Upload all models and assets for ab (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 +1,562 @@
+---
+language: ab
+language_name: AB
+language_family: caucasian_northwest
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-caucasian_northwest
+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: 4.203
+  - name: best_isotropy
+    type: isotropy
+    value: 0.8443
+  - name: vocabulary_size
+    type: vocab
+    value: 34914
+generated: 2025-12-27
+---
+
+# AB - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **AB** 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** | 3.211x | 3.15 | 0.1756% | 257,918 |
+| **16k** | 3.553x | 3.49 | 0.1943% | 233,133 |
+| **32k** | 3.880x | 3.81 | 0.2122% | 213,462 |
+| **64k** | 4.203x 🏆 | 4.13 | 0.2299% | 197,072 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `Ѫ, ѫ — кириллтәи аҩыратә архаикатә иажәхьоу нбан. 
+
+Азхьарԥшқәа
+ Graphemica (Ѫ)
+...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ ѫ , ▁ ѫ ▁— ▁кириллтәи ▁аҩыратә ▁архаикатә ▁иажәхьоу ... (+11 more)` | 21 |
+| 16k | `▁ ѫ , ▁ ѫ ▁— ▁кириллтәи ▁аҩыратә ▁архаикатә ▁иажәхьоу ... (+11 more)` | 21 |
+| 32k | `▁ ѫ , ▁ ѫ ▁— ▁кириллтәи ▁аҩыратә ▁архаикатә ▁иажәхьоу ... (+11 more)` | 21 |
+| 64k | `▁ ѫ , ▁ ѫ ▁— ▁кириллтәи ▁аҩыратә ▁архаикатә ▁иажәхьоу ... (+11 more)` | 21 |
+
+**Sample 2:** `Аби́а () — ҵиаа. Ашәыр. Ашәырҵла.
+
+Ахьарԥшқәа
+
+б`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁аби ́ а ▁() ▁— ▁ҵиаа . ▁ашәыр . ▁аш ... (+5 more)` | 15 |
+| 16k | `▁аби ́ а ▁() ▁— ▁ҵиаа . ▁ашәыр . ▁ашәырҵ ... (+4 more)` | 14 |
+| 32k | `▁аби ́а ▁() ▁— ▁ҵиаа . ▁ашәыр . ▁ашәырҵла . ... (+2 more)` | 12 |
+| 64k | `▁аби ́а ▁() ▁— ▁ҵиаа . ▁ашәыр . ▁ашәырҵла . ... (+2 more)` | 12 |
+
+**Sample 3:** `Ҝ, ҝ — кириллтәи аҩыратә нбан.`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ ҝ , ▁ ҝ ▁— ▁кириллтәи ▁аҩыратә ▁нбан .` | 10 |
+| 16k | `▁ ҝ , ▁ ҝ ▁— ▁кириллтәи ▁аҩыратә ▁нбан .` | 10 |
+| 32k | `▁ ҝ , ▁ ҝ ▁— ▁кириллтәи ▁аҩыратә ▁нбан .` | 10 |
+| 64k | `▁ ҝ , ▁ ҝ ▁— ▁кириллтәи ▁аҩыратә ▁нбан .` | 10 |
+
+
+### Key Findings
+
+- **Best Compression:** 64k achieves 4.203x compression
+- **Lowest UNK Rate:** 8k with 0.1756% 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** | 2,750 🏆 | 11.43 | 13,494 | 35.3% | 57.9% |
+| **2-gram** | 464 🏆 | 8.86 | 5,850 | 56.1% | 94.4% |
+| **3-gram** | 2,460 | 11.26 | 16,782 | 38.6% | 56.9% |
+| **3-gram** | 3,385 | 11.72 | 40,776 | 25.5% | 64.3% |
+| **4-gram** | 3,267 | 11.67 | 27,732 | 37.4% | 51.5% |
+| **4-gram** | 13,192 | 13.69 | 145,474 | 16.1% | 43.3% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `акатегориа :` | 5,231 |
+| 2 | `рыԥсҭазаара иалҵит` | 3,971 |
+| 3 | `иит рыԥсҭазаара` | 3,938 |
+| 4 | `нанҳәамза цәыббрамза` | 3,601 |
+| 5 | `жәабранмза хәажәкырамза` | 3,601 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `иит рыԥсҭазаара иалҵит` | 3,938 |
+| 2 | `ажьырныҳәамза жәабранмза хәажәкырамза` | 3,601 |
+| 3 | `жьҭаарамза абҵарамза ԥхынҷкәынмза` | 3,601 |
+| 4 | `мшаԥымза лаҵарамза рашәарамза` | 3,601 |
+| 5 | `ԥхынгәымза нанҳәамза цәыббрамза` | 3,601 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `лаҵарамза рашәарамза ԥхынгәымза нанҳәамза` | 3,601 |
+| 2 | `рашәарамза ԥхынгәымза нанҳәамза цәыббрамза` | 3,601 |
+| 3 | `ахҭысқəа ажьырныҳәамза жәабранмза хәажәкырамза` | 3,601 |
+| 4 | `мшаԥымза лаҵарамза рашәарамза ԥхынгәымза` | 3,601 |
+| 5 | `ԥхынгәымза нанҳәамза цәыббрамза жьҭаарамза` | 3,601 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 464
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~43% 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** | 0.5772 | 1.492 | 3.62 | 99,604 | 42.3% |
