Upload all models and assets for bcl (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 +1,560 @@
+---
+language: bcl
+language_name: BCL
+language_family: austronesian_philippine_central
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-austronesian_philippine_central
+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: 4.640
+  - name: best_isotropy
+    type: isotropy
+    value: 0.8200
+  - name: vocabulary_size
+    type: vocab
+    value: 139464
+generated: 2025-12-28
+---
+
+# BCL - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **BCL** 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** | 3.849x | 3.74 | 0.0148% | 391,873 |
+| **16k** | 4.154x | 4.04 | 0.0160% | 363,086 |
+| **32k** | 4.421x | 4.30 | 0.0170% | 341,132 |
+| **64k** | 4.640x 🏆 | 4.51 | 0.0178% | 325,066 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `REDIRECT An Sanduguan`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁re dire ct ▁an ▁sand ug uan` | 7 |
+| 16k | `▁re dire ct ▁an ▁sand ug uan` | 7 |
+| 32k | `▁re direct ▁an ▁sand uguan` | 5 |
+| 64k | `▁re direct ▁an ▁sand uguan` | 5 |
+
+**Sample 2:** `An  sarong komyun asin banwaan sa Provincia nin Cosenza sa rehiyon Calabria kan ...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁an ▁sarong ▁komyun ▁asin ▁banwaan ▁sa ▁provincia ▁nin ▁cos enza ... (+6 more)` | 16 |
+| 16k | `▁an ▁sarong ▁komyun ▁asin ▁banwaan ▁sa ▁provincia ▁nin ▁cosenza ▁sa ... (+5 more)` | 15 |
+| 32k | `▁an ▁sarong ▁komyun ▁asin ▁banwaan ▁sa ▁provincia ▁nin ▁cosenza ▁sa ... (+5 more)` | 15 |
+| 64k | `▁an ▁sarong ▁komyun ▁asin ▁banwaan ▁sa ▁provincia ▁nin ▁cosenza ▁sa ... (+5 more)` | 15 |
+
+**Sample 3:** `An  sarong taon sa Gregoryanong kalendaryo.
+
+ Enero 
+Pebrero
+ Marso 
+Abril
+ Mayo...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁an ▁sarong ▁taon ▁sa ▁gregoryanong ▁kalendaryo . ▁enero ▁pebrero ▁marso ... (+9 more)` | 19 |
+| 16k | `▁an ▁sarong ▁taon ▁sa ▁gregoryanong ▁kalendaryo . ▁enero ▁pebrero ▁marso ... (+9 more)` | 19 |
+| 32k | `▁an ▁sarong ▁taon ▁sa ▁gregoryanong ▁kalendaryo . ▁enero ▁pebrero ▁marso ... (+9 more)` | 19 |
+| 64k | `▁an ▁sarong ▁taon ▁sa ▁gregoryanong ▁kalendaryo . ▁enero ▁pebrero ▁marso ... (+9 more)` | 19 |
+
+
+### Key Findings
+
+- **Best Compression:** 64k achieves 4.640x compression
+- **Lowest UNK Rate:** 8k with 0.0148% 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** | 31,343 🏆 | 14.94 | 180,870 | 14.8% | 31.9% |
+| **2-gram** | 262 🏆 | 8.03 | 8,566 | 68.4% | 98.8% |
+| **3-gram** | 108,578 | 16.73 | 332,655 | 6.5% | 18.2% |
+| **3-gram** | 2,285 | 11.16 | 64,437 | 30.5% | 69.8% |
+| **4-gram** | 210,030 | 17.68 | 511,491 | 6.6% | 14.4% |
+| **4-gram** | 13,379 | 13.71 | 345,622 | 17.2% | 41.0% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `. an` | 41,934 |
+| 2 | `sa mga` | 30,441 |
+| 3 | `an mga` | 27,397 |
+| 4 | `, asin` | 26,685 |
+| 5 | `, an` | 24,473 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `kategorya : mga` | 16,293 |
+| 2 | `. an mga` | 6,827 |
+| 3 | `panluwas na takod` | 5,537 |
+| 4 | `mga panluwas na` | 4,931 |
+| 5 | `toltolan kategorya :` | 4,124 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `mga panluwas na takod` | 4,635 |
+| 2 | `toltolan kategorya : mga` | 2,861 |
+| 3 | `toltolan mga panluwas na` | 2,801 |
+| 4 | `— — — —` | 2,785 |
+| 5 | `. igwa ining sukol` | 2,225 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 262
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~41% 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** | 0.6497 | 1.569 | 5.59 | 379,065 | 35.0% |
+| **1** | 1.0949 | 2.136 | 6.69 | 6,611 | 0.0% |
+| **2** | 0.3654 | 1.288 | 2.19 | 2,116,590 | 63.5% |
+| **2** | 0.6035 | 1.519 | 3.87 | 44,194 | 39.6% |
+| **3** | 0.1662 | 1.122 | 1.36 | 4,629,958 | 83.4% |
+| **3** | 0.7134 | 1.640 | 3.84 | 171,168 | 28.7% |
+| **4** | 0.0685 🏆 | 1.049 | 1.12 | 6,293,312 | 93.1% |
+| **4** | 0.6518 🏆 | 1.571 | 2.96 | 656,881 | 34.8% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `, sarong law jack white house of eastern europe award hale sa ' affaire jean nabiribid`
+2. `sa mga padalian na english rosalía nagpirma sa banwaan kan ikasampulong kabilogan nin edukasyon si a...`
+3. `na mga pagpreparar nin mayor na pigtuturing kan huring pararawitdawit , o tungkod " ) .`
+
+**Context Size 2:**
+
+1. `. an designadong zip code kaini iyo . susog ki milagros perfecto sanchez sa halipot na usipon`
+2. `sa mga osipon sa pilipino na may titulong paghinanyog man , siya nagpoon na mag - audition`
+3. `an mga botelya , pakete nin kakanon asin an responsibilidad . sa ibaba sa kabtang kaini .`
+
+**Context Size 3:**
+
+1. `kategorya : mga 2016 na kagadanan kategorya : mga tataramon na mansakan , iyo an pinagrekonstruhir n...`
+2. `. an mga bitis nin manok sarong seryosong peligro nin pagkahilo sa susunod na taon huli sa iyo`
+3. `panluwas na takod philatlas . com philippine standard geographic code local governance performance m...`
+
+**Context Size 4:**
+
+1. `mga panluwas na takod inactive volcanoes page ( arkibo ) kategorya : mga unibersidad asin kolehiyo s...`
+2. `toltolan kategorya : mga armadong sanga kan mga partido pulitika kategorya : mga organisasyon natugd...`
+3. `toltolan mga panluwas na takod philatlas . com philippine standard geographic code local governance ...`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 93.1% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (656,881 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 | 139,464 |
+| Total Tokens | 6,306,562 |
+| Mean Frequency | 45.22 |
+| Median Frequency | 4 |
+| Frequency Std Dev | 1750.04 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | sa | 340,332 |
+| 2 | na | 337,956 |
+| 3 | an | 230,638 |
+| 4 | kan | 226,231 |
+| 5 | mga | 183,688 |
+| 6 | nin | 132,320 |
+| 7 | asin | 125,887 |
+| 8 | sarong | 62,639 |
+| 9 | si | 54,499 |
+| 10 | the | 44,508 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | zhaparova | 2 |
+| 2 | altynbekov | 2 |
+| 3 | wanatabe | 2 |
+| 4 | megapaniki | 2 |
+| 5 | kordon | 2 |
+| 6 | sobringaran | 2 |
+| 7 | khanid | 2 |
+| 8 | ganish | 2 |
+| 9 | archdioceseofcaceres | 2 |
+| 10 | niceno | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 1.0291 |
+| R² (Goodness of Fit) | 0.993065 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 41.8% |
+| Top 1,000 | 62.8% |
+| Top 5,000 | 79.1% |
+| Top 10,000 | 85.2% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9931 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 41.8% of corpus
+- **Long Tail:** 129,464 words needed for remaining 14.8% 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** | 78,307 | 32 | 3.325 | 0.855 | 0.8200 🏆 |
+| **mono_64d** | 78,307 | 64 | 3.871 | 0.899 | 0.8194 |
+| **mono_128d** | 78,307 | 128 | 4.639 | 0.920 | 0.8065 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.8200 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 78,307 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 (4.64x) with low UNK rate |
+| N-gram | **5-gram** | Lowest perplexity (262) |
+| Markov | **Context-4** | Highest predictability (93.1%) |
+| 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 00:25:48*

