Upload all models and assets for crh (20251001)

README.md

@@ -23,14 +23,14 @@ dataset_info:
 metrics:
   - name: best_compression_ratio
     type: compression
-    value: 4.462
+    value: 4.773
   - name: best_isotropy
     type: isotropy
-    value: 0.7580
+    value: 0.6920
   - name: vocabulary_size
     type: vocab
-    value: 53689
-generated: 2025-12-28
+    value: 0
+generated: 2026-01-03
 ---
 
 # CRH - Wikilangs Models
@@ -44,12 +44,13 @@ We analyze tokenizers, n-gram models, Markov chains, vocabulary statistics, and
 ### Models & Assets
 
 - Tokenizers (8k, 16k, 32k, 64k)
-- N-gram models (2, 3, 4-gram)
-- Markov chains (context of 1, 2, 3 and 4)
+- 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
+- Embeddings in various sizes and dimensions (aligned and unaligned)
 - Language Vocabulary
 - Language Statistics
+
 ![Performance Dashboard](visualizations/performance_dashboard.png)
 
 ### Analysis and Evaluation
@@ -59,7 +60,8 @@ We analyze tokenizers, n-gram models, Markov chains, vocabulary statistics, and
 - [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)
+- [6. Morphological Analysis (Experimental)](#6-morphological-analysis)
+- [7. Summary & Recommendations](#7-summary--recommendations)
 - [Metrics Glossary](#appendix-metrics-glossary--interpretation-guide)
 - [Visualizations Index](#visualizations-index)
 
@@ -68,51 +70,57 @@ We analyze tokenizers, n-gram models, Markov chains, vocabulary statistics, and
 
 ![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** | 3.468x | 3.42 | 0.2080% | 235,121 |
-| **16k** | 3.842x | 3.78 | 0.2305% | 212,188 |
-| **32k** | 4.188x | 4.13 | 0.2512% | 194,672 |
-| **64k** | 4.462x 🏆 | 4.39 | 0.2676% | 182,737 |
+| **8k** | 3.643x | 3.65 | 0.2020% | 214,351 |
+| **16k** | 4.073x | 4.08 | 0.2258% | 191,723 |
+| **32k** | 4.453x | 4.46 | 0.2469% | 175,361 |
+| **64k** | 4.773x 🏆 | 4.78 | 0.2647% | 163,591 |
 
 ### Tokenization Examples
 
 Below are sample sentences tokenized with each vocabulary size:
 
-**Sample 1:** `Novotroyevka () - Rusiyeniñ Belgorod vilâyetinde Koroça rayonında bir köy. Ealis...`
+**Sample 1:** `Şaburovo () - Rusiyeniñ Altay ülkesinde Solton rayonında bir qasaba. Ealisiniñ s...`
 
 | Vocab | Tokens | Count |
 |-------|--------|-------|
-| 8k | `▁novo tr oy evka ▁() ▁- ▁rusiyeniñ ▁belgorod ▁vilâyetinde ▁koroça ... (+17 more)` | 27 |
-| 16k | `▁novotr oy evka ▁() ▁- ▁rusiyeniñ ▁belgorod ▁vilâyetinde ▁koroça ▁rayonında ... (+16 more)` | 26 |
-| 32k | `▁novotroy evka ▁() ▁- ▁rusiyeniñ ▁belgorod ▁vilâyetinde ▁koroça ▁rayonında ▁bir ... (+15 more)` | 25 |
-| 64k | `▁novotroy evka ▁() ▁- ▁rusiyeniñ ▁belgorod ▁vilâyetinde ▁koroça ▁rayonında ▁bir ... (+15 more)` | 25 |
+| 8k | `▁şab urovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁solton ▁rayonında ▁bir ... (+12 more)` | 22 |
+| 16k | `▁şab urovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁solton ▁rayonında ▁bir ... (+12 more)` | 22 |
+| 32k | `▁şab urovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁solton ▁rayonında ▁bir ... (+12 more)` | 22 |
+| 64k | `▁şaburovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁solton ▁rayonında ▁bir ▁qasaba ... (+11 more)` | 21 |
 
