Upload all models and assets for arc (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: arc
+language_name: ARC
+language_family: semitic_aramaic
+tags:
+  - wikilangs
+  - nlp
+  - tokenizer
+  - embeddings
+  - n-gram
+  - markov
+  - wikipedia
+  - monolingual
+  - family-semitic_aramaic
+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: 4.512
+  - name: best_isotropy
+    type: isotropy
+    value: 0.2995
+  - name: vocabulary_size
+    type: vocab
+    value: 6528
+generated: 2025-12-27
+---
+
+# ARC - Wikilangs Models
+## Comprehensive Research Report & Full Ablation Study
+
+This repository contains NLP models trained and evaluated by Wikilangs, specifically on **ARC** 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** | 3.534x | 3.51 | 0.0853% | 59,794 |
+| **16k** | 3.932x | 3.90 | 0.0949% | 53,742 |
+| **32k** | 4.512x 🏆 | 4.48 | 0.1089% | 46,835 |
+
+### Tokenization Examples
+
+Below are sample sentences tokenized with each vocabulary size:
+
+**Sample 1:** `R (ܙܥܘܪܬܐ r) ܗܝ ܐܬܘܬܐ ܕܐܠܦܒܝܬ ܠܐܛܝܢܝܐ܀`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁r ▁( ܙܥܘܪܬܐ ▁r ) ▁ܗܝ ▁ܐܬܘܬܐ ▁ܕܐܠܦܒܝܬ ▁ܠܐܛܝܢܝܐ܀` | 9 |
+| 16k | `▁r ▁( ܙܥܘܪܬܐ ▁r ) ▁ܗܝ ▁ܐܬܘܬܐ ▁ܕܐܠܦܒܝܬ ▁ܠܐܛܝܢܝܐ܀` | 9 |
+| 32k | `▁r ▁( ܙܥܘܪܬܐ ▁r ) ▁ܗܝ ▁ܐܬܘܬܐ ▁ܕܐܠܦܒܝܬ ▁ܠܐܛܝܢܝܐ܀` | 9 |
+
+**Sample 2:** `1847 ܗܘܬ ܫܢܬܐ܀
+
+ܓܕܫ̈ܐ
+
+ ܐܬܝܠܕ 
+
+ ܡܝܬ 
+
+ܣܕܪܐ:ܕܪܐ ܬܫܥܣܪܝܢܝܐ`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ 1 8 4 7 ▁ܗܘܬ ▁ܫܢܬܐ܀ ▁ܓܕܫ̈ܐ ▁ܐܬܝܠܕ ▁ܡܝܬ ... (+5 more)` | 15 |
+| 16k | `▁ 1 8 4 7 ▁ܗܘܬ ▁ܫܢܬܐ܀ ▁ܓܕܫ̈ܐ ▁ܐܬܝܠܕ ▁ܡܝܬ ... (+5 more)` | 15 |
+| 32k | `▁ 1 8 4 7 ▁ܗܘܬ ▁ܫܢܬܐ܀ ▁ܓܕܫ̈ܐ ▁ܐܬܝܠܕ ▁ܡܝܬ ... (+4 more)` | 14 |
+
+**Sample 3:** `ܗܘܦܪܟܝܐ ܕܒܝܠܓܝܟ ܗܝ ܗܘܦܪܟܝܐ ܒܛܘܪܩܝܐ܀
+
+ܣܕܪܐ:ܗܘܦܪܟܝܣ ܕܛܘܪܩܝܐ`
+
+| Vocab | Tokens | Count |
+|-------|--------|-------|
+| 8k | `▁ܗܘܦܪܟܝܐ ▁ܕܒܝܠ ܓ ܝܟ ▁ܗܝ ▁ܗܘܦܪܟܝܐ ▁ܒܛܘܪܩܝܐ܀ ▁ܣܕܪܐ : ܗܘܦܪܟܝܣ ... (+1 more)` | 11 |
+| 16k | `▁ܗܘܦܪܟܝܐ ▁ܕܒܝܠ ܓܝܟ ▁ܗܝ ▁ܗܘܦܪܟܝܐ ▁ܒܛܘܪܩܝܐ܀ ▁ܣܕܪܐ : ܗܘܦܪܟܝܣ ▁ܕܛܘܪܩܝܐ` | 10 |
+| 32k | `▁ܗܘܦܪܟܝܐ ▁ܕܒܝܠܓܝܟ ▁ܗܝ ▁ܗܘܦܪܟܝܐ ▁ܒܛܘܪܩܝܐ܀ ▁ܣܕܪܐ : ܗܘܦܪܟܝܣ ▁ܕܛܘܪܩܝܐ` | 9 |
+
+
+### Key Findings
+
+- **Best Compression:** 32k achieves 4.512x compression
+- **Lowest UNK Rate:** 8k with 0.0853% 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** | 836 🏆 | 9.71 | 1,994 | 37.5% | 82.7% |
+| **2-gram** | 405 🏆 | 8.66 | 2,501 | 57.6% | 95.6% |
+| **3-gram** | 1,500 | 10.55 | 2,669 | 27.2% | 73.4% |
+| **3-gram** | 2,617 | 11.35 | 11,822 | 27.5% | 65.5% |
+| **4-gram** | 2,666 | 11.38 | 4,604 | 22.0% | 58.3% |
+| **4-gram** | 9,085 | 13.15 | 32,191 | 14.3% | 42.7% |
+
+### Top 5 N-grams by Size
+
+**2-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `̈ ܐ` | 2,050 |
+| 2 | `ܣܕܪܐ :` | 1,195 |
+| 3 | `܀ ܣܕܪܐ` | 593 |
+| 4 | `) ܗܝ` | 445 |
+| 5 | `̈ ܝܐ` | 356 |
+
+**3-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `܀ ܣܕܪܐ :` | 593 |
+| 2 | `ܐܢܫ ̈ ܐ` | 135 |
+| 3 | `܀ ܐܦ ܚܙܝ` | 134 |
+| 4 | `̈ ܐ ܀` | 127 |
+| 5 | `ܣܕܪܐ : ܝܘܠܦܢ` | 117 |
+
+**4-grams:**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `ܣܕܪܐ : ܝܘܠܦܢ ܨܪܘܝܘܬܐ` | 115 |
