Upload all models and assets for ba (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 +1,557 @@
+---
+language: ba
+language_name: BA
+language_family: turkic_kipchak
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-turkic_kipchak
+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: 4.068
+  - name: best_isotropy
+    type: isotropy
+    value: 0.7712
+  - name: vocabulary_size
+    type: vocab
+    value: 417410
+generated: 2025-12-27
+---
+
+# BA - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **BA** 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** | 3.248x | 3.21 | 0.2917% | 1,877,404 |
+| **16k** | 3.576x | 3.53 | 0.3212% | 1,705,077 |
+| **32k** | 3.852x | 3.81 | 0.3460% | 1,582,867 |
+| **64k** | 4.068x 🏆 | 4.02 | 0.3653% | 1,499,077 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `Сыңрау торна:
+ Сыңрау торна (йыр) — башҡорт халыҡ йыры.
+ Сыңрау торна — өс актлы...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁с ың рау ▁тор на : ▁с ың рау ▁тор ... (+23 more)` | 33 |
+| 16k | `▁сың рау ▁торна : ▁сың рау ▁торна ▁( йыр ) ... (+17 more)` | 27 |
+| 32k | `▁сың рау ▁торна : ▁сың рау ▁торна ▁( йыр ) ... (+16 more)` | 26 |
+| 64k | `▁сың рау ▁торна : ▁сың рау ▁торна ▁( йыр ) ... (+16 more)` | 26 |
+
+**Sample 2:** `Бөйөк БританияБөйөк Британия`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁бөйөк ▁британия бөйөк ▁британия` | 4 |
+| 16k | `▁бөйөк ▁британия бөйөк ▁британия` | 4 |
+| 32k | `▁бөйөк ▁британия бөйөк ▁британия` | 4 |
+| 64k | `▁бөйөк ▁британия бөйөк ▁британия` | 4 |
+
+**Sample 3:** `Австралия — Көньяҡ ярымшарҙарҙа урынлашҡан дәүләт.
+ Австралия (ҡитға) — Көнсығыш...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁австр алия ▁— ▁көньяҡ ▁ярым шар ҙарҙа ▁урынлашҡан ▁дәүләт . ... (+18 more)` | 28 |
+| 16k | `▁австралия ▁— ▁көньяҡ ▁ярым шар ҙарҙа ▁урынлашҡан ▁дәүләт . ▁австралия ... (+15 more)` | 25 |
+| 32k | `▁австралия ▁— ▁көньяҡ ▁ярымшар ҙарҙа ▁урынлашҡан ▁дәүләт . ▁австралия ▁( ... (+12 more)` | 22 |
+| 64k | `▁австралия ▁— ▁көньяҡ ▁ярымшар ҙарҙа ▁урынлашҡан ▁дәүләт . ▁австралия ▁( ... (+11 more)` | 21 |
+
+
+### Key Findings
+
+- **Best Compression:** 64k achieves 4.068x compression
+- **Lowest UNK Rate:** 8k with 0.2917% 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** | 46,308 🏆 | 15.50 | 552,281 | 13.1% | 33.2% |
+| **2-gram** | 586 🏆 | 9.20 | 17,419 | 48.7% | 94.9% |
+| **3-gram** | 107,842 | 16.72 | 1,025,675 | 10.6% | 27.4% |
+| **3-gram** | 5,171 | 12.34 | 163,683 | 17.5% | 55.2% |
+| **4-gram** | 184,106 | 17.49 | 1,793,503 | 10.9% | 25.9% |
+| **4-gram** | 26,442 | 14.69 | 914,560 | 9.8% | 31.4% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `категория :` | 197,952 |
+| 2 | `. —` | 100,523 |
+| 3 | `) .` | 79,614 |
+| 4 | `) —` | 77,384 |
+| 5 | `) ,` | 71,872 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `% d0 %` | 38,175 |
+| 2 | `йылға бассейны —` | 29,475 |
+| 3 | `. а .` | 21,364 |
+| 4 | `йылғалары категория :` | 20,772 |
+| 5 | `һыу реестры мәғлүмәттәре` | 20,323 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `рәсәй дәүләт һыу реестры` | 20,195 |
+| 2 | `мәғлүмәттәре рәсәй дәүләт һыу` | 20,169 |
+| 3 | `реестры мәғлүмәттәре рәсәй дәүләт` | 20,169 |
+| 4 | `һыу реестры мәғлүмәттәре рәсәй` | 20,166 |
+| 5 | `дәүләт һыу реестрында һыу` | 20,160 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 586
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~31% 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** | 0.7118 | 1.638 | 7.12 | 1,081,802 | 28.8% |