-**Sample 2:** `Holodna Balka () - Ukrainanıñ Ades vilâyetinde Ades rayonında bir köy. Ealisiniñ...`
+**Sample 2:** `Polzunovo () - Rusiyeniñ Altay ülkesinde Barnaul şeer bölgesinda bir stantsiya. ...`
 
 | Vocab | Tokens | Count |
 |-------|--------|-------|
-| 8k | `▁hol od na ▁balka ▁() ▁- ▁ukrainanıñ ▁ades ▁vilâyetinde ▁ades ... (+18 more)` | 28 |
-| 16k | `▁hol od na ▁balka ▁() ▁- ▁ukrainanıñ ▁ades ▁vilâyetinde ▁ades ... (+18 more)` | 28 |
-| 32k | `▁hol odna ▁balka ▁() ▁- ▁ukrainanıñ ▁ades ▁vilâyetinde ▁ades ▁rayonında ... (+17 more)` | 27 |
-| 64k | `▁holodna ▁balka ▁() ▁- ▁ukrainanıñ ▁ades ▁vilâyetinde ▁ades ▁rayonında ▁bir ... (+16 more)` | 26 |
+| 8k | `▁pol z unovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁barnaul ▁şeer ... (+17 more)` | 27 |
+| 16k | `▁pol z unovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁barnaul ▁şeer ... (+17 more)` | 27 |
+| 32k | `▁pol z unovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁barnaul ▁şeer ... (+17 more)` | 27 |
+| 64k | `▁polzunovo ▁() ▁- ▁rusiyeniñ ▁altay ▁ülkesinde ▁barnaul ▁şeer ▁bölgesinda ▁bir ... (+15 more)` | 25 |
 
-**Sample 3:** `Tereşpil () - Ukrainanıñ Vinnıtsâ vilâyetinde Hmilnık rayonında bir köy. Ealisin...`
+**Sample 3:** `Bobliv () - Ukrainanıñ Vinnıtsâ vilâyetinde Vinnıtsâ rayonında bir köy. Ealisini...`
 
 | Vocab | Tokens | Count |
 |-------|--------|-------|
-| 8k | `▁ter eş pil ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁hmilnık ▁rayonında ... (+16 more)` | 26 |
-| 16k | `▁tereş pil ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁hmilnık ▁rayonında ▁bir ... (+15 more)` | 25 |
-| 32k | `▁tereş pil ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁hmilnık ▁rayonında ▁bir ... (+15 more)` | 25 |
-| 64k | `▁tereşpil ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁hmilnık ▁rayonında ▁bir ▁köy ... (+14 more)` | 24 |
+| 8k | `▁bob liv ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁vinnıtsâ ▁rayonında ▁bir ... (+12 more)` | 22 |
+| 16k | `▁bob liv ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁vinnıtsâ ▁rayonında ▁bir ... (+12 more)` | 22 |
+| 32k | `▁bob liv ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁vinnıtsâ ▁rayonında ▁bir ... (+12 more)` | 22 |
+| 64k | `▁bobliv ▁() ▁- ▁ukrainanıñ ▁vinnıtsâ ▁vilâyetinde ▁vinnıtsâ ▁rayonında ▁bir ▁köy ... (+11 more)` | 21 |
 
 
 ### Key Findings
 
-- **Best Compression:** 64k achieves 4.462x compression
-- **Lowest UNK Rate:** 8k with 0.2080% unknown tokens
+- **Best Compression:** 64k achieves 4.773x compression
+- **Lowest UNK Rate:** 8k with 0.2020% unknown tokens
 - **Trade-off:** Larger vocabularies improve compression but increase model size
 - **Recommendation:** 32k vocabulary provides optimal balance for production use
 
@@ -121,57 +129,89 @@ Below are sample sentences tokenized with each vocabulary size:
 
 ![N-gram Perplexity](visualizations/ngram_perplexity.png)
 
