ائتمیشدیر . روبنسین`
+3. `. قایناقلار بؤلمه : هیندوستان کندلری en : kalawad`
+
+**Context Size 4:**
+
+1. `. قایناقلار بؤلمه : هیندوستان کندلری en : balowal`
+2. `ایشلدنلری طرفیندن یارانمیش « crabtree ' s catalyst » ، مقاله ‌ سیندن گؤتورولوبدور . ( ۷ سپتامبر ۲۰۱۷`
+3. `، مقاله ‌ سیندن گؤتورولوبدور . ( ۱۲ ژوئن ۲۰۱۸ تاریخینده یو ْ خلانیلیبدیر ) . بؤلمه : ایستادیوملار`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 93.6% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (1,333,406 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 | 317,640 |
+| Total Tokens | 16,697,106 |
+| Mean Frequency | 52.57 |
+| Median Frequency | 3 |
+| Frequency Std Dev | 1499.80 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | بؤلمه | 391,840 |
+| 2 | و | 291,285 |
+| 3 | اینگیلیسجه | 188,551 |
+| 4 | بیر | 169,796 |
+| 5 | او | 145,694 |
+| 6 | قایناقلار | 143,368 |
+| 7 | سی | 139,966 |
+| 8 | لار | 138,388 |
+| 9 | دیر | 132,038 |
+| 10 | the | 130,277 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | سئندن | 2 |
+| 2 | ائششکین | 2 |
+| 3 | onager | 2 |
+| 4 | داشناکلارلا | 2 |
+| 5 | قۇلان | 2 |
+| 6 | آسینۇس | 2 |
+| 7 | هئمیو | 2 |
+| 8 | تاپؽلمیشدیر | 2 |
+| 9 | kulan | 2 |
+| 10 | کسا | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 1.1842 |
+| R² (Goodness of Fit) | 0.995333 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 36.8% |
+| Top 1,000 | 66.7% |
+| Top 5,000 | 81.2% |
+| Top 10,000 | 85.9% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9953 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 36.8% of corpus
+- **Long Tail:** 307,640 words needed for remaining 14.1% 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** | 134,712 | 32 | 5.098 | 1.243 | 0.8242 🏆 |
+| **mono_64d** | 134,712 | 64 | 5.525 | 1.165 | 0.7928 |
+| **mono_128d** | 134,712 | 128 | 6.056 | 1.052 | 0.7520 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.8242 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 134,712 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 (4.20x) with low UNK rate |
+| N-gram | **5-gram** | Lowest perplexity (655) |
+| Markov | **Context-4** | Highest predictability (93.6%) |
+| 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 22:57:16*

models/embeddings/monolingual/azb_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e36e34c4aa5231270647ed20effcf7427bf66e34bf81f236183cc4d2910eef2d
+size 1164785458

models/embeddings/monolingual/azb_128d.meta.json

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

models/embeddings/monolingual/azb_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "azb",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 134712
+}

models/embeddings/monolingual/azb_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:24dbd964d896595164ff03aa344de312318ab9024fadf1dda4173202ae37885b
+size 293326642

models/embeddings/monolingual/azb_32d.meta.json

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

models/embeddings/monolingual/azb_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "azb",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 134712
+}

models/embeddings/monolingual/azb_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6193385dacda8dfbeee9fb60f5514bc20a74571e0a03c1d6372443118dd7271a
+size 583812914

models/embeddings/monolingual/azb_64d.meta.json

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

models/embeddings/monolingual/azb_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "azb",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 134712
+}

models/subword_markov/azb_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b5b4c16c0cc262c904b79b490c2cb1ed8fa1b624111ee815c8a153b6470de540
+size 249202

models/subword_markov/azb_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "azb",
+  "unique_contexts": 2934,
+  "total_transitions": 117347430
+}

models/subword_markov/azb_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bb9b0eb974612ded94c2843ab693fd0d8bae85b9e9719cc399abf8806f0d8e6
+size 2049903

models/subword_markov/azb_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "azb",
+  "unique_contexts": 31498,
+  "total_transitions": 117102967
+}

models/subword_markov/azb_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:364a895a59b1416e468680327f44afdb5268c191c6d191e6b325691c0f3c3920
+size 10556130

models/subword_markov/azb_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "azb",
+  "unique_contexts": 262151,
+  "total_transitions": 116858504
+}

