Upload all models and assets for bew (20251201)

.gitattributes

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

README.md

@@ -0,0 +1,554 @@
+---
+language: bew
+language_name: BEW
+language_family: austronesian_malay
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-austronesian_malay
+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: 4.487
+  - name: best_isotropy
+    type: isotropy
+    value: 0.7802
+  - name: vocabulary_size
+    type: vocab
+    value: 18978
+generated: 2025-12-28
+---
+
+# BEW - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **BEW** 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** | 3.702x | 3.68 | 0.1337% | 174,282 |
+| **16k** | 3.991x | 3.96 | 0.1441% | 161,647 |
+| **32k** | 4.245x | 4.22 | 0.1533% | 152,003 |
+| **64k** | 4.487x 🏆 | 4.46 | 0.1620% | 143,793 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `Pondok Melati entu kecamatan nyang ada di Bekasi Kota. Ni kecamatan ngejenggar a...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁pondok ▁melati ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bekasi ▁kota . ... (+18 more)` | 28 |
+| 16k | `▁pondok ▁melati ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bekasi ▁kota . ... (+18 more)` | 28 |
+| 32k | `▁pondok ▁melati ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bekasi ▁kota . ... (+18 more)` | 28 |
+| 64k | `▁pondok ▁melati ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bekasi ▁kota . ... (+18 more)` | 28 |
+
+**Sample 2:** `Tanah Sarèal entu kecamatan nyang ada di Bogor Kota. Ni kecamatan ngejenggar amp...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁tanah ▁sar è al ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bogor ... (+18 more)` | 28 |
+| 16k | `▁tanah ▁sar è al ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bogor ... (+18 more)` | 28 |
+| 32k | `▁tanah ▁sarè al ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bogor ▁kota ... (+17 more)` | 27 |
+| 64k | `▁tanah ▁sarèal ▁entu ▁kecamatan ▁nyang ▁ada ▁di ▁bogor ▁kota . ... (+16 more)` | 26 |
+
+**Sample 3:** `Forum lingkar pena ielah`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁forum ▁lingkar ▁pena ▁i elah` | 5 |
+| 16k | `▁forum ▁lingkar ▁pena ▁ielah` | 4 |
+| 32k | `▁forum ▁lingkar ▁pena ▁ielah` | 4 |
+| 64k | `▁forum ▁lingkar ▁pena ▁ielah` | 4 |
+
+
+### Key Findings
+
+- **Best Compression:** 64k achieves 4.487x compression
+- **Lowest UNK Rate:** 8k with 0.1337% 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** | 2,734 🏆 | 11.42 | 11,166 | 32.8% | 60.2% |
+| **2-gram** | 288 🏆 | 8.17 | 2,851 | 67.0% | 98.5% |
+| **3-gram** | 3,250 | 11.67 | 13,829 | 31.3% | 56.8% |
+| **3-gram** | 2,150 | 11.07 | 19,026 | 28.2% | 72.7% |
+| **4-gram** | 4,868 | 12.25 | 21,467 | 28.6% | 50.8% |
+| **4-gram** | 9,513 | 13.22 | 75,515 | 16.1% | 45.6% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `gundul :` | 3,517 |
+| 2 | `category :` | 3,400 |
+| 3 | `( hurup` | 3,387 |
+| 4 | `arab gundul` | 3,312 |
+| 5 | `hurup arab` | 3,190 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `arab gundul :` | 3,306 |
+| 2 | `hurup arab gundul` | 3,176 |
+| 3 | `( hurup arab` | 3,132 |
+| 4 | `ruju ' an` | 2,848 |
+| 5 | `. ruju '` | 2,632 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `hurup arab gundul :` | 3,171 |
+| 2 | `( hurup arab gundul` | 3,124 |
+| 3 | `. ruju ' an` | 2,631 |
+| 4 | `' an category :` | 1,377 |
+| 5 | `ruju ' an category` | 1,375 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 288
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~46% 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** | 0.7488 | 1.680 | 4.84 | 43,321 | 25.1% |
+| **1** | 0.9960 | 1.994 | 6.07 | 1,469 | 0.4% |
+| **2** | 0.2696 | 1.205 | 1.61 | 209,140 | 73.0% |
+| **2** | 0.8519 | 1.805 | 4.62 | 8,912 | 14.8% |
+| **3** | 0.0871 | 1.062 | 1.16 | 336,473 | 91.3% |
+| **3** | 0.7791 | 1.716 | 3.44 | 41,145 | 22.1% |
+| **4** | 0.0352 🏆 | 1.025 | 1.07 | 389,729 | 96.5% |
+| **4** | 0.5441 🏆 | 1.458 | 2.28 | 141,337 | 45.6% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `. barung nyang tinggal di awalnya ga tau - laèn - terusan . 15 agustus 1945`
+2. `, jadinya orang / 347 sm , canada . ada punya waktu junior , sirilik :`
+3. `: rectum , 3 taon ( uè ] ˤəʔ ( 2005 ) ièlah pahlawan nasional ,`
+
+**Context Size 2:**
+
+1. `gundul : ارسخ ; basa katalan : espanya ; basa pin ama swèd , cuman enni kué`
+2. `category : papua nugini category : kota di bilangan èropa kidul - wètan . ni basa bunyi`
+3. `( hurup arab gundul : کبليک ) – ibukota ( indo . ) , atawa nyang seantéro`
+
+**Context Size 3:**
+
+1. `arab gundul : ارلوجى کنتوڠ ) atawa horloji kantong ( hurup arab gundul : تورون بروء ) ièlah`
+2. `hurup arab gundul : موبل بک ) atawa losbak ( hurup arab gundul : خرکف ; rus :`
+3. `( hurup arab gundul : غزة ; arab : تشاد tsyād ) atawa resminya kiblik nigèr ( hurup`
+
+**Context Size 4:**
+
+1. `hurup arab gundul : چلان ) entu pakéan luaran nyang nutupin pinggang ampé kekiongan , kadang cuman e...`
+2. `( hurup arab gundul : ) atawa gedong lèr pèjèngan ( hurup arab gundul : ) entu kejadian cuaca`
+3. `. ruju ' an sènggètan luar martunis sarbini di instagram martunis ronaldo di instagram martunis rona...`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 96.5% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (141,337 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 | 18,978 |
+| Total Tokens | 362,916 |
+| Mean Frequency | 19.12 |
+| Median Frequency | 4 |
+| Frequency Std Dev | 165.48 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | di | 10,336 |
+| 2 | nyang | 9,101 |
+| 3 | ama | 5,533 |
+| 4 | entu | 5,338 |
+| 5 | ada | 4,150 |
+| 6 | atawa | 3,974 |
+| 7 | ni | 3,950 |
+| 8 | orang | 3,897 |
+| 9 | punya | 3,836 |
+| 10 | hurup | 3,665 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | abi | 2 |
+| 2 | gelanggang | 2 |
+| 3 | m² | 2 |
+| 4 | writing | 2 |
+| 5 | syaamil | 2 |
+| 6 | fermentasi | 2 |
+| 7 | xl | 2 |
+| 8 | oase | 2 |
+| 9 | sos | 2 |
+| 10 | litt | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 1.0759 |
+| R² (Goodness of Fit) | 0.994882 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 41.4% |
+| Top 1,000 | 69.4% |
+| Top 5,000 | 87.6% |
+| Top 10,000 | 94.3% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9949 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 41.4% of corpus
+- **Long Tail:** 8,978 words needed for remaining 5.7% 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** | 7,863 | 32 | 3.543 | 0.852 | 0.7802 🏆 |
+| **mono_64d** | 7,863 | 64 | 3.726 | 0.807 | 0.4496 |
+| **mono_128d** | 7,863 | 128 | 3.776 | 0.808 | 0.1049 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.7802 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 7,863 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 (4.49x) with low UNK rate |
+| N-gram | **5-gram** | Lowest perplexity (288) |
+| Markov | **Context-4** | Highest predictability (96.5%) |
+| 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 02:18:06*