+![N-gram Unique](visualizations/ngram_unique.png)
+
 ![N-gram Coverage](visualizations/ngram_coverage.png)
 
 ### Results
 
-| N-gram | Perplexity | Entropy | Unique N-grams | Top-100 Coverage | Top-1000 Coverage |
-|--------|------------|---------|----------------|------------------|-------------------|
-| **2-gram** | 1,152 🏆 | 10.17 | 18,186 | 53.8% | 71.2% |
-| **2-gram** | 405 🏆 | 8.66 | 4,814 | 59.9% | 97.2% |
-| **3-gram** | 1,983 | 10.95 | 27,457 | 46.4% | 65.5% |
-| **3-gram** | 2,478 | 11.28 | 35,169 | 32.1% | 69.8% |
-| **4-gram** | 4,721 | 12.21 | 55,382 | 36.7% | 54.6% |
-| **4-gram** | 8,182 | 13.00 | 157,163 | 26.4% | 52.7% |
+| N-gram | Variant | Perplexity | Entropy | Unique N-grams | Top-100 Coverage | Top-1000 Coverage |
+|--------|---------|------------|---------|----------------|------------------|-------------------|
+| **2-gram** | Word | 851 | 9.73 | 10,225 | 56.0% | 74.3% |
+| **2-gram** | Subword | 349 🏆 | 8.45 | 3,905 | 63.4% | 98.0% |
+| **3-gram** | Word | 1,276 | 10.32 | 13,312 | 49.1% | 71.8% |
+| **3-gram** | Subword | 2,227 | 11.12 | 29,302 | 33.0% | 71.7% |
+| **4-gram** | Word | 4,192 | 12.03 | 31,529 | 31.9% | 54.7% |
+| **4-gram** | Subword | 7,868 | 12.94 | 131,434 | 26.0% | 52.2% |
 
 ### Top 5 N-grams by Size
 
-**2-grams:**
+**2-grams (Word):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `ealisiniñ sayısı` | 20,731 |
+| 2 | `rayonında bir` | 17,343 |
+| 3 | `meskün yerler` | 12,883 |
+| 4 | `bir köy` | 10,053 |
+| 5 | `köy ealisiniñ` | 9,130 |
+
+**3-grams (Word):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `rayonında bir köy` | 9,305 |
+| 2 | `köy ealisiniñ sayısı` | 9,130 |
+| 3 | `bir köy ealisiniñ` | 9,130 |
+| 4 | `rayonındaki meskün yerler` | 5,591 |
+| 5 | `kişi meskün yerler` | 4,604 |
+
+**4-grams (Word):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `bir köy ealisiniñ sayısı` | 9,130 |
+| 2 | `rayonında bir köy ealisiniñ` | 8,976 |
+| 3 | `bir köydir ealisiniñ sayısı` | 4,601 |
+| 4 | `rayonında bir köydir ealisiniñ` | 4,565 |
+| 5 | `i̇htar rayonındaki meskün yerler` | 3,615 |
+
+**2-grams (Subword):**
 
 | Rank | N-gram | Count |
 |------|--------|-------|
-| 1 | `kategoriya :` | 32,150 |
-| 2 | `( )` | 23,218 |
-| 3 | `) -` | 21,379 |
-| 4 | `ealisiniñ sayısı` | 20,740 |
-| 5 | `. ealisiniñ` | 20,734 |
+| 1 | `i n` | 101,180 |
+| 2 | `e r` | 95,484 |
+| 3 | `a _` | 88,670 |
+| 4 | `r _` | 84,656 |
+| 5 | `. _` | 80,946 |
 
-**3-grams:**
+**3-grams (Subword):**
 
 | Rank | N-gram | Count |
 |------|--------|-------|
-| 1 | `. ealisiniñ sayısı` | 20,734 |
-| 2 | `( ) -` | 19,732 |
-| 3 | `. kategoriya :` | 16,456 |
-| 4 | `kişi . kategoriya` | 14,755 |
-| 5 | `bir köy .` | 10,054 |
+| 1 | `i ñ _` | 43,476 |
+| 2 | `n i ñ` | 42,980 |
+| 3 | `l e r` | 42,878 |
+| 4 | `n d e` | 35,841 |
+| 5 | `e t i` | 35,623 |
 
