Upload all models and assets for cr (20251001)

.gitattributes

@@ -33,3 +33,6 @@ 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

README.md

@@ -0,0 +1,668 @@
+---
+language: cr
+language_name: CR
+language_family: american_algonquian
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-american_algonquian
+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: 3.182
+  - name: best_isotropy
+    type: isotropy
+    value: 0.0381
+  - name: vocabulary_size
+    type: vocab
+    value: 0
+generated: 2026-01-03
+---
+
+# CR - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **CR** 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, 5-gram)
+- Markov chains (context of 1, 2, 3, 4 and 5)
+- Subword N-gram and Markov chains
+- Embeddings in various sizes and dimensions (aligned and unaligned)
+- 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. Morphological Analysis (Experimental)](#6-morphological-analysis)
+- [7. Summary & Recommendations](#7-summary--recommendations)
+- [Metrics Glossary](#appendix-metrics-glossary--interpretation-guide)
+- [Visualizations Index](#visualizations-index)
+
+---
+## 1. Tokenizer Evaluation
+
+![Tokenizer Compression](visualizations/tokenizer_compression.png)
+
+![Tokenizer Fertility](visualizations/tokenizer_fertility.png)
+
+![Tokenizer OOV](visualizations/tokenizer_oov.png)
+
+![Total Tokens](visualizations/tokenizer_total_tokens.png)
+
+### Results
+
+| Vocab Size | Compression | Avg Token Len | UNK Rate | Total Tokens |
+|------------|-------------|---------------|----------|--------------|
+| **8k** | 3.182x 🏆 | 3.19 | 2.9567% | 6,629 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `ᐊᓐ ᐊᒋᐦᑖᓱᓐ ᐯᔭᒄ ᑲ ᐃᔑᓂᐦᑳᑌᒡ, ᐋᐸᑎᓐ ᒉ ᒌ ᐃᑣᓅᐦᒡ ᐯᔭᒄ ᒉᒀᓐ ᒫᒃ ᐊᐌᓐ᙮ ᐊᓐ ᒫᒃ ᐊᒋᐦᑖᓱᓐ ᐯᔭᒄ, ᐁᐅᑯᓐ ᓃ...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ᐊᓐ ▁ᐊᒋᐦᑖᓱᓐ ▁ᐯᔭᒄ ▁ᑲ ▁ᐃᔑᓂᐦᑳᑌᒡ , ▁ᐋᐸᑎᓐ ▁ᒉ ▁ᒌ ▁ᐃᑣᓅᐦᒡ ... (+19 more)` | 29 |
+
+**Sample 2:** `ᓀᐦᐃᔭᐁᐧᐃᐧᐣ ᑕᐣᓯ ᑲ ᐃᓯᐲᑭᐢᑫᐧᕁ ᓵᓴᕀ ᐳᓂ ᐱᑭᐢᑫᐧᐃᐧᐣ ᐱᐦᒑᔨᕁ ᑳᓇᑕ. ᓵᓴᕀ ᐳᓂ ᐱᑭᐢᑫᐧᐃᐧᐣ ᓇᐊᐧᐨ ᐳᑯ ᒌᑳᐦᑕ...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ᓀᐦᐃᔭᐁᐧᐃᐧᐣ ▁ᑕᐣᓯ ▁ᑲ ▁ᐃᓯᐲᑭᐢᑫᐧᕁ ▁ᓵᓴᕀ ▁ᐳᓂ ▁ᐱᑭᐢᑫᐧᐃᐧᐣ ▁ᐱᐦᒑᔨᕁ ▁ᑳᓇᑕ . ... (+11 more)` | 21 |
+
+**Sample 3:** `ᒨᔅ, Muus, Mush ( ; ) n.a. ᐊᐧᐁᓰᔅ ᐆ᙮ ᒨᔅ ᒥᐦᒑᐱᔅᒋᓲ᙮ ᓂᒥᑕᐦᐊᒻ ᑲᔦᐦ᙮ ᐸᐹᒦᒋᓲ᙮ ᒨᒥᓀᐤ᙮ ᒨᔅ ᒦᒎ ᓂᐦ...`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ᒨᔅ , ▁muus , ▁mush ▁( ▁; ▁) ▁n . ... (+17 more)` | 27 |
+
+
+### Key Findings
+
+- **Best Compression:** 8k achieves 3.182x compression
+- **Lowest UNK Rate:** 8k with 2.9567% 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 Unique](visualizations/ngram_unique.png)
+
+![N-gram Coverage](visualizations/ngram_coverage.png)
+
+### Results
+
+| N-gram | Variant | Perplexity | Entropy | Unique N-grams | Top-100 Coverage | Top-1000 Coverage |
+|--------|---------|------------|---------|----------------|------------------|-------------------|
+| **2-gram** | Word | 16 | 4.04 | 17 | 100.0% | 100.0% |
+| **2-gram** | Subword | 492 | 8.94 | 848 | 48.2% | 100.0% |
+| **3-gram** | Word | 15 🏆 | 3.88 | 16 | 100.0% | 100.0% |
+| **3-gram** | Subword | 1,528 | 10.58 | 1,986 | 19.4% | 75.4% |
+| **4-gram** | Word | 163 | 7.35 | 166 | 62.1% | 100.0% |
+| **4-gram** | Subword | 3,131 | 11.61 | 3,878 | 11.9% | 50.9% |