+| 2 | `܀ ܣܕܪܐ : ܝܘܠܦܢ` | 97 |
+| 3 | `̈ ܐ ܒܪ ̈` | 91 |
+| 4 | `ܐ ܒܪ ̈ ܝܐ` | 90 |
+| 5 | `ܐ ܀ ܣܕܪܐ :` | 66 |
+
+
+### Key Findings
+
+- **Best Perplexity:** 2-gram with 405
+- **Entropy Trend:** Decreases with larger n-grams (more predictable)
+- **Coverage:** Top-1000 patterns cover ~43% 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** | 0.5575 | 1.472 | 3.10 | 18,087 | 44.3% |
+| **1** | 1.3634 | 2.573 | 8.68 | 797 | 0.0% |
+| **2** | 0.1553 | 1.114 | 1.32 | 55,465 | 84.5% |
+| **2** | 0.9613 | 1.947 | 4.38 | 6,904 | 3.9% |
+| **3** | 0.0630 | 1.045 | 1.11 | 72,203 | 93.7% |
+| **3** | 0.6343 | 1.552 | 2.52 | 30,176 | 36.6% |
+| **4** | 0.0270 🏆 | 1.019 | 1.04 | 78,995 | 97.3% |
+| **4** | 0.3633 🏆 | 1.286 | 1.71 | 75,950 | 63.7% |
+
+### Generated Text Samples
+
+Below are text samples generated from each Markov chain model:
+
+**Context Size 1:**
+
+1. `̈ ܠܐ ܀ ܣܕܪܐ : ܐܘܢܓܠܝܘܢ ܕܡܪܩܘܣ ܘܐܘܢܓܠܝܘܢ ܕܡܪܩܘܣ ܀ ܣܕܪܐ : ܡܐܢܐ ܕܐܝܬ ܠܗ ܬܪܬܝܢ`
+2. `: ܕܝܬܝܩܝ ܥܬܝܩܬܐ ܘܗܝ ܚܕܐ ܡܢ ܐܠܗܐ ܫܪܝܪܐ ܝܠܝܕܐ ܘܠܐ ܛܥܢܢ ܠܡܕܟܪ ܕܟܢܘܫܬܐ ܡܪܕܘܬܢܝܬܐ ܣܘܪܝܝܬܐ ܐܪܬܘܕܟܣܝܬܐ`
+3. `ܐ ܣܢܝܩܐ ܝܘܚ ܠܡܚܒܢ ̈ ܬܐ ܐܚܪ ̈ ܐ ܩܕܡ ܡܫܝܚܐ ܥܕܡܐ ܠܫܢܬ 1919ܡ ܘܒܡܕܒܚ ̈`
+
+**Context Size 2:**
+
+1. `̈ ܐ ܒܥܠܡܐ . ܘܦܪܣܐ ܒܝܬܝܪ ܡܢ ܠܫܢܐ ܣܘܪܝܝܐ ܘܐܪܡܢܝܐ ܡܬܬܗܪܓܝܢ ܒܡܕܪ ̈ ܫܬܐ ܬܝܪܝܟܝܬܐ ܡܪܘ ̈`
+2. `ܣܕܪܐ : ܛܪܘܢܐ ܣܕܪܐ : ܡܕܝܢܬܐ ܕܥܝܪܐܩ ܣܕܪܐ : ܛܪܘܢܐ ܣܕܪܐ : ܗܘܐ ܒܬܫܪܝܢ ܐܚܪܝܣܕܪܐ : ܒܬܫܪܝܢ`
+3. `܀ ܣܕܪܐ : ܣܘܪܝܐ ܣܕܪܐ : ܒܝܬ ܢܗܪܝܢ ܣܕܪܐ : ܝܗܘܕܝܘܬܐ ܣܕܪܐ : ܡܐܢܐ ܡܘܣܝܩܝܐ . ܒܥܕܬܐ`
+
+**Context Size 3:**
+
+1. `܀ ܣܕܪܐ : ܝܘܠܦܢ ܨܪܘܝܘܬܐ ܣܕܪܐ : ܥܝܢܐ ( ܝܘܠܦܢ ܨܪܘܝܘܬܐ ) ܣܕܪܐ : ܡܫܝܚܝܘܬܐ ܣܕܪܐ : ܕܝܬܝܩܝ`
+2. `ܐܢܫ ̈ ܐ ܒܓܘܪܓܝܐ ܢܡܠܠܘܢ ܓܘܪܓܐܝܬ ܀`
+3. `܀ ܐܦ ܚܙܝ ܓܪܡܐ ܣܕܪܐ : ܝܘܠܦܢ ܟܝܢܝܬܐ`
+
+**Context Size 4:**
+
+1. `܀ ܣܕܪܐ : ܝܘܠܦܢ ܨܪܘܝܘܬܐ ܣܕܪܐ : ܝܘܠܦܢ ܨܪܘܝܘܬܐ ܣܕܪܐ : ܓܪܡܐ`
+2. `̈ ܐ ܒܪ ̈ ܝܐ ܡܓܠܬܐ 1 ܘ 2 ‌ ܘ ܘܡܓܠܬܐ 3 ܕܓܢܙܐ ܪܒܐ ܒܠܫܢܐ ܣܘܪܝܝܐ .`
+3. `ܐ ܒܪ ̈ ܝܐ ܐܓܪܬܐ ܩܕܡܝܬܐ ܕܦܘܠܘܣ ܫܠܝܚܐ ܕܠܘܬ ܛܝܡܬܐܘܣ ܕܬܪܬܝܢ ܚܕܐ ܡܢ ܐܓܪ ̈ ܬܐ ܕܕܝܬܝܩܝ ܚܕܬܐ .`
+
+
+### Key Findings
+
+- **Best Predictability:** Context-4 with 97.3% predictability
+- **Branching Factor:** Decreases with context size (more deterministic)
+- **Memory Trade-off:** Larger contexts require more storage (75,950 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 | 6,528 |
+| Total Tokens | 65,426 |
+| Mean Frequency | 10.02 |
+| Median Frequency | 3 |
+| Frequency Std Dev | 48.74 |
+
+### Most Common Words
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | ܐ | 2,433 |
+| 2 | ܡܢ | 1,300 |
+| 3 | ܣܕܪܐ | 1,205 |
+| 4 | ܐܘ | 1,034 |
+| 5 | ܗܝ | 1,024 |
+| 6 | ܗܘ | 1,023 |
+| 7 | ܐܝܬ | 520 |
+| 8 | ܗܘܐ | 408 |
+| 9 | ܬܐ | 376 |
+| 10 | ܝܐ | 369 |
+
+### Least Common Words (from vocabulary)
+
+| Rank | Word | Frequency |
+|------|------|-----------|
+| 1 | ܟܢܘܢܝܐ | 2 |
+| 2 | ܘܟ | 2 |
+| 3 | ܦܩ | 2 |
+| 4 | ܕܚܘ | 2 |
+| 5 | ܒܐܘ | 2 |
+| 6 | ܪܚ | 2 |
+| 7 | ܐܘܟܝܬܐ | 2 |
+| 8 | ܕܠܥ | 2 |
+| 9 | ܕܒܘ | 2 |
+| 10 | ܠܨܡ | 2 |
+
+### Zipf's Law Analysis
+
+| Metric | Value |
+|--------|-------|
+| Zipf Coefficient | 0.9501 |
+| R² (Goodness of Fit) | 0.985114 |
+| Adherence Quality | **excellent** |
+
+### Coverage Analysis
+
