Upload all models and assets for bpy (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 +1,555 @@
+---
+language: bpy
+language_name: BPY
+language_family: indoaryan_eastern
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-indoaryan_eastern
+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: 5.072
+  - name: best_isotropy
+    type: isotropy
+    value: 0.7157
+  - name: vocabulary_size
+    type: vocab
+    value: 23871
+generated: 2025-12-28
+---
+
+# BPY - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **BPY** 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** | 4.628x | 4.39 | 0.2028% | 107,503 |
+| **16k** | 4.794x | 4.55 | 0.2101% | 103,773 |
+| **32k** | 4.937x | 4.69 | 0.2164% | 100,762 |
+| **64k** | 5.072x 🏆 | 4.81 | 0.2222% | 98,093 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `অক্টোবর ৮, গ্রেগরিয়ান পাঞ্জী হান ইলয়া আজি বসরর ২৮১তম (অধিবর্ষত ২৮২তম) দিন হান।...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁অক্টোবর ▁৮ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২৮১ ... (+22 more)` | 32 |
+| 16k | `▁অক্টোবর ▁৮ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২৮১ ... (+22 more)` | 32 |
+| 32k | `▁অক্টোবর ▁৮ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২৮১ ... (+22 more)` | 32 |
+| 64k | `▁অক্টোবর ▁৮ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২৮১ ... (+22 more)` | 32 |
+
+**Sample 2:** `হোসেনপুর ইউনিয়ন, পলাশবাড়ী
+ হোসেনপুর ইউনিয়ন, রাজৈর`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁হোসেন পুর ▁ইউনিয়ন , ▁পলাশবাড়ী ▁হোসেন পুর ▁ইউনিয়ন , ▁রাজৈর` | 10 |
+| 16k | `▁হোসেনপুর ▁ইউনিয়ন , ▁পলাশবাড়ী ▁হোসেনপুর ▁ইউনিয়ন , ▁রাজৈর` | 8 |
+| 32k | `▁হোসেনপুর ▁ইউনিয়ন , ▁পলাশবাড়ী ▁হোসেনপুর ▁ইউনিয়ন , ▁রাজৈর` | 8 |
+| 64k | `▁হোসেনপুর ▁ইউনিয়ন , ▁পলাশবাড়ী ▁হোসেনপুর ▁ইউনিয়ন , ▁রাজৈর` | 8 |
+
+**Sample 3:** `আগষ্ট ২৫, গ্রেগরিয়ান পাঞ্জী হান ইলয়া আজি বসরর ২৩৭তম (অধিবর্ষত ২৩৮তম) দিন হান। ...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁আগষ্ট ▁২৫ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২৩৭ ... (+22 more)` | 32 |
+| 16k | `▁আগষ্ট ▁২৫ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২���৭ ... (+22 more)` | 32 |
+| 32k | `▁আগষ্ট ▁২৫ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২৩৭ ... (+22 more)` | 32 |
+| 64k | `▁আগষ্ট ▁২৫ , ▁গ্রেগরিয়ান ▁পাঞ্জী ▁হান ▁ইলয়া ▁আজি ▁বসরর ▁২৩৭ ... (+22 more)` | 32 |
+
+
+### Key Findings
+
+- **Best Compression:** 64k achieves 5.072x compression
+- **Lowest UNK Rate:** 8k with 0.2028% 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** | 586 🏆 | 9.19 | 23,018 | 53.4% | 92.4% |
+| **2-gram** | 401 🏆 | 8.65 | 5,122 | 58.3% | 97.5% |
+| **3-gram** | 1,915 | 10.90 | 78,850 | 32.4% | 80.1% |
+| **3-gram** | 1,738 | 10.76 | 36,042 | 31.5% | 79.9% |
+| **4-gram** | 3,658 | 11.84 | 192,863 | 25.6% | 72.1% |
+| **4-gram** | 3,941 | 11.94 | 148,690 | 23.3% | 68.9% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `া র` | 368,952 |
+| 2 | `া ন` | 252,549 |
+| 3 | `ম া` | 171,696 |
+| 4 | `হ া` | 165,000 |
+| 5 | `য ়` | 137,023 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `ি য ়` | 76,882 |
+| 2 | `র ম া` | 75,352 |
+| 3 | `ব া র` | 75,053 |
+| 4 | `া র ো` | 70,050 |
+| 5 | `ম া ন` | 58,522 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `ব া র ো` | 69,326 |