+| **1** | 1.5567 | 2.942 | 13.88 | 876 | 0.0% |
+| **2** | 0.1878 | 1.139 | 1.43 | 360,470 | 81.2% |
+| **2** | 1.2241 | 2.336 | 6.90 | 12,157 | 0.0% |
+| **3** | 0.0635 | 1.045 | 1.11 | 515,280 | 93.6% |
+| **3** | 0.7258 | 1.654 | 3.34 | 83,923 | 27.4% |
+| **4** | 0.0257 🏆 | 1.018 | 1.04 | 573,219 | 97.4% |
+| **4** | 0.4863 🏆 | 1.401 | 2.16 | 280,678 | 51.4% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `, аил - маклаи ихьӡ зху аҟәатәи аҳәынҭқарратә педагогтә институт . кёльн - рико ) ,`
+2. `. алитература ахырхарҭала . уи азҵаара азыҳәан қьалышь - 1528 ашықәсқәа рзы агазет « titus andronicu...`
+3. `- зшәышықәса агьама змоу акоуп азеипш гәабзиарахьчара аусхк аҿы ԥаҵаду ҳәа иашьҭан . акатегориа : в`
+
+**Context Size 2:**
+
+1. `акатегориа : аԥсны аиҭагаҩцәа акатегориа : аҩада — атерриториа атерриториа – . ақалақьқәа ақалақь га...`
+2. `иит рыԥсҭазаара иалҵит : друз иулии цезарь – германики агриппинәи рԥа ( дыԥсит ? ? ) азхьарԥшқәа`
+3. `мшаԥымза лаҵарамза рашәарамза ԥхынгәымза нанҳәамза цәыббрамза жьҭаарамза абҵарамза ԥхынҷкәынмза иит ...`
+
+**Context Size 3:**
+
+1. `мшаԥымза лаҵарамза рашәарамза ԥхынгәымза нанҳәамза цәыббрамза жьҭаарамза абҵарамза ԥхынҷкәынмза иит ...`
+2. `рашәарамза ԥхынгәымза нанҳәамза цәыббрамза жьҭаарамза абҵарамза ԥхынҷкәынмза иит рыԥсҭазаара иалҵит ...`
+3. `жьҭаарамза абҵарамза ԥхынҷкәынмза иит рыԥсҭазаара иалҵит арминии – германиатә херуски аимшьҭра рхада...`
+
+**Context Size 4:**
+
+1. `мшаԥымза лаҵарамза рашәарамза ԥхынгәымза нанҳәамза цәыббрамза жьҭаарамза абҵарамза ԥхынҷкәынмза иит ...`
+2. `нанҳәамза цәыббрамза жьҭаарамза абҵарамза ԥхынҷкәынмза иит рыԥсҭазаара иалҵит арминии – германиатә х...`
+3. `цәыббрамза жьҭаарамза абҵарамза ԥхынҷкәынмза иит рыԥсҭазаара иалҵит арминии – германиатә херуски аим...`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 97.4% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (280,678 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 | 34,914 |
+| Total Tokens | 483,415 |
+| Mean Frequency | 13.85 |
+| Median Frequency | 3 |
+| Frequency Std Dev | 106.12 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | акатегориа | 5,263 |
+| 2 | уи | 4,164 |
+| 3 | рыԥсҭазаара | 4,025 |
+| 4 | иит | 3,987 |
+| 5 | иалҵит | 3,980 |
+| 6 | лаҵарамза | 3,888 |
+| 7 | жәабранмза | 3,837 |
+| 8 | хәажәкырамза | 3,833 |
+| 9 | ԥхынҷкәынмза | 3,805 |
+| 10 | абҵарамза | 3,804 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | адрес | 2 |
+| 2 | extended | 2 |
+| 3 | stream | 2 |
+| 4 | block | 2 |
+| 5 | stru | 2 |
+| 6 | compressed | 2 |
+| 7 | draft | 2 |
+| 8 | preston | 2 |
+| 9 | видеохәмарроуп | 2 |
+| 10 | авидеохәмаррақәа | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 0.9724 |
+| R² (Goodness of Fit) | 0.994461 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 30.1% |
+| Top 1,000 | 55.4% |
+| Top 5,000 | 76.6% |
+| Top 10,000 | 85.3% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9945 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 30.1% of corpus
+- **Long Tail:** 24,914 words needed for remaining 14.7% 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** | 12,418 | 32 | 3.919 | 0.892 | 0.8443 🏆 |
+| **mono_64d** | 12,418 | 64 | 4.225 | 0.826 | 0.5913 |
+| **mono_128d** | 12,418 | 128 | 4.285 | 0.827 | 0.1726 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.8443 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 12,418 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 (464) |
+| Markov | **Context-4** | Highest predictability (97.4%) |
+| 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 04:31:24*