models/embeddings/monolingual/bcl_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b68fde739946da5ad381854c13e0c7e9f3ae925a20b45114dead92b2d7a5ec7
+size 1105550844

models/embeddings/monolingual/bcl_128d.meta.json

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

models/embeddings/monolingual/bcl_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bcl",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 78307
+}

models/embeddings/monolingual/bcl_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c5888b050fda02741996a649ef4e574917bf0c70c8bcd7b003fd8a04df805ab
+size 277411068

models/embeddings/monolingual/bcl_32d.meta.json

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

models/embeddings/monolingual/bcl_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bcl",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 78307
+}

models/embeddings/monolingual/bcl_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:73e0742024fc7fef32624a6cdeb09e3dc030122f49523813287cd56514f45ca0
+size 553457660

models/embeddings/monolingual/bcl_64d.meta.json

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

models/embeddings/monolingual/bcl_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "bcl",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 78307
+}

models/subword_markov/bcl_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:739fa35135e9500445bce8bcb51a25f5c3997c1ee417f4e594aafd5b60d2ad0f
+size 286254

models/subword_markov/bcl_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "bcl",
+  "unique_contexts": 6611,
+  "total_transitions": 42024331
+}

models/subword_markov/bcl_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f1fa5543c495a6dc7b592e6ed9db597421ed8a10f87195a57c3a5ae0adb04538
+size 1378883