+
+### Top 5 N-grams by Size
+
+**2-grams (Word):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `some articles` | 10 |
+| 2 | `articles in` | 10 |
+| 3 | `ēkwa mīna` | 8 |
+| 4 | `list of` | 8 |
+| 5 | `of articles` | 8 |
+
+**3-grams (Word):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `some articles in` | 10 |
+| 2 | `list of articles` | 8 |
+| 3 | `dialect list of` | 6 |
+| 4 | `cree iso 639` | 5 |
+| 5 | `ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ` | 5 |
+
+**4-grams (Word):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `dialect list of articles` | 6 |
+| 2 | `in standard roman orthography` | 5 |
+| 3 | `written in standard roman` | 5 |
+| 4 | `ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ` | 4 |
+| 5 | `of articles some articles` | 3 |
+
+**2-grams (Subword):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `i n` | 215 |
+| 2 | `, _` | 213 |
+| 3 | `_ ᐊ` | 179 |
+| 4 | `i k` | 168 |
+| 5 | `n _` | 165 |
+
+**3-grams (Subword):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `i n _` | 61 |
+| 2 | `a n i` | 49 |
+| 3 | `w i n` | 48 |
+| 4 | `_ k i` | 47 |
+| 5 | `k a n` | 46 |
+
+**4-grams (Subword):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `w a k _` | 33 |
+| 2 | `w i n _` | 27 |
+| 3 | `t i o n` | 24 |
+| 4 | `k a n i` | 23 |
+| 5 | `i k a n` | 22 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 3-gram (word) with 15
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~51% of corpus
+- **Recommendation:** 4-gram or 5-gram for best predictive performance
+
+---
+## 3. Markov Chain Evaluation
+
+![Markov Entropy](visualizations/markov_entropy.png)
+
+![Markov Contexts](visualizations/markov_contexts.png)
+
+![Markov Branching](visualizations/markov_branching.png)
+
+### Results
+
+| Context | Variant | Avg Entropy | Perplexity | Branching Factor | Unique Contexts | Predictability |
+|---------|---------|-------------|------------|------------------|-----------------|----------------|
+| **1** | Word | 0.2827 | 1.216 | 1.47 | 1,787 | 71.7% |
+| **1** | Subword | 1.9100 | 3.758 | 10.53 | 273 | 0.0% |
+| **2** | Word | 0.0424 | 1.030 | 1.05 | 2,607 | 95.8% |
+| **2** | Subword | 0.6919 | 1.615 | 2.63 | 2,872 | 30.8% |
+| **3** | Word | 0.0178 | 1.012 | 1.02 | 2,724 | 98.2% |
+| **3** | Subword | 0.3559 | 1.280 | 1.57 | 7,557 | 64.4% |
+| **4** | Word | 0.0086 🏆 | 1.006 | 1.01 | 2,765 | 99.1% |
+| **4** | Subword | 0.1591 | 1.117 | 1.21 | 11,842 | 84.1% |
+
+### Generated Text Samples (Word-based)
+
+Below are text samples generated from each word-based Markov chain model:
+
+**Context Size 1:**
+
+1. `ᐁ ᐊᐧᐃᐢᑮᐦᐃᑲᐣ ᐃᐧᐊ ᐁᔥᐃᐦᑕᒧᐃᐧᐣ ᑳ ᐋᐸᐦᐄᔥᑌᒡ english and montana some articles in iyuw iyimuun natuashish dia...`
+2. `e kašcihot e wîcit e iskwewit mâk atimwa wes namawîy nataweyihtam cecî cisceyihtâkwaniyic ekw wenâpe...`
+3. `of nonkilling channel on l nehirâmowin qc r s t u v w ᐌ ᐎ ᐒ`
+
+**Context Size 2:**
+
+1. `some articles in ininiwi išikišwēwin eastern dialect la romaine mingan natashquan pakuashipi and she...`
+2. `articles in lehlueun western dialect list of articles some articles in nīhithawīwin list of articles...`