+| Top N Words | Coverage |
+|-------------|----------|
+| Top 100 | 35.0% |
+| Top 1,000 | 70.1% |
+| Top 5,000 | 95.3% |
+| Top 10,000 | 0.0% |
+
+### Key Findings
+
+- **Zipf Compliance:** R²=0.9851 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 35.0% of corpus
+- **Long Tail:** -3,472 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)
+
+### Model Comparison
+
+| Model | Vocab Size | Dimension | Avg Norm | Std Norm | Isotropy |
+|-------|------------|-----------|----------|----------|----------|
+| **mono_32d** | 1,958 | 32 | 3.019 | 0.712 | 0.2995 🏆 |
+| **mono_64d** | 1,958 | 64 | 2.997 | 0.742 | 0.0596 |
+| **mono_128d** | 1,958 | 128 | 2.998 | 0.754 | 0.0093 |
+| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+
+### Key Findings
+
+- **Best Isotropy:** mono_32d with 0.2995 (more uniform distribution)
+- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
+- **Vocabulary Coverage:** All models cover 1,958 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 (4.51x) with low UNK rate |
+| N-gram | **5-gram** | Lowest perplexity (405) |
+| Markov | **Context-4** | Highest predictability (97.3%) |
+| Embeddings | **100d** | Balanced semantic capture and isotropy |
+
+---
+## Appendix: Metrics Glossary & Interpretation Guide
+
+This section provides definitions, intuitions, and guidance for interpreting the metrics used throughout this report.
+
+### Tokenizer Metrics
+
+**Compression Ratio**
+> *Definition:* The ratio of characters to tokens (chars/token). Measures how efficiently the tokenizer represents text.
+>
+> *Intuition:* Higher compression means fewer tokens needed to represent the same text, reducing sequence lengths for downstream models. A 3x compression means ~3 characters per token on average.
+>
+> *What to seek:* Higher is generally better for efficiency, but extremely high compression may indicate overly aggressive merging that loses morphological information.
+
+**Average Token Length (Fertility)**
+> *Definition:* Mean number of characters per token produced by the tokenizer.
+>
+> *Intuition:* Reflects the granularity of tokenization. Longer tokens capture more context but may struggle with rare words; shorter tokens are more flexible but increase sequence length.
+>
+> *What to seek:* Balance between 2-5 characters for most languages. Arabic/morphologically-rich languages may benefit from slightly longer tokens.
+
+**Unknown Token Rate (OOV Rate)**
+> *Definition:* Percentage of tokens that map to the unknown/UNK token, indicating words the tokenizer cannot represent.
+>
+> *Intuition:* Lower OOV means better vocabulary coverage. High OOV indicates the tokenizer encounters many unseen character sequences.
+>
+> *What to seek:* Below 1% is excellent; below 5% is acceptable. BPE tokenizers typically achieve very low OOV due to subword fallback.
+
+### N-gram Model Metrics
+
+**Perplexity**
+> *Definition:* Measures how "surprised" the model is by test data. Mathematically: 2^(cross-entropy). Lower values indicate better prediction.