models/embeddings/monolingual/bew_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4335b040a535b0bd154cb31255037f3ed12d90b8411695a2d0d4ad1728d4dc87
+size 1032183649

models/embeddings/monolingual/bew_128d.meta.json

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

models/embeddings/monolingual/bew_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bew",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 7863
+}

models/embeddings/monolingual/bew_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2244ac75f23113c3ac11f29736e5e29c03b5e1e667fd5fd810422aa4e47da53
+size 258144865

models/embeddings/monolingual/bew_32d.meta.json

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

models/embeddings/monolingual/bew_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bew",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 7863
+}

models/embeddings/monolingual/bew_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f38e536291d4ea1e5f53da361e36edb36a40a0edd8714d2da36cec0726e5b622
+size 516157793

models/embeddings/monolingual/bew_64d.meta.json

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

models/embeddings/monolingual/bew_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bew",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 7863
+}

models/subword_markov/bew_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bf262c92b34ec5661d8db620bc74ff3485373c52fa29ff8c7d2422a8136da30
+size 73122

models/subword_markov/bew_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "bew",
+  "unique_contexts": 1469,
+  "total_transitions": 2526460
+}

models/subword_markov/bew_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6554a7de4144be71a218ddddea00779aef61fc58f934c3f0191ca8c80355c2f
+size 335416