models/subword_markov/bcl_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "bcl",
+  "unique_contexts": 44194,
+  "total_transitions": 42002807
+}

models/subword_markov/bcl_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:21e6e12757b0460aafcbbdf777b4947f1a68cd9d5a25f399807b67cd4a70bd88
+size 5038178

models/subword_markov/bcl_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "bcl",
+  "unique_contexts": 171168,
+  "total_transitions": 41981283
+}

models/subword_markov/bcl_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d8aaf71ea0ea25c963c31d64367f987031042ce65e722e9f5e8475e81791c7af
+size 15230088

models/subword_markov/bcl_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "bcl",
+  "unique_contexts": 656881,
+  "total_transitions": 41959759
+}

models/subword_ngram/bcl_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5afe288d387fa175c7c566a01acbddcc3e251edf63334aafe7f54ed4dee114e5
+size 112319

models/subword_ngram/bcl_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "bcl",
+  "unique_ngrams": 8566,
+  "total_ngrams": 42024331
+}

models/subword_ngram/bcl_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b01ba4b34f80ef57927352ca4ad9437096837f0d321a57fb99d039db1ecf0d5
+size 793658

models/subword_ngram/bcl_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "bcl",
+  "unique_ngrams": 64437,
+  "total_ngrams": 42002807
+}

models/subword_ngram/bcl_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa935e3bfc1d4549c3217dc3d673805015911677b09755efe6121a07a25deb22
+size 3810759

models/subword_ngram/bcl_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "bcl",
+  "unique_ngrams": 345622,
+  "total_ngrams": 41981283
+}

models/tokenizer/bcl_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eb0c1d6ec9b37aa48ab4580025133b7115a3a321b90a538df159371b6b346dc5
+size 504162

models/tokenizer/bcl_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45ed3fdf3b5b2424886cfa7b6371ad864d24da591648b25007a712969fcd7bf9
+size 779719

models/tokenizer/bcl_tokenizer_64k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bca4456326070a24d1192d3df157782d08cdd45ec794b4e39075351072d4119
+size 1339713

models/tokenizer/bcl_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:001d9a2fb8dd93cbd5b03c53f63b6fda18503cff21a9fdcf488356c5ddaba705
+size 370216

models/vocabulary/bcl_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:377f83e39987b4da8cfa29b2470f276d727552a7628b8a387f78dd8541a70ab0
+size 2359026

models/vocabulary/bcl_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "bcl",
+  "vocabulary_size": 139464,
+  "statistics": {
+    "type_token_ratio": 0.05786067345010461,
+    "coverage": {
+      "top_100": 0.4028147626471449,
+      "top_1000": 0.6046892341630454,
+      "top_5000": 0.7619812262587947,
+      "top_10000": 0.8206949599325984
+    },
+    "hapax_count": 239283,
+    "hapax_ratio": 0.6317753011905045,
+    "total_documents": 21524
+  }
+}

models/word_markov/bcl_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:aa0505ff75b4bb563a9c6248c9ad624724bc60f8e390de4c1480153c67bb974e
+size 19592785

models/word_markov/bcl_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "bcl",
+  "unique_contexts": 379065,
+  "total_transitions": 7959439
+}

models/word_markov/bcl_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:017bca82c0d603a73ea84c30f277a13bf8138f46a8bd609c2ca20192bc0339c2
+size 51192571

models/word_markov/bcl_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "bcl",
+  "unique_contexts": 2116590,
+  "total_transitions": 7937915
+}

models/word_markov/bcl_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2c15d097a7dfd722a8f8ff785ce89195257c0b0623ae7522c58fc7ad017fe2a1
+size 85270277

models/word_markov/bcl_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "bcl",
+  "unique_contexts": 4629958,
+  "total_transitions": 7916395
+}

models/word_markov/bcl_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:06def7271772bbb24e5855c6419398ebbd5ee42dac50cd2db0ae0e82b52681ec
+size 107623142

models/word_markov/bcl_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "bcl",
+  "unique_contexts": 6293312,
+  "total_transitions": 7894880
+}

models/word_ngram/bcl_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:804450949415c5c59b5cff49dc971557289a8e8c8df8fe79908549dace5f80d9
+size 2384371

models/word_ngram/bcl_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "bcl",
+  "unique_ngrams": 180870,
+  "total_ngrams": 7959439
+}

models/word_ngram/bcl_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bffb92dfe9eb20b87645783d0ec648103f9ae0fafa9ccd37d6e397d295243c2a
+size 4734062

models/word_ngram/bcl_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "bcl",
+  "unique_ngrams": 332655,
+  "total_ngrams": 7937915
+}

models/word_ngram/bcl_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bb52d0e8bc028c69de74ba58fb8dc5143ff79cc0d378ce50a1d7f95e483fa874
+size 7776301

models/word_ngram/bcl_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "bcl",
+  "unique_ngrams": 511491,
+  "total_ngrams": 7916395
+}