-**4-grams:**
+**4-grams (Subword):**
 
 | Rank | N-gram | Count |
 |------|--------|-------|
-| 1 | `kişi . kategoriya :` | 14,755 |
-| 2 | `rayonında bir köy .` | 9,313 |
-| 3 | `( ) - rusiyeniñ` | 9,196 |
-| 4 | `bir köy . ealisiniñ` | 9,139 |
-| 5 | `köy . ealisiniñ sayısı` | 9,139 |
+| 1 | `n i ñ _` | 42,723 |
+| 2 | `i n d e` | 34,201 |
+| 3 | `y e t i` | 30,809 |
+| 4 | `ı n d a` | 30,136 |
+| 5 | `_ b i r` | 29,657 |
 
 
 ### Key Findings
 
-- **Best Perplexity:** 2-gram with 405
+- **Best Perplexity:** 2-gram (subword) with 349
 - **Entropy Trend:** Decreases with larger n-grams (more predictable)
-- **Coverage:** Top-1000 patterns cover ~53% of corpus
+- **Coverage:** Top-1000 patterns cover ~52% of corpus
 - **Recommendation:** 4-gram or 5-gram for best predictive performance
 
 ---
@@ -179,55 +219,86 @@ Below are sample sentences tokenized with each vocabulary size:
 
 ![Markov Entropy](visualizations/markov_entropy.png)
 
+![Markov Contexts](visualizations/markov_contexts.png)
+
 ![Markov Branching](visualizations/markov_branching.png)
 
 ### Results
 
-| Context | Avg Entropy | Perplexity | Branching Factor | Unique Contexts | Predictability |
-|---------|-------------|------------|------------------|-----------------|----------------|
-| **1** | 0.5665 | 1.481 | 3.09 | 136,749 | 43.3% |
-| **1** | 1.1866 | 2.276 | 9.27 | 1,393 | 0.0% |
-| **2** | 0.1665 | 1.122 | 1.39 | 422,458 | 83.3% |
-| **2** | 0.9666 | 1.954 | 5.70 | 12,902 | 3.3% |
-| **3** | 0.0665 | 1.047 | 1.14 | 585,200 | 93.4% |
-| **3** | 0.8194 | 1.765 | 3.81 | 73,541 | 18.1% |
-| **4** | 0.0328 🏆 | 1.023 | 1.07 | 663,395 | 96.7% |
-| **4** | 0.5748 🏆 | 1.489 | 2.40 | 280,480 | 42.5% |
+| Context | Variant | Avg Entropy | Perplexity | Branching Factor | Unique Contexts | Predictability |
+|---------|---------|-------------|------------|------------------|-----------------|----------------|
+| **1** | Word | 0.6260 | 1.543 | 3.00 | 128,781 | 37.4% |
+| **1** | Subword | 0.8874 | 1.850 | 6.86 | 1,506 | 11.3% |
+| **2** | Word | 0.1303 | 1.094 | 1.24 | 384,958 | 87.0% |
+| **2** | Subword | 0.9042 | 1.872 | 5.57 | 10,319 | 9.6% |
+| **3** | Word | 0.0387 | 1.027 | 1.07 | 475,888 | 96.1% |
+| **3** | Subword | 0.8152 | 1.760 | 3.87 | 57,491 | 18.5% |
+| **4** | Word | 0.0241 🏆 | 1.017 | 1.05 | 504,736 | 97.6% |
+| **4** | Subword | 0.6067 | 1.523 | 2.54 | 222,424 | 39.3% |
+
+### Generated Text Samples (Word-based)
+
+Below are text samples generated from each word-based Markov chain model:
+
+**Context Size 1:**
+
+1. `bir köy ealisiniñ sayısı 102 1 ya ayaqlı bir köydir ealisiniñ sayısı kişi vilâyetindeki qasabalar ma...`
+2. `kişi vilâyetindeki meskün yerleri şeer şeklinde qasabalar albörikent accı suv miktarınıñ az olğanına...`
+3. `sayısı 919 kişi vilâyetindeki şeerler boksitogorsk rayonında bir dürki türkiyede oturğan 2 567 kişi ...`
+
+**Context Size 2:**
+
+1. `ealisiniñ sayısı 3 900 kişi muhtar vilâyetindeki şeer şeklinde qasabalar bazarnıy karabulak rayonını...`
+2. `rayonında bir şeer gornozavodsk rayonınıñ merkezi ealisiniñ sayısı 562 kişi vilâyetindeki köyler ray...`
+3. `bir köy ealisiniñ sayısı 952 kişi vilâyetindeki meskün yerler köyler atıflar rayonındaki meskün yerl...`
+
+**Context Size 3:**
+
+1. `rayonında bir köy ealisiniñ sayısı kişi i̇htar rayonındaki meskün yerler köyler atıflar rayonındaki ...`
+2. `köy ealisiniñ sayısı 183 kişi i̇htar rayonındaki meskün yerler köyler atıflar rayonındaki meskün yer...`
+3. `bir köy ealisiniñ sayısı 7 kişi i̇htar rayonındaki meskün yerler şeer şeklinde qasabalar verhovye gl...`
+
+**Context Size 4:**
+
+1. `bir köy ealisiniñ sayısı kişi vilâyetindeki şeer şeklinde qasabalar şeer bölgesindeki meskün yerler ...`
+2. `rayonında bir köy ealisiniñ sayısı 572 kişi i̇htar rayonındaki meskün yerler vilâyetindeki şeer şekl...`
+3. `bir köydir ealisiniñ sayısı 306 kişi meskün yerler`
 