+3. `ēkwa mīna otaskānitik e ka naskahtamēw nikiskihcēta anihi tahki itēhk kā pēhtahkik tānpahtiwin ē mic...`
+
+**Context Size 3:**
+
+1. `some articles in nēhiyawēwin âpihtâkosisânak kâ isiwepahki maskisin ᐸᐦᑵᓯᑲᐣ pimîhkân tipahikan itasin...`
+2. `list of articles wikipedias in other native american languages atikamekw avañe ẽ aymar choctaw ꮳꮃꭹ c...`
+3. `dialect list of articles some articles in ililîmowin list of articles ᐃᓕᓖᒧᐎᓐ ililîmowin ililîmowin p...`
+
+**Context Size 4:**
+
+1. `dialect list of articles some articles in iyuw iyimuun kawawachikamach dialect list of articles nīhi...`
+2. `written in standard roman orthography`
+3. `ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᑎᐸᐦᐄᑲᓐ ᐋᐱᐦᑖᒌᔑᑳᐤ ᐋᐱᐦᑖᑎᐱᔅᑳᐤ 1 05 ᐯᔭᒄ ᑎᐸᐦᐄᑲᓐ ᒦᓐ ᓂᔮᔪ ᒥᓂᑯᔥ ᓂᔮᔪ ᒥᓂᑯᔥ ᒥᔮ...`
+
+
+### Generated Text Samples (Subword-based)
+
+Below are text samples generated from each subword-based Markov chain model:
+
+**Context Size 1:**
+
+1. `_ᒉᒀᓐᓂᓕᐅᕝᕙᓪᓗ_ᐃᓐᓂᓂ`
+2. `ik;_ᑲᐤ_(_ᑕᐦᐁᐧᔭᐍᑎ`
+3. `am_ᐁr_ē-nata_ost`
+
+**Context Size 2:**
+
+1. `inēhiyiy-âyot_ayi`
+2. `,_ᐱᔪᓐᓇᖅ_ᖂᑉ_ᒪᓕᒋᐊᓕᖕ`
+3. `_ᐊᑕᐦᑐᒥᒃ_ᑐᒃᓯᓪᓗᓂ_ᐊᓯ`
+
+**Context Size 3:**
+
+1. `in_nešt_mâk_ekwa_a`
+2. `anininisiniw._pask`
+3. `win_okiskān_tipēna`
+
+**Context Size 4:**
+
+1. `wak_tāpihikan_ᐆᒪ_ᐊᐢ`
+2. `win_(statistics_(10`
+3. `tion_métis_federati`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 (word) with 99.1% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (11,842 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 | 489 |
+| Total Tokens | 1,731 |
+| Mean Frequency | 3.54 |
+| Median Frequency | 2 |
+| Frequency Std Dev | 3.40 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | ᐁ | 34 |
+| 2 | e | 30 |
+| 3 | and | 22 |
+| 4 | in | 22 |
+| 5 | of | 22 |
+| 6 | pîsim | 19 |
+| 7 | articles | 18 |
+| 8 | cree | 16 |
+| 9 | dialect | 14 |
+| 10 | kîsikâw | 14 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | ᐸᑦᑕᖕᓂᑦ | 2 |
+| 2 | ordinateur | 2 |
+| 3 | demandez | 2 |
+| 4 | le | 2 |
+| 5 | programme | 2 |
+| 6 | eurêka | 2 |
+| 7 | culture | 2 |
+| 8 | 18 | 2 |
+| 9 | août | 2 |
+| 10 | ᖃᐅᔨᓴᖅᑎᐅᔪᓄᑦ | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 0.5522 |
+| R² (Goodness of Fit) | 0.947702 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 47.6% |
+| Top 1,000 | 0.0% |
+| Top 5,000 | 0.0% |
+| Top 10,000 | 0.0% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9477 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 47.6% of corpus
+- **Long Tail:** -9,511 words needed for remaining 100.0% 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)
+
+
+### 5.1 Cross-Lingual Alignment
+
+> *Note: Multilingual alignment visualization not available for this language.*
+
+
+### 5.2 Model Comparison
+
+| Model | Dimension | Isotropy | Semantic Density | Alignment R@1 | Alignment R@10 |
+|-------|-----------|----------|------------------|---------------|----------------|
+| **mono_32d** | 32 | 0.0381 🏆 | 0.0000 | N/A | N/A |
+| **mono_64d** | 64 | 0.0033 | 0.0000 | N/A | N/A |
+| **mono_128d** | 128 | 0.0000 | 0.0000 | N/A | N/A |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.0381 (more uniform distribution)