models/subword_markov/azb_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6bbc3fbabaaf693d43117cfaccb8266ae07186aa1c883be3dd208a2f9aad8724
+size 35507016

models/subword_markov/azb_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "azb",
+  "unique_contexts": 1333406,
+  "total_transitions": 116614041
+}

models/subword_ngram/azb_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ccea8a8d09e5e2c50495929c3ed6209d6250e502dbf821570083b812edd44ab9
+size 198369

models/subword_ngram/azb_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "azb",
+  "unique_ngrams": 14131,
+  "total_ngrams": 117347430
+}

models/subword_ngram/azb_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e3575791a79055ab8760da4b584e9186ebd76add76ef9131f8b1c690f6f619c
+size 1733384

models/subword_ngram/azb_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "azb",
+  "unique_ngrams": 141425,
+  "total_ngrams": 117102967
+}

models/subword_ngram/azb_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2d7bf39c2289a572f7d6d41c6ee5b70e899bc025d9c7d73b34bf4e38cac70f4
+size 9977940

models/subword_ngram/azb_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "azb",
+  "unique_ngrams": 804296,
+  "total_ngrams": 116858504
+}

models/tokenizer/azb_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3884a029fd3957ce39121cce73cb5a8166d611ce94cef5675b43779699c0bd8a
+size 524898

models/tokenizer/azb_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7da271a7e77b591daaf850d5e0ca617ac859c9d9797bfc37fd65791b5cfc15fe
+size 819972

models/tokenizer/azb_tokenizer_64k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54cad8dd709976bcc93887547846baf55cbc48f7161ddab74a438fd09acb2685
+size 1430728

models/tokenizer/azb_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2ecb3b8a28e5f01ac0a0633151869481ce9b8a3af8ad9c54007045e09589ce83
+size 382716

models/vocabulary/azb_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c0ed41ec149e3751f6a1ace2372fcee3d368b53c6355b2765ae7b98caed8924
+size 4872673

models/vocabulary/azb_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "azb",
+  "vocabulary_size": 317640,
+  "statistics": {
+    "type_token_ratio": 0.05786680977434657,
+    "coverage": {
+      "top_100": 0.35375568505036664,
+      "top_1000": 0.6410291673928461,
+      "top_5000": 0.7797111502624887,
+      "top_10000": 0.8250270483868463
+    },
+    "hapax_count": 688404,
+    "hapax_ratio": 0.6842682825005666,
+    "total_documents": 244463
+  }
+}

models/word_markov/azb_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f8d5ae87507c80aafd715379fb38b0179c5049568d6f9e2b033dffefdfb0395
+size 49308871

models/word_markov/azb_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "azb",
+  "unique_contexts": 1005885,
+  "total_transitions": 23557605
+}

models/word_markov/azb_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c8a73136462b690d7c5514e30b6f6b274d36f9f121fbafeb5e99c8af7ec34c3
+size 115872504

models/word_markov/azb_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "azb",
+  "unique_contexts": 4302287,
+  "total_transitions": 23313143
+}

models/word_markov/azb_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acd2b58441a2312fe6aff4e924e858f5d85a5843063ac2adf427f032e6628d22
+size 173864269

models/word_markov/azb_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "azb",
+  "unique_contexts": 7369809,
+  "total_transitions": 23068685
+}

models/word_markov/azb_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:849a5c01ecae01be263fc6446cd008e4ee8cd861a0e83cd8e5da46a5f178d5ae
+size 222555153

models/word_markov/azb_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "azb",
+  "unique_contexts": 9387722,
+  "total_transitions": 22824244
+}

models/word_ngram/azb_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d9b9580f5793b668288ab3a7773338440486cda750c383b557c68ac0bc9221f
+size 4287214

models/word_ngram/azb_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "azb",
+  "unique_ngrams": 268529,
+  "total_ngrams": 23557605
+}

models/word_ngram/azb_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec70d427e623958a19690dad68f77970e4b6ff77368e7285059f2db52e46e7e6
+size 9355723

models/word_ngram/azb_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "azb",
+  "unique_ngrams": 512828,
+  "total_ngrams": 23313143
+}

models/word_ngram/azb_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b30786e6c43d7ce3c871df1db1b870820ce26c16155e9f336633c143d3099b5
+size 19762130

models/word_ngram/azb_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "azb",
+  "unique_ngrams": 979944,
+  "total_ngrams": 23068685
+}