+>
+> *Intuition:* If perplexity is 100, the model is as uncertain as if choosing uniformly among 100 options at each step. A perplexity of 10 means effectively choosing among 10 equally likely options.
+>
+> *What to seek:* Lower is better. Perplexity decreases with larger n-grams (more context). Values vary widely by language and corpus size.
+
+**Entropy**
+> *Definition:* Average information content (in bits) needed to encode the next token given the context. Related to perplexity: perplexity = 2^entropy.
+>
+> *Intuition:* High entropy means high uncertainty/randomness; low entropy means predictable patterns. Natural language typically has entropy between 1-4 bits per character.
+>
+> *What to seek:* Lower entropy indicates more predictable text patterns. Entropy should decrease as n-gram size increases.
+
+**Coverage (Top-K)**
+> *Definition:* Percentage of corpus occurrences explained by the top K most frequent n-grams.
+>
+> *Intuition:* High coverage with few patterns indicates repetitive/formulaic text; low coverage suggests diverse vocabulary usage.
+>
+> *What to seek:* Depends on use case. For language modeling, moderate coverage (40-60% with top-1000) is typical for natural text.
+
+### Markov Chain Metrics
+
+**Average Entropy**
+> *Definition:* Mean entropy across all contexts, measuring average uncertainty in next-word prediction.
+>
+> *Intuition:* Lower entropy means the model is more confident about what comes next. Context-1 has high entropy (many possible next words); Context-4 has low entropy (few likely continuations).
+>
+> *What to seek:* Decreasing entropy with larger context sizes. Very low entropy (<0.1) indicates highly deterministic transitions.
+
+**Branching Factor**
+> *Definition:* Average number of unique next tokens observed for each context.
+>
+> *Intuition:* High branching = many possible continuations (flexible but uncertain); low branching = few options (predictable but potentially repetitive).
+>
+> *What to seek:* Branching factor should decrease with context size. Values near 1.0 indicate nearly deterministic chains.
+
+**Predictability**
+> *Definition:* Derived metric: (1 - normalized_entropy) × 100%. Indicates how deterministic the model's predictions are.
+>
+> *Intuition:* 100% predictability means the next word is always certain; 0% means completely random. Real text falls between these extremes.
+>
+> *What to seek:* Higher predictability for text generation quality, but too high (>98%) may produce repetitive output.
+
+### Vocabulary & Zipf's Law Metrics
+
+**Zipf's Coefficient**
+> *Definition:* The slope of the log-log plot of word frequency vs. rank. Zipf's law predicts this should be approximately -1.
+>
+> *Intuition:* A coefficient near -1 indicates the corpus follows natural language patterns where a few words are very common and most words are rare.
+>
+> *What to seek:* Values between -0.8 and -1.2 indicate healthy natural language distribution. Deviations may suggest domain-specific or artificial text.