+- **Semantic Density:** Average pairwise similarity of 0.0000. Lower values indicate better semantic separation.
+- **Alignment Quality:** No aligned models evaluated in this run.
+- **Recommendation:** 128d aligned for best cross-lingual performance
+
+---
+## 6.  Morphological Analysis (Experimental)
+
+> ⚠️ **Warning:** This language shows low morphological productivity. The statistical signals used for this analysis may be noisy or less reliable than for morphologically rich languages.
+
+This section presents an automated morphological analysis derived from the statistical divergence between word-level and subword-level models. By analyzing where subword predictability spikes and where word-level coverage fails, we can infer linguistic structures without supervised data.
+
+### 6.1 Productivity & Complexity
+
+| Metric | Value | Interpretation | Recommendation |
+|--------|-------|----------------|----------------|
+| Productivity Index | **0.000** | Low morphological productivity | ⚠️ Likely unreliable |
+| Idiomaticity Gap | **-1.000** | Low formulaic content | - |
+
+### 6.2 Affix Inventory (Productive Units)
+
+These are the most productive prefixes and suffixes identified by sampling the vocabulary for global substitutability patterns. A unit is considered an affix if stripping it leaves a valid stem that appears in other contexts.
+
+*No productive affixes detected.*
+
+
+### 6.3 Bound Stems (Lexical Roots)
+
+Bound stems are high-frequency subword units that are semantically cohesive but rarely appear as standalone words. These often correspond to the 'core' of a word that requires inflection or derivation to be valid.
+
+*No significant bound stems detected.*
+
+
+### 6.4 Affix Compatibility (Co-occurrence)
+
+This table shows which prefixes and suffixes most frequently co-occur on the same stems, revealing the 'stacking' rules of the language's morphology.
+
+*No significant affix co-occurrences detected.*
+
+
+### 6.5 Recursive Morpheme Segmentation
+
+Using **Recursive Hierarchical Substitutability**, we decompose complex words into their constituent morphemes. This approach handles nested affixes (e.g., `prefix-prefix-root-suffix`).
+
+*Insufficient data for recursive segmentation.*
+
+
+### 6.6 Linguistic Interpretation
+
+> **Automated Insight:**
+The language CR appears to be more isolating or has a highly fixed vocabulary. Word-level models perform nearly as well as subword models, indicating fewer productive morphological processes.
+
+---
+## 7. Summary & Recommendations
+
+![Performance Dashboard](visualizations/performance_dashboard.png)
+
+### Production Recommendations
+
+| Component | Recommended | Rationale |
+|-----------|-------------|-----------|
+| Tokenizer | **8k BPE** | Best compression (3.18x) |
+| N-gram | **3-gram** | Lowest perplexity (15) |
+| Markov | **Context-4** | Highest predictability (99.1%) |
+| 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},
+  doi = {10.5281/zenodo.18073153},
+  publisher = {Zenodo},
+  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)
+- 🤝 Sponsor: [Featherless AI](https://featherless.ai)
+---
+*Generated by Wikilangs Models Pipeline*
+
+*Report Date: 2026-01-03 10:19:03*