models/embeddings/monolingual/ab_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:70a5bb9264c0f53018d482c2616f76c0bf9699ff0b3215c919337610e18f88a9
+size 1037017106

models/embeddings/monolingual/ab_128d.meta.json

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

models/embeddings/monolingual/ab_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ab",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 12418
+}

models/embeddings/monolingual/ab_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f4040b5f28ef7e1b3cf96524fad4c6f4249e5107c0f78bdd6a3133c867ead2aa
+size 259480082

models/embeddings/monolingual/ab_32d.meta.json

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

models/embeddings/monolingual/ab_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ab",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 12418
+}

models/embeddings/monolingual/ab_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59bba3121b8000d938989dc2a115057a7fd481c440ecd2c4af0c5f7711a53e1d
+size 518659090

models/embeddings/monolingual/ab_64d.meta.json

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

models/embeddings/monolingual/ab_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ab",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 12418
+}

models/subword_markov/ab_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5b0a600b840e3db1ee3806d9be588a8213f99ef3cce2c1cd544318a9585f986
+size 93410

models/subword_markov/ab_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "ab",
+  "unique_contexts": 876,
+  "total_transitions": 4575952
+}

models/subword_markov/ab_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ea56c5cf197c6e7366129c2324fad6e8e710aca5379b47abfec22e6edd82fa6
+size 586863

models/subword_markov/ab_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "ab",
+  "unique_contexts": 12157,
+  "total_transitions": 4569421
+}

models/subword_markov/ab_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d3626fdfb84a1b24e4b45c30124c54337aabdb6a8585a36031b40d4b2aea4c4
+size 2014070

models/subword_markov/ab_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "ab",
+  "unique_contexts": 83923,
+  "total_transitions": 4562890
+}

models/subword_markov/ab_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:142c381e03e008f540daba013f253cbcc8bcf7cfe2a7bac360d368f678c4c000
+size 5510132

models/subword_markov/ab_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "ab",
+  "unique_contexts": 280678,
+  "total_transitions": 4556359
+}