-### Generated Text Samples
 
-Below are text samples generated from each Markov chain model:
+### Generated Text Samples (Subword-based)
+
+Below are text samples generated from each subword-based Markov chain model:
 
 **Context Size 1:**
 
-1. `. menbalar türkiye i ̇ zmalkovo krasnoye rayonında bir qasaba . ealisiniñ sayısı 27 , başqırtistan`
-2. `- rusiyeniñ belgorod vilâyetinde podilsk rayonı ( , cin æmæ jyn kad ! bu ağnıñ esas`
-3. `, latin elifbesiniñ 19 ( ) , cıyınlarda cıyınlarnıñ cıyıntıgı cıyıntıq alına kelgen medinet es herma...`
+1. `_revıñ_moviñmur_`
+2. `a_()._ay.._obeto`
+3. `i_робу_yıñ_()_fi`
 
 **Context Size 2:**
 
-1. `kategoriya : troitskoye rayonındaki meskün yerler kategoriya : primorye ülkesindeki meskün yerler ka...`
-2. `( ) - rusiyeniñ altay ülkesinde şelaboliha rayonında bir köy . ealisiniñ sayısı 47 kişi . kategoriya`
-3. `) - rusiyede , başqırtistan cumhuriyetiniñ miyeke rayonında bir hutor . ealisiniñ sayısı 177 kişi . ...`
+1. `inovaq_ümalmannıñ`
+2. `a_başqırınde_talo`
+3. `r_rus_graynay_the`
 
 **Context Size 3:**
 
-1. `. ealisiniñ sayısı 485 kişi . kategoriya : herson vilâyeti`
-2. `( ) - rusiyeniñ brânsk vilâyetinde karaçev rayonında bir köy . ealisiniñ sayısı 423 kişi . kategoriy...`
-3. `. kategoriya : başqırtistandaki meskün yerler`
+1. `iñ_brânskaya_yañız`
+2. `niñ_araman_sürtüyl`
+3. `nde_bek-tatar_vilâ`
 