models/embeddings/monolingual/cr_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:322799eab6ee84e7d28a0f76de24efb5b70f9e1fcd0eef88c9b61829ce272397
+size 1024067754

models/embeddings/monolingual/cr_128d.meta.json

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

models/embeddings/monolingual/cr_128d_metadata.json

@@ -0,0 +1,15 @@
+{
+  "language": "cr",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "algorithm": "skipgram",
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5,
+    "encoding_method": "rope",
+    "dim": 128
+  },
+  "vocab_size": 65
+}

models/embeddings/monolingual/cr_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:095584a1fc693fcab2c3fd7f0ac831d533933eb58ce565c3b082866e303db9f6
+size 256017834

models/embeddings/monolingual/cr_32d.meta.json

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

models/embeddings/monolingual/cr_32d_metadata.json

@@ -0,0 +1,15 @@
+{
+  "language": "cr",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "algorithm": "skipgram",
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5,
+    "encoding_method": "rope",
+    "dim": 32
+  },
+  "vocab_size": 65
+}

models/embeddings/monolingual/cr_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fdd5c40a0696cfa8faa26c56bda982ab2a5efbbc0bc7d30fcea9c944b571b542
+size 512034474

models/embeddings/monolingual/cr_64d.meta.json

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

models/embeddings/monolingual/cr_64d_metadata.json

@@ -0,0 +1,15 @@
+{
+  "language": "cr",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "algorithm": "skipgram",
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5,
+    "encoding_method": "rope",
+    "dim": 64
+  },
+  "vocab_size": 65
+}

models/subword_markov/cr_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d336a63f8ad5e082086c17ec41db170270495929a8dd340a6abcd2e5998bc0e6
+size 19362

models/subword_markov/cr_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "cr",
+  "unique_contexts": 273,
+  "total_transitions": 21066
+}

models/subword_markov/cr_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:88cd25c549a06a864e6af9c0a79d29d91815a704038e65f2ce7d2c821349c095
+size 54769

models/subword_markov/cr_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "cr",
+  "unique_contexts": 2872,
+  "total_transitions": 21041
+}

models/subword_markov/cr_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:493e48b8388399c0693ca1644064c6aff29ebd8768e02dc362bd13a9beb9923c
+size 109170

models/subword_markov/cr_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "cr",
+  "unique_contexts": 7557,
+  "total_transitions": 21016
+}