+
+**R² (Coefficient of Determination)**
+> *Definition:* Measures how well the linear fit explains the frequency-rank relationship. Ranges from 0 to 1.
+>
+> *Intuition:* R² near 1.0 means the data closely follows Zipf's law; lower values indicate deviation from expected word frequency patterns.
+>
+> *What to seek:* R² > 0.95 is excellent; > 0.99 indicates near-perfect Zipf adherence typical of large natural corpora.
+
+**Vocabulary Coverage**
+> *Definition:* Cumulative percentage of corpus tokens accounted for by the top N words.
+>
+> *Intuition:* Shows how concentrated word usage is. If top-100 words cover 50% of text, the corpus relies heavily on common words.
+>
+> *What to seek:* Top-100 covering 30-50% is typical. Higher coverage indicates more repetitive text; lower suggests richer vocabulary.
+
+### Word Embedding Metrics
+
+**Isotropy**
+> *Definition:* Measures how uniformly distributed vectors are in the embedding space. Computed as the ratio of minimum to maximum singular values.
+>
+> *Intuition:* High isotropy (near 1.0) means vectors spread evenly in all directions; low isotropy means vectors cluster in certain directions, reducing expressiveness.
+>
+> *What to seek:* Higher isotropy generally indicates better-quality embeddings. Values > 0.1 are reasonable; > 0.3 is good. Lower-dimensional embeddings tend to have higher isotropy.
+
+**Average Norm**
+> *Definition:* Mean magnitude (L2 norm) of word vectors in the embedding space.
+>
+> *Intuition:* Indicates the typical "length" of vectors. Consistent norms suggest stable training; high variance may indicate some words are undertrained.
+>
+> *What to seek:* Relatively consistent norms across models. The absolute value matters less than consistency (low std deviation).
+
+**Cosine Similarity**
+> *Definition:* Measures angular similarity between vectors, ranging from -1 (opposite) to 1 (identical direction).
+>
+> *Intuition:* Words with similar meanings should have high cosine similarity. This is the standard metric for semantic relatedness in embeddings.
+>
+> *What to seek:* Semantically related words should score > 0.5; unrelated words should be near 0. Synonyms often score > 0.7.
+
+**t-SNE Visualization**
+> *Definition:* t-Distributed Stochastic Neighbor Embedding - a dimensionality reduction technique that preserves local structure for visualization.
+>
+> *Intuition:* Clusters in t-SNE plots indicate groups of semantically related words. Spread indicates vocabulary diversity; tight clusters suggest semantic coherence.
+>
+> *What to seek:* Meaningful clusters (e.g., numbers together, verbs together). Avoid over-interpreting distances - t-SNE preserves local, not global, structure.
+
+### General Interpretation Guidelines
+
+1. **Compare within model families:** Metrics are most meaningful when comparing models of the same type (e.g., 8k vs 64k tokenizer).
+2. **Consider trade-offs:** Better performance on one metric often comes at the cost of another (e.g., compression vs. OOV rate).