+| **1** | 1.2737 | 2.418 | 9.88 | 4,928 | 0.0% |
+| **2** | 0.3181 | 1.247 | 2.06 | 7,695,897 | 68.2% |
+| **2** | 0.9811 | 1.974 | 7.09 | 48,690 | 1.9% |
+| **3** | 0.1344 | 1.098 | 1.31 | 15,852,609 | 86.6% |
+| **3** | 0.8666 | 1.823 | 4.73 | 344,963 | 13.3% |
+| **4** | 0.0616 🏆 | 1.044 | 1.12 | 20,837,559 | 93.8% |
+| **4** | 0.6873 🏆 | 1.610 | 3.25 | 1,631,641 | 31.3% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `. а . һыу бассейны ) . — шул таштар ҡыҙһа , 1998 ) — 0`
+2. `, суданды иҫәпләмәйенсә ) , 1978 йылда 768 ( номеры ) 9 октябрь 1918 йылдан административ`
+3. `— 28 тайфун ) , табак магнаты , 2006 ) — 13 ғинуарында бәләкәй йылға двина`
+
+**Context Size 2:**
+
+1. `категория : ҡабарҙы - балҡар йылғалары категория : алфавит буйынса шәхестәр категория : рәсәй субъек...`
+2. `. — мәскәү 762 « сапсан » санкт - петербург собор майҙаны ансамбле , солист сифатында саҡыралар`
+3. `) . памятный знак на месте ќ , яңғырау диапозоны киң ( 30 сентябрь 1960 йыл ,`
+
+**Context Size 3:**
+
+1. `% d0 % b5 % d1 % 83 % d0 % b0 % d1 % 86 % d1`
+2. `йылға бассейны — печора һәм обь йылғалары араһындағы , баренц диңгеҙенә ҡойоусы , йылғалар бассейны ...`
+3. `. а . токарев тәҡдим итәләр . немецтарҙы аптыратып , сталин документтар өсөн түләргә ризалаша . әзер...`
+
+**Context Size 4:**
+
+1. `рәсәй дәүләт һыу реестры мәғлүмәте буйынса йылға үрге обь һыу бассейны округында урынлашҡан , һыу ху...`
+2. `реестры мәғлүмәттәре рәсәй дәүләт һыу реестры мәғлүмәте буйынса йылға үрге обь һыу бассейны округынд...`
+3. `мәғлүмәттәре рәсәй дәүләт һыу реестры мәғлүмәте буйынса йылға түбәнге волга һыу бассейны округында у...`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 93.8% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (1,631,641 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 | 417,410 |
+| Total Tokens | 23,479,822 |
+| Mean Frequency | 56.25 |
+| Median Frequency | 4 |
+| Frequency Std Dev | 1249.29 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | һәм | 442,975 |
+| 2 | буйынса | 199,955 |
+| 3 | категория | 198,342 |
+| 4 | һыу | 168,429 |
+| 5 | менән | 154,744 |
+| 6 | йылға | 141,138 |
+| 7 | йылда | 136,378 |
+| 8 | рәсәй | 111,896 |
+| 9 | йыл | 97,392 |
+| 10 | йылдың | 89,845 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | совкомбанк | 2 |
+| 2 | маркетплейстың | 2 |
+| 3 | суларға | 2 |
+| 4 | кишлак | 2 |
+| 5 | пацанский | 2 |
+| 6 | мунден | 2 |
+| 7 | гертфордшир | 2 |
+| 8 | кроуға | 2 |
+| 9 | франклоу | 2 |
+| 10 | алтынкүлдән | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 1.0644 |
+| R² (Goodness of Fit) | 0.989157 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 23.2% |
+| Top 1,000 | 52.0% |
+| Top 5,000 | 71.9% |
+| Top 10,000 | 78.9% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9892 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 23.2% of corpus
+- **Long Tail:** 407,410 words needed for remaining 21.1% 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** | 246,880 | 32 | 3.693 | 1.255 | 0.7645 |
+| **mono_64d** | 246,880 | 64 | 4.152 | 1.202 | 0.7712 🏆 |
+| **mono_128d** | 246,880 | 128 | 4.739 | 1.156 | 0.7517 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_64d with 0.7712 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 246,880 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 (4.07x) with low UNK rate |
+| N-gram | **5-gram** | Lowest perplexity (586) |
+| Markov | **Context-4** | Highest predictability (93.8%) |
+| 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 23:45:09*