models/subword_ngram/ab_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:25a3ad73144d6ab005e2dfb7deb33bbc6f3db63a36d07ff42172d7b2be73b608
+size 72236

models/subword_ngram/ab_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "ab",
+  "unique_ngrams": 5850,
+  "total_ngrams": 4575952
+}

models/subword_ngram/ab_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e7da8578639b16b8512cca6705fe65f303df54fe9d9e019e89159e0e2b7f2ff5
+size 516357

models/subword_ngram/ab_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "ab",
+  "unique_ngrams": 40776,
+  "total_ngrams": 4569421
+}

models/subword_ngram/ab_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e898894c24a5ef426ee5c2cd23250aabe7e24d18ae03c6bbee902c09cf9ec33
+size 1774272

models/subword_ngram/ab_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "ab",
+  "unique_ngrams": 145474,
+  "total_ngrams": 4562890
+}

models/tokenizer/ab_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b92a09ffb9d9e16d3cfcd9d14625e37f4a5f64b96aa51c52d35d12a333ad633b
+size 582079

models/tokenizer/ab_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5030df6b09c2e503b39eb01d7fa0fccf197e83ae7fdb9fd7b16866de09bababe
+size 937553

models/tokenizer/ab_tokenizer_64k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:049440a566c70344456ff4c82202e7e102f4915ace284706ad7cc4c5fc73dba9
+size 1697167

models/tokenizer/ab_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47045e15113d3f3907abf69824acfc10797a5b6e9942db1f6c5d4990df2dd043
+size 405395

models/vocabulary/ab_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b85dbc4e6a7540ba1ff9dfa0ac21d279d0fce68b4180e8d6df6b29deb668b84
+size 683448

models/vocabulary/ab_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "ab",
+  "vocabulary_size": 34914,
+  "statistics": {
+    "type_token_ratio": 0.18177207522936784,
+    "coverage": {
+      "top_100": 0.2651855284354094,
+      "top_1000": 0.48867710079779325,
+      "top_5000": 0.6758529345765748,
+      "top_10000": 0.7518667778310897
+    },
+    "hapax_count": 64722,
+    "hapax_ratio": 0.649584487534626,
+    "total_documents": 6531
+  }
+}

models/word_markov/ab_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1bf752e561b655e36306643fcd5f9ee780835f3c8ab87825116a8028fcef88a
+size 4988023

models/word_markov/ab_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "ab",
+  "unique_contexts": 99604,
+  "total_transitions": 715219
+}

models/word_markov/ab_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f00af198c4f86326bb4f959b22443089a350d189a2ac8d5415a26f3dcfb7f220
+size 10940272

models/word_markov/ab_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "ab",
+  "unique_contexts": 360470,
+  "total_transitions": 708688
+}

models/word_markov/ab_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7bac1826a063cba6bbd8f61df21c4be2dba68dac0abfc311b80cd598e258a68e
+size 14657425

models/word_markov/ab_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "ab",
+  "unique_contexts": 515280,
+  "total_transitions": 702160
+}

models/word_markov/ab_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17958e76510075d2200f79a9911c3d1e9a3f4d0a87cd2a4cc0331afca431b300
+size 17354156

models/word_markov/ab_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "ab",
+  "unique_contexts": 573219,
+  "total_transitions": 695636
+}

models/word_ngram/ab_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:acdfd16e121b263288a429d93a717f252969d4848648d80287275db5e16edc95
+size 282764

models/word_ngram/ab_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "ab",
+  "unique_ngrams": 13494,
+  "total_ngrams": 715219
+}

models/word_ngram/ab_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dab1f98b0cb7bd2d28f99f706169e76bd8a06a19b00b43438a7ba2c241b8d463
+size 389587

models/word_ngram/ab_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "ab",
+  "unique_ngrams": 16782,
+  "total_ngrams": 708688
+}

models/word_ngram/ab_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:592dd7484ba41a21c3e7a4f4fc8fc748b19346f94898e746c0776742ba664e4d
+size 699987

models/word_ngram/ab_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "ab",
+  "unique_ngrams": 27732,
+  "total_ngrams": 702160
+}