+3. **Context matters:** Optimal values depend on downstream tasks. Text generation may prioritize different metrics than classification.
+4. **Corpus influence:** All metrics are influenced by corpus characteristics. Wikipedia text differs from social media or literature.
+5. **Language-specific patterns:** Morphologically rich languages (like Arabic) may show different optimal ranges than analytic languages.
+
+
+### Visualizations Index
+
+| Visualization | Description |
+|---------------|-------------|
+| Tokenizer Compression | Compression ratios by vocabulary size |
+| Tokenizer Fertility | Average token length by vocabulary |
+| Tokenizer OOV | Unknown token rates |
+| Tokenizer Total Tokens | Total tokens by vocabulary |
+| N-gram Perplexity | Perplexity by n-gram size |
+| N-gram Entropy | Entropy by n-gram size |
+| N-gram Coverage | Top pattern coverage |
+| N-gram Unique | Unique n-gram counts |
+| Markov Entropy | Entropy by context size |
+| Markov Branching | Branching factor by context |
+| Markov Contexts | Unique context counts |
+| Zipf's Law | Frequency-rank distribution with fit |
+| Vocab Frequency | Word frequency distribution |
+| Top 20 Words | Most frequent words |
+| Vocab Coverage | Cumulative coverage curve |
+| Embedding Isotropy | Vector space uniformity |
+| Embedding Norms | Vector magnitude distribution |
+| Embedding Similarity | Word similarity heatmap |
+| Nearest Neighbors | Similar words for key terms |
+| t-SNE Words | 2D word embedding visualization |
+| t-SNE Sentences | 2D sentence embedding visualization |
+| Position Encoding | Encoding method comparison |
+| Model Sizes | Storage requirements |
+| Performance Dashboard | Comprehensive performance overview |
+
+---
+## About This Project
+
+### Data Source
+
+Models trained on [wikipedia-monthly](https://huggingface.co/datasets/omarkamali/wikipedia-monthly) - a monthly snapshot of Wikipedia articles across 300+ languages.
+
+### Project
+
+A project by **[Wikilangs](https://wikilangs.org)** - Open-source NLP models for every Wikipedia language.
+
+### Maintainer
+
+[Omar Kamali](https://omarkamali.com) - [Omneity Labs](https://omneitylabs.com)
+
+### Citation
+
+If you use these models in your research, please cite:
+
+```bibtex
+@misc{wikilangs2025,
+  author = {Kamali, Omar},
+  title = {Wikilangs: Open NLP Models for Wikipedia Languages},
+  year = {2025},
+  publisher = {HuggingFace},
+  url = {https://huggingface.co/wikilangs}
+  institution = {Omneity Labs}
+}
+```
+
+### License
+
+MIT License - Free for academic and commercial use.
+
+### Links
+
+- 🌐 Website: [wikilangs.org](https://wikilangs.org)
+- 🤗 Models: [huggingface.co/wikilangs](https://huggingface.co/wikilangs)
+- 📊 Data: [wikipedia-monthly](https://huggingface.co/datasets/omarkamali/wikipedia-monthly)
+- 👤 Author: [Omar Kamali](https://huggingface.co/omarkamali)
+---
+*Generated by Wikilangs Models Pipeline*
+
+*Report Date: 2025-12-27 16:35:06*