+| 2 | `ইউন ি য ়` | 55,842 |
+| 3 | `ম া ন ু` | 51,209 |
+| 4 | `া ত ্ ত` | 49,012 |
+| 5 | `া র ম া` | 48,203 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 401
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~69% 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** | 0.4826 | 1.397 | 3.74 | 51,595 | 51.7% |
+| **1** | 1.2901 | 2.445 | 11.60 | 918 | 0.0% |
+| **2** | 0.2087 | 1.156 | 2.09 | 192,535 | 79.1% |
+| **2** | 1.0212 | 2.030 | 6.12 | 10,624 | 0.0% |
+| **3** | 0.1700 | 1.125 | 1.65 | 401,469 | 83.0% |
+| **3** | 0.7832 | 1.721 | 3.64 | 64,912 | 21.7% |
+| **4** | 0.1274 🏆 | 1.092 | 1.40 | 660,290 | 87.3% |
+| **4** | 0.5823 🏆 | 1.497 | 2.36 | 236,324 | 41.8% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `া ৪৮ . ২গ ঘর পর ্ ত ি হ া জ ে দ ে ব`
+2. `ি বর থ া ন । এর ে শর জ ি ল া র ১৬ ,`
+3. `র ে র ম ু ন ৬৭ % । আয ় া র ো ৭২০২গ গ`
+
+**Context Size 2:**
+
+1. `া র া ষ ্ ট ্ র া ঘ ি ম া ন ) ১ ,`
+2. `া ন ু ল া র ম া প া হ া ন । খ া লয`
+3. `ম া হ া রহ া ন আস ে । চ ৌ দ ্ র া ষ`
+
+**Context Size 3:**
+
+1. `ি য ় ন , ১৮৭ হ া ন ব া র ো দ ্ র া জ`
+2. `র ম া ম ু ন ি ৫২ % , ব া র ো প া ন ্`
+3. `ব া র ো ' language death ' উল ্ ল ে শব া র ্ ক ি`
+
+**Context Size 4:**
+
+1. `ব া র ো ম ৌ জ া ইউন ি য ় নর স া ক ্ ষরত া`
+2. `ইউন ি য ় ন এগত গ া ঙ : ২১ হ া ন ব া র ো ম`
+3. `ম া ন ু ১৭ , ৬৭৩গ শহর ে দ ে ব া র ো হ ু ক া`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 87.3% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (236,324 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 | 23,871 |
+| Total Tokens | 5,192,993 |
+| Mean Frequency | 217.54 |
+| Median Frequency | 3 |
+| Frequency Std Dev | 6063.65 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | র | 542,779 |
+| 2 | ন | 370,575 |
+| 3 | ম | 283,266 |
+| 4 | ত | 212,745 |
+| 5 | য | 205,066 |
+| 6 | ক | 196,409 |
+| 7 | হ | 185,131 |
+| 8 | ল | 177,046 |
+| 9 | প | 164,231 |
+| 10 | ব | 161,639 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | নঅও | 2 |
+| 2 | অহতই | 2 |
+| 3 | আপত | 2 |
+| 4 | থকভ | 2 |
+| 5 | হগই | 2 |
+| 6 | হসর | 2 |
+| 7 | পরমপ | 2 |
+| 8 | হবন | 2 |
+| 9 | আকগও | 2 |
+| 10 | সযন | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 1.3958 |
+| R² (Goodness of Fit) | 0.983849 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 89.1% |
+| Top 1,000 | 97.2% |
+| Top 5,000 | 98.8% |
+| Top 10,000 | 99.3% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9838 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 89.1% of corpus
+- **Long Tail:** 13,871 words needed for remaining 0.7% 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** | 12,408 | 32 | 4.835 | 0.825 | 0.7157 🏆 |
+| **mono_64d** | 12,408 | 64 | 5.039 | 0.789 | 0.5338 |
+| **mono_128d** | 12,408 | 128 | 5.101 | 0.751 | 0.2644 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.7157 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 12,408 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 (5.07x) with low UNK rate |
+| N-gram | **5-gram** | Lowest perplexity (401) |
+| Markov | **Context-4** | Highest predictability (87.3%) |
+| 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: 2025-12-28 07:50:18*