models/subword_markov/bew_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "bew",
+  "unique_contexts": 8912,
+  "total_transitions": 2523179
+}

models/subword_markov/bew_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11737a41163ad2d3dc78a88b3374643f4ab61b54f3acdd192526ea1a5ee7772c
+size 1094363

models/subword_markov/bew_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "bew",
+  "unique_contexts": 41145,
+  "total_transitions": 2519898
+}

models/subword_markov/bew_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16cda91fbc05488cedf0da6299c8e562dc38d26e9a3e36cfcdc4dd6d72338234
+size 2664959

models/subword_markov/bew_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "bew",
+  "unique_contexts": 141337,
+  "total_transitions": 2516617
+}

models/subword_ngram/bew_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d11f42c7b2a133d183fa06b571114f5938df285f4b5141918f3b5efff0030d2
+size 38137

models/subword_ngram/bew_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "bew",
+  "unique_ngrams": 2851,
+  "total_ngrams": 2526460
+}

models/subword_ngram/bew_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:48279f1d7817a148fc1255e0da6f55d1433148f1da9d1236709123c496ee8de8
+size 234049

models/subword_ngram/bew_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "bew",
+  "unique_ngrams": 19026,
+  "total_ngrams": 2523179
+}

models/subword_ngram/bew_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:258a4beb45713e7bf934e170923447b31b46b1ce23f58937f21f56f082969f23
+size 897268

models/subword_ngram/bew_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "bew",
+  "unique_ngrams": 75515,
+  "total_ngrams": 2519898
+}

models/tokenizer/bew_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f3a1b400084e8023377501b802845cb1cf711d79ac5febd723de2b21cb129d8e
+size 503204

models/tokenizer/bew_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e6232b0f1128785af7cc902e6ccd07e21b7b7b98c247d8d99404f7ad262866c1
+size 778728

models/tokenizer/bew_tokenizer_64k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a43c4038dd4030c982e33b34e05748d9a148f335d8f2efba64b7c0538588aed
+size 1366171

models/tokenizer/bew_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef7e45246fddc1dfe48d9fe97af9c3c7abf4c75cfebb0b0038e553937f8383e4
+size 369839

models/vocabulary/bew_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5ee57d5cd1890a06d6b99c4a39ea33a34f0319ef69950667ae5e96b854e97e04
+size 319583

models/vocabulary/bew_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "bew",
+  "vocabulary_size": 18978,
+  "statistics": {
+    "type_token_ratio": 0.11140219657569815,
+    "coverage": {
+      "top_100": 0.38824255858956175,
+      "top_1000": 0.6505553445616538,
+      "top_5000": 0.8214500706614272,
+      "top_10000": 0.884600976083626
+    },
+    "hapax_count": 24141,
+    "hapax_ratio": 0.5598691991929312,
+    "total_documents": 3281
+  }
+}

models/word_markov/bew_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3cdf9da068448c4bfbde696835827ac01e4a769c1ec6964ab6bd9945242f7276
+size 1573718

models/word_markov/bew_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "bew",
+  "unique_contexts": 43321,
+  "total_transitions": 502833
+}

models/word_markov/bew_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f8057ab45a27f71e7f400d4c915c286a5afb7564f853cbd23531ee51ffb80882
+size 3691116

models/word_markov/bew_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "bew",
+  "unique_contexts": 209140,
+  "total_transitions": 499553
+}

models/word_markov/bew_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23e99493fac0cadc2995899df423385a80a06621183aebb7e369e7e8e8037117
+size 5335565

models/word_markov/bew_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "bew",
+  "unique_contexts": 336473,
+  "total_transitions": 496273
+}

models/word_markov/bew_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87f64763bb79f2a0ab4341bd982916a269745d23f99b7c968f4da43a80c0bd07
+size 6324742

models/word_markov/bew_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "bew",
+  "unique_contexts": 389729,
+  "total_transitions": 492993
+}

models/word_ngram/bew_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd99b15683b2e4627c958b5bdce81a6e91e3f8835310f01ce330b91f6b97a00c
+size 151156

models/word_ngram/bew_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "bew",
+  "unique_ngrams": 11166,
+  "total_ngrams": 502833
+}

models/word_ngram/bew_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c39fbaf39f746f27447ed0c0828a54ef2a669878f0ec7bffe54e6814f269aa6c
+size 207491

models/word_ngram/bew_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "bew",
+  "unique_ngrams": 13829,
+  "total_ngrams": 499553
+}

models/word_ngram/bew_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2518c9520f8d848bf3480605868ff5db2743aa855b5a1f2076c9d33ecccba861
+size 356085

models/word_ngram/bew_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "bew",
+  "unique_ngrams": 21467,
+  "total_ngrams": 496273
+}