models/embeddings/monolingual/arc_128d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4837cbf17dc8afc088fdf1e21419e03ca91e8ea6df0e96425e80b5d9a487ffcc
+size 1026044955

models/embeddings/monolingual/arc_128d.meta.json

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

models/embeddings/monolingual/arc_128d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "arc",
+  "dimension": 128,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 128,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 1958
+}

models/embeddings/monolingual/arc_32d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:913460969e1240ed6f202c66ac801c20a82745d7723e7bb5674cf6e4e44ce378
+size 256541211

models/embeddings/monolingual/arc_32d.meta.json

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

models/embeddings/monolingual/arc_32d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "arc",
+  "dimension": 32,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 32,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 1958
+}

models/embeddings/monolingual/arc_64d.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:653a7b15a84735fd70f3c904bdb3fbeaeefd0db41e9d318bc421f7789fe732d5
+size 513042459

models/embeddings/monolingual/arc_64d.meta.json

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

models/embeddings/monolingual/arc_64d_metadata.json

@@ -0,0 +1,13 @@
+{
+  "language": "arc",
+  "dimension": 64,
+  "version": "monolingual",
+  "training_params": {
+    "dim": 64,
+    "min_count": 5,
+    "window": 5,
+    "negative": 5,
+    "epochs": 5
+  },
+  "vocab_size": 1958
+}

models/subword_markov/arc_markov_ctx1_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec81048f0b697c5ef8814bf4a212babdecfeffc2132c582028ae9497f4e263d2
+size 52917

models/subword_markov/arc_markov_ctx1_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "subword",
+  "language": "arc",
+  "unique_contexts": 797,
+  "total_transitions": 419545
+}

models/subword_markov/arc_markov_ctx2_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:640b341980a08ec099ab4ada8f63a1d3fe2ab634ed70a1ac9775fb4f48741db8
+size 216906

models/subword_markov/arc_markov_ctx2_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "subword",
+  "language": "arc",
+  "unique_contexts": 6904,
+  "total_transitions": 417586
+}

models/subword_markov/arc_markov_ctx3_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e817885697fb90519987face639656e80f4a711be45aa058aa7deb391c5b2a01
+size 631662

models/subword_markov/arc_markov_ctx3_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "subword",
+  "language": "arc",
+  "unique_contexts": 30176,
+  "total_transitions": 415627
+}

models/subword_markov/arc_markov_ctx4_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ecb5b54df5a44bacfd37832cde268bb71a63c8764fe087dd4679398d7a74c655
+size 1267490

models/subword_markov/arc_markov_ctx4_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "subword",
+  "language": "arc",
+  "unique_contexts": 75950,
+  "total_transitions": 413668
+}

models/subword_ngram/arc_2gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:af4e8cc6802d1add1ff93363c8ccc163a89472ff8b3b69a137ae643d3fe26fc8
+size 31635

models/subword_ngram/arc_2gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "subword",
+  "language": "arc",
+  "unique_ngrams": 2501,
+  "total_ngrams": 419545
+}

models/subword_ngram/arc_3gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6ff6ceec88fedf4d9696b4a775251332e446c4fd0b6b3ad44cecfee8f91db33
+size 148445

models/subword_ngram/arc_3gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "subword",
+  "language": "arc",
+  "unique_ngrams": 11822,
+  "total_ngrams": 417586
+}