models/subword_markov/cr_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:77ef47f052b70e08ed38cb704ef21e780c36aa884a719e5d4963f672dd6af637
+size 165104

models/subword_markov/cr_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "cr",
+  "unique_contexts": 11842,
+  "total_transitions": 20991
+}

models/subword_ngram/cr_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ec4f4bbb6a07a46393df0aae6d9e7264c3913c983048c199b187ccf1637509f
+size 10150

models/subword_ngram/cr_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "cr",
+  "unique_ngrams": 848,
+  "total_ngrams": 21066
+}

models/subword_ngram/cr_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a341e935103c45c5dd7be03811b9df507bbbcb941614f0a4d3949665defe332
+size 22181

models/subword_ngram/cr_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "cr",
+  "unique_ngrams": 1986,
+  "total_ngrams": 21041
+}

models/subword_ngram/cr_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36e06026aa7388f2d7e58a4f337a1ff541dac38696cfa56eaaaec940804c17a1
+size 46333

models/subword_ngram/cr_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "cr",
+  "unique_ngrams": 3878,
+  "total_ngrams": 21016
+}

models/tokenizer/cr_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a0973f38231d8be6cc17f5bb11b4adcec87ca7e994a2d91d59f4a7f15ea4655f
+size 379309

models/vocabulary/cr_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c9a339a558f9a91e507cf4c6a9afd7657b8a650a75b8594311ae35047326f22c
+size 10707

models/vocabulary/cr_vocabulary_metadata.json

@@ -0,0 +1,15 @@
+{
+  "language": "cr",
+  "vocabulary_size": 489,
+  "variant": "full",
+  "statistics": {
+    "type_token_ratio": 0.5915817165406116,
+    "coverage": {
+      "top_100": 0.2709634988490628,
+      "top_1000": 0.7372574810917462
+    },
+    "hapax_count": 1310,
+    "hapax_ratio": 0.7281823235130628,
+    "total_documents": 25
+  }
+}

models/word_markov/cr_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b9cad226d4f9247d3628f1a8d45e964b5ac7aee3588940f10b795c7db079363
+size 51656

models/word_markov/cr_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "cr",
+  "unique_contexts": 1787,
+  "total_transitions": 3016
+}

models/word_markov/cr_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1f017105092b235e6c58f3c540d7b81c2ec84d6afd12ad9d42b7f06f32711cb
+size 68727

models/word_markov/cr_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "cr",
+  "unique_contexts": 2607,
+  "total_transitions": 2991
+}

models/word_markov/cr_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba438b38b527c374838426729a544c3af06986da131dcd7b6efc96c6174fdf64
+size 78351

models/word_markov/cr_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "cr",
+  "unique_contexts": 2724,
+  "total_transitions": 2966
+}

models/word_markov/cr_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:de165eca955a9457fb7d104a709440b37d8fe11272266a6eaf29d8aca13bcb96
+size 86120

models/word_markov/cr_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "cr",
+  "unique_contexts": 2765,
+  "total_transitions": 2941
+}

models/word_ngram/cr_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:52de354c79218bb55b0791d690a0c3f82e884bbb743d648d216165104095eb1f
+size 2755

models/word_ngram/cr_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "cr",
+  "unique_ngrams": 17,
+  "total_ngrams": 3016
+}

models/word_ngram/cr_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59a0f2119f21ff2bd23e3728beb8036aff3979776cb563f4f9372cec5961037d
+size 3122

models/word_ngram/cr_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "cr",
+  "unique_ngrams": 16,
+  "total_ngrams": 2991
+}

models/word_ngram/cr_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8c1ac43c0e3b24b9b058c232c3156ef3f4650c2f5098663b05aff7249ed3efd3
+size 5830

models/word_ngram/cr_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "cr",
+  "unique_ngrams": 166,
+  "total_ngrams": 2966
+}