models/embeddings/monolingual/bpy_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4fb6b15d9bd093df27b0fd9a4efcd269979e163e27ef6ef6754ac07eb74d1ab2
+size 1037046131

models/embeddings/monolingual/bpy_128d.meta.json

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

models/embeddings/monolingual/bpy_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bpy",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 12408
+}

models/embeddings/monolingual/bpy_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47e22726d92510c004d40792ae23c82129d179cc6367e68d751f9764a7ca0df1
+size 259516787

models/embeddings/monolingual/bpy_32d.meta.json

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

models/embeddings/monolingual/bpy_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bpy",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 12408
+}

models/embeddings/monolingual/bpy_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:69a920bfb09617e816f807e477568361980d5eb7d59bc6fce4f483f40da1725b
+size 518693235

models/embeddings/monolingual/bpy_64d.meta.json

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

models/embeddings/monolingual/bpy_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bpy",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 12408
+}

models/subword_markov/bpy_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bdb54276207e5b6285187f2b86c36ad177fb87faaaaefdb2ccb4c908e1f0cb67
+size 87617

models/subword_markov/bpy_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "bpy",
+  "unique_contexts": 918,
+  "total_transitions": 14885921
+}

models/subword_markov/bpy_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4f249b093261db16fcadfa715db9f8802b43064e74e1ba8719ca675247806d36
+size 535820

models/subword_markov/bpy_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "bpy",
+  "unique_contexts": 10624,
+  "total_transitions": 14860752
+}

models/subword_markov/bpy_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ff090decea3ecf5111ffe1a1235436699266f613b38361d41312345bb838b7e
+size 1775613

models/subword_markov/bpy_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "bpy",
+  "unique_contexts": 64912,
+  "total_transitions": 14835583
+}

models/subword_markov/bpy_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52d356df5651a0f5f54331cd5e76f74df23d7adf3534dae88c6e119d9b761ee9
+size 5047184

models/subword_markov/bpy_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "bpy",
+  "unique_contexts": 236324,
+  "total_transitions": 14810414
+}

models/subword_ngram/bpy_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cdabbb125dfaa15729dfa32510019102c91c18a767800511da74d1f6d7f74b99
+size 71525

models/subword_ngram/bpy_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "bpy",
+  "unique_ngrams": 5122,
+  "total_ngrams": 14885921
+}

models/subword_ngram/bpy_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4170250afd8b7e81e3bb514e2ea7e5003b5a4cac40d1d8cf1b944941fd5763b
+size 488221

models/subword_ngram/bpy_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "bpy",
+  "unique_ngrams": 36042,
+  "total_ngrams": 14860752
+}

models/subword_ngram/bpy_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:49e5e1be83e58be9ef9945b7e74642ea9638babffc0bcefcc28e74284c2eca28
+size 1849563

models/subword_ngram/bpy_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "bpy",
+  "unique_ngrams": 148690,
+  "total_ngrams": 14835583
+}

models/tokenizer/bpy_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b1b87ae27874d0e0deb8c8a9507b9010354c9e4be4f579eaf9efb10233bc2b7
+size 608584

models/tokenizer/bpy_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8549dc42b070078e9b8bf38d5cb50e4361f0b9c8ba4de0cb3db16ff7947e4703
+size 1001952

models/tokenizer/bpy_tokenizer_64k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8984f34241873e65994cc434c8d728913d71f46df6299666ebba4dab4b121e30
+size 1889033

models/tokenizer/bpy_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17966726c85a44c946d110bbc967aa3995eace9b617553b974f23d2e598fe035
+size 419939

models/vocabulary/bpy_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68b8a4ca35238e27244b1e44260e17a48c67a189db1d781d8c200c28fb378b13
+size 368371

models/vocabulary/bpy_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "bpy",
+  "vocabulary_size": 23871,
+  "statistics": {
+    "type_token_ratio": 0.009857897178592595,
+    "coverage": {
+      "top_100": 0.8861177653236629,
+      "top_1000": 0.9663878729322724,
+      "top_5000": 0.9828984715508948,
+      "top_10000": 0.9880168624748257
+    },
+    "hapax_count": 27593,
+    "hapax_ratio": 0.5361612000621794,
+    "total_documents": 25169
+  }
+}

models/word_markov/bpy_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1d5036170691d1160320935b9f8ba2016467e19c1053c73425833658f88971c
+size 1613321

models/word_markov/bpy_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "bpy",
+  "unique_contexts": 51595,
+  "total_transitions": 9871256
+}

models/word_markov/bpy_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:131007aca3a521d9f995c3b68152c37568e6a3043b27cee57c0170453d3ad10b
+size 3751981

models/word_markov/bpy_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "bpy",
+  "unique_contexts": 192535,
+  "total_transitions": 9846087
+}

models/word_markov/bpy_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ab0b79c46c94f67d583a0221361f6c4add6cf9a150d20c2a40f4daaef196f21e
+size 7021353

models/word_markov/bpy_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "bpy",
+  "unique_contexts": 401469,
+  "total_transitions": 9820919
+}

models/word_markov/bpy_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:092aa31f510790527aca2846851d191e9138ff743310cdb5e7d9e612f2bc0b69
+size 10993473

models/word_markov/bpy_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "bpy",
+  "unique_contexts": 660290,
+  "total_transitions": 9795751
+}

models/word_ngram/bpy_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8df7389a29eaa5b721ff0e95520427f6b27d7ebab6cd7ffdd086c22978342cb
+size 320237

models/word_ngram/bpy_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "bpy",
+  "unique_ngrams": 23018,
+  "total_ngrams": 9871256
+}

models/word_ngram/bpy_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f719fdf4863b7285c7d7700f0109a2991c67303c3a99f375996b5ce90543fc58
+size 1063287

models/word_ngram/bpy_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "bpy",
+  "unique_ngrams": 78850,
+  "total_ngrams": 9846087
+}

models/word_ngram/bpy_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a0ccd76524932bacdba4faae94a4465412313cd3d52a2e7b8449b2b1881217f
+size 2632657

models/word_ngram/bpy_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "bpy",
+  "unique_ngrams": 192863,
+  "total_ngrams": 9820919
+}