models/subword_ngram/arc_4gram_subword.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e13852e81422d6b99190605904715f9839adb8a8b216c7576bb6fd42b19100d8
+size 446279

models/subword_ngram/arc_4gram_subword_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "subword",
+  "language": "arc",
+  "unique_ngrams": 32191,
+  "total_ngrams": 415627
+}

models/tokenizer/arc_tokenizer_16k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:367e3b2a22375ddd3d88817ba1709bd7643670ad0106e2eb141344253a7fb454
+size 551251

models/tokenizer/arc_tokenizer_32k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5823939f3ffb240c668a02808069bb97f33262098d8601e83ea34f1a899be0b5
+size 905690

models/tokenizer/arc_tokenizer_8k.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:67f3e230ce64dd72a320d2684f32479b3bd61e6f0ba27e151727416757e65509
+size 389381

models/vocabulary/arc_vocabulary.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b45cacd75fde3f5334c33a28bfc4333498d0ac3db4b00832e14d4ffa2dff9c9
+size 103543

models/vocabulary/arc_vocabulary_metadata.json

@@ -0,0 +1,16 @@
+{
+  "language": "arc",
+  "vocabulary_size": 6528,
+  "statistics": {
+    "type_token_ratio": 0.2340763088766938,
+    "coverage": {
+      "top_100": 0.2975890140185701,
+      "top_1000": 0.5967515410023668,
+      "top_5000": 0.8110744102577441,
+      "top_10000": 0.8959660849436917
+    },
+    "hapax_count": 11472,
+    "hapax_ratio": 0.6373333333333333,
+    "total_documents": 1959
+  }
+}

models/word_markov/arc_markov_ctx1_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be0301162ceeb31bf05250db417e7a78133aa599465993cff73117e95292164a
+size 575823

models/word_markov/arc_markov_ctx1_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 1,
+  "variant": "word",
+  "language": "arc",
+  "unique_contexts": 18087,
+  "total_transitions": 98674
+}

models/word_markov/arc_markov_ctx2_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:91579b48342a7c4a5e2042ea24e67219434aa6d0e3c1f757801d321393a4642f
+size 1199157

models/word_markov/arc_markov_ctx2_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 2,
+  "variant": "word",
+  "language": "arc",
+  "unique_contexts": 55465,
+  "total_transitions": 96715
+}

models/word_markov/arc_markov_ctx3_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e49b6befa755f07916a3b98b954b006f24cc3d8c685cd363a0d80b6068712fd2
+size 1603718

models/word_markov/arc_markov_ctx3_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 3,
+  "variant": "word",
+  "language": "arc",
+  "unique_contexts": 72203,
+  "total_transitions": 94756
+}

models/word_markov/arc_markov_ctx4_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9a54a5ffb22a9f5f303e36dd991688e4ab528b6254681a949db801002249f59e
+size 1904517

models/word_markov/arc_markov_ctx4_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "context_size": 4,
+  "variant": "word",
+  "language": "arc",
+  "unique_contexts": 78995,
+  "total_transitions": 92798
+}

models/word_ngram/arc_2gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4bfab18a90cce87e63b943074ea22993c759d1f45144ccb7cf6a68f8b6bf8211
+size 35430

models/word_ngram/arc_2gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 2,
+  "variant": "word",
+  "language": "arc",
+  "unique_ngrams": 1994,
+  "total_ngrams": 98674
+}

models/word_ngram/arc_3gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8a7068fff152c1beb637797f6f800103cf8355b1cd9afd9f1e191b20b5a17f3
+size 54386

models/word_ngram/arc_3gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 3,
+  "variant": "word",
+  "language": "arc",
+  "unique_ngrams": 2669,
+  "total_ngrams": 96715
+}

models/word_ngram/arc_4gram_word.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc77168295f018ed264ddcf16570c851b899416ced88b358010acf98e223b687
+size 103678

models/word_ngram/arc_4gram_word_metadata.json

@@ -0,0 +1,7 @@
+{
+  "n": 4,
+  "variant": "word",
+  "language": "arc",
+  "unique_ngrams": 4604,
+  "total_ngrams": 94756
+}