models/embeddings/monolingual/ba_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5ab9a23748938ba7a7d67175df00ad9b8f68caab526a73b491efb79bbbbe158
+size 1283130940

models/embeddings/monolingual/ba_128d.meta.json

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

models/embeddings/monolingual/ba_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ba",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 246880
+}

models/embeddings/monolingual/ba_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04654a56597263edcecef3b373e2a46dc6600d993eb9bd089dd4f30b4f6271fc
+size 325527100

models/embeddings/monolingual/ba_32d.meta.json

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

models/embeddings/monolingual/ba_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ba",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 246880
+}

models/embeddings/monolingual/ba_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a93872afbd8f9317ccf9f48907b1875deb4c323daca798bc87cd8b1031f5280d
+size 644728380

models/embeddings/monolingual/ba_64d.meta.json

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

models/embeddings/monolingual/ba_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "ba",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 246880
+}

models/subword_markov/ba_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d6868878ce8c3d2fd2c44240f2a970b42cb2586dbfde7025d1e72a3b6ac70b3
+size 342219

models/subword_markov/ba_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "ba",
+  "unique_contexts": 4928,
+  "total_transitions": 182147511
+}

models/subword_markov/ba_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a89552d8eccadab10a1ddfcdbc4ef36a2c61eb6c16185355b87ec80138c89456
+size 2589714

models/subword_markov/ba_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "ba",
+  "unique_contexts": 48690,
+  "total_transitions": 182083612
+}

models/subword_markov/ba_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe121d741c5aa034400c27fbbc59abf2e556ee7421504c98b9d515f7a84ad7bb
+size 12785656

models/subword_markov/ba_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "ba",
+  "unique_contexts": 344963,
+  "total_transitions": 182019713
+}

models/subword_markov/ba_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ae5d341812bb14f14b59e593a901e72ed00651851f4e438b227da8ad19c4b3b
+size 45384065

models/subword_markov/ba_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "ba",
+  "unique_contexts": 1631641,
+  "total_transitions": 181955814
+}

models/subword_ngram/ba_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1fc455ec80702bbdfc394b20ad7e53af994541f5739559a039c1f14797a9da11
+size 239143

models/subword_ngram/ba_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "ba",
+  "unique_ngrams": 17419,
+  "total_ngrams": 182147511
+}

models/subword_ngram/ba_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5c40130c1d50bed91350758c922cb88d6c733a82bc1c682c711ddb6b3a967810
+size 2055831

models/subword_ngram/ba_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "ba",
+  "unique_ngrams": 163683,
+  "total_ngrams": 182083612
+}

models/subword_ngram/ba_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69d73f681e1e459c2bb8a107c96c5c890e3a2e44d6ab89934d8b31a552557178
+size 11914292

models/subword_ngram/ba_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "ba",
+  "unique_ngrams": 914560,
+  "total_ngrams": 182019713
+}

models/tokenizer/ba_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9f7f8c6111aa25b0301c425ef2ca81bd1c492beb384a92794a1bb5cc8894883
+size 585441

models/tokenizer/ba_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6e332a25893c33db32e04b3f994922558d7dc8b80336f210ced52c2fd41f20f5
+size 959495

models/tokenizer/ba_tokenizer_64k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cea1f24f48084bfadca804cc73807fcc76d7f6bb1dff92a5c5e66974005805d8
+size 1732060

models/tokenizer/ba_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1e002b5188a90a30bfff5759ab095f75651e042f7c82b1813539e5174c356ccc
+size 406416

models/vocabulary/ba_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:649c9e7928c6d049c1807e44ee84ba39de0da8e0afe944c65f92c5e67dc152ea
+size 7166077

models/vocabulary/ba_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "ba",
+  "vocabulary_size": 417410,
+  "statistics": {
+    "type_token_ratio": 0.04478519689532156,
+    "coverage": {
+      "top_100": 0.22514075312093942,
+      "top_1000": 0.505577357431498,
+      "top_5000": 0.6994683745965613,
+      "top_10000": 0.7677274461586074
+    },
+    "hapax_count": 663870,
+    "hapax_ratio": 0.6139667801124593,
+    "total_documents": 63899
+  }
+}

models/word_markov/ba_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4875129953afdc863c3b7840b357671100ecec9e86e7b3b1f5291e35e34721f
+size 96022509

models/word_markov/ba_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "ba",
+  "unique_contexts": 1081802,
+  "total_transitions": 31976815
+}

models/word_markov/ba_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:40cea10bfc8003e0da9502d75d5a3f40994d832dc0531213e610dc8d43e240b8
+size 318818703

models/word_markov/ba_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "ba",
+  "unique_contexts": 7695897,
+  "total_transitions": 31912918
+}

models/word_markov/ba_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f02c513c6045c41dd1b65bd8350a8ae70c94cc93ed3a0e138a41a538bcc05b9
+size 525618960

models/word_markov/ba_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "ba",
+  "unique_contexts": 15852609,
+  "total_transitions": 31849021
+}

models/word_markov/ba_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:285ccd82b56012a9d822c5a84b5bd795f9f01fdd006e41e3d182acacb79cd943
+size 670487523

models/word_markov/ba_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "ba",
+  "unique_contexts": 20837559,
+  "total_transitions": 31785127
+}

models/word_ngram/ba_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cfe67bcedd2ac32d53a8a2dd06d0192ad020853a8783d4bb0a6899c0ea5f9839
+size 13228712

models/word_ngram/ba_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "ba",
+  "unique_ngrams": 552281,
+  "total_ngrams": 31976815
+}

models/word_ngram/ba_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8dfd471500e30bd3120ae58431c82aac2f9ea6b68195d44bb94fde6f64830a54
+size 27689470

models/word_ngram/ba_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "ba",
+  "unique_ngrams": 1025675,
+  "total_ngrams": 31912918
+}

models/word_ngram/ba_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:577732cb5b7771e05860e02c4596a4b43feba90e0cbe682d7bca85459e12e11a
+size 51575545

models/word_ngram/ba_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "ba",
+  "unique_ngrams": 1793503,
+  "total_ngrams": 31849021
+}