 **Context Size 4:**
 
-1. `kişi . kategoriya : ades vilâyetindeki köyler`
-2. `rayonında bir köy . ealisiniñ sayısı 448 kişi . i ̇ htar kategoriya : tahtamukay rayonındaki meskün ...`
-3. `( ) - rusiyeniñ amur vilâyetinde şimanovsk rayonında bir köy . ealisiniñ sayısı 654 kişi . kategoriy...`
+1. `niñ_stepanovskiy_—_`
+2. `inde_ölümlerinen_so`
+3. `yetindeki_meskün_ye`
 
 
 ### Key Findings
 
-- **Best Predictability:** Context-4 with 96.7% predictability
+- **Best Predictability:** Context-4 (word) with 97.6% predictability
 - **Branching Factor:** Decreases with context size (more deterministic)
-- **Memory Trade-off:** Larger contexts require more storage (280,480 contexts)
+- **Memory Trade-off:** Larger contexts require more storage (222,424 contexts)
 - **Recommendation:** Context-3 or Context-4 for text generation
 
 ---
@@ -243,64 +314,64 @@ Below are text samples generated from each Markov chain model:
 
 | Metric | Value |
 |--------|-------|
-| Vocabulary Size | 53,689 |
-| Total Tokens | 889,654 |
-| Mean Frequency | 16.57 |
+| Vocabulary Size | 51,581 |
+| Total Tokens | 778,307 |
+| Mean Frequency | 15.09 |
 | Median Frequency | 3 |
-| Frequency Std Dev | 308.97 |
+| Frequency Std Dev | 271.68 |
 
 ### Most Common Words
 
 | Rank | Word | Frequency |
 |------|------|-----------|
-| 1 | kategoriya | 32,152 |
-| 2 | bir | 27,919 |
-| 3 | kişi | 20,861 |
-| 4 | sayısı | 20,822 |
-| 5 | ealisiniñ | 20,770 |
-| 6 | rayonında | 17,392 |
-| 7 | i | 13,962 |
-| 8 | meskün | 13,507 |
-| 9 | yerler | 12,928 |
-| 10 | vilâyetinde | 12,440 |
+| 1 | bir | 27,780 |
+| 2 | kişi | 20,845 |
+| 3 | sayısı | 20,811 |
+| 4 | ealisiniñ | 20,761 |
+| 5 | rayonında | 17,383 |
+| 6 | meskün | 13,506 |
+| 7 | yerler | 12,928 |
+| 8 | vilâyetinde | 12,431 |
+| 9 | köy | 10,895 |
+| 10 | rusiyeniñ | 9,597 |
 
 ### Least Common Words (from vocabulary)
 
 | Rank | Word | Frequency |
 |------|------|-----------|
-| 1 | зияде | 2 |
-| 2 | atalarnıñ | 2 |
-| 3 | kotsubınskıylar | 2 |
-| 4 | yüneskonıñ | 2 |
-| 5 | اورمودا | 2 |
-| 6 | دیللر | 2 |
-| 7 | ازبری | 2 |
-| 8 | اولان | 2 |
-| 9 | قیزی | 2 |
-| 10 | samançı | 2 |
+| 1 | ekranlar | 2 |
+| 2 | oem | 2 |
+| 3 | macbook | 2 |
+| 4 | mahsuldarlıq | 2 |
+| 5 | planşetler | 2 |
+| 6 | fatemeh | 2 |
+| 7 | movaghar | 2 |
+| 8 | پریسا | 2 |
+| 9 | موقر | 2 |
+| 10 | slammer | 2 |
 
 ### Zipf's Law Analysis
 
 | Metric | Value |
 |--------|-------|
-| Zipf Coefficient | 1.0203 |
-| R² (Goodness of Fit) | 0.996904 |
+| Zipf Coefficient | 0.9849 |
+| R² (Goodness of Fit) | 0.998059 |
 | Adherence Quality | **excellent** |
 
 ### Coverage Analysis
 
 | Top N Words | Coverage |
 |-------------|----------|
-| Top 100 | 46.7% |
-| Top 1,000 | 65.0% |
-| Top 5,000 | 79.6% |
-| Top 10,000 | 85.3% |
+| Top 100 | 45.4% |
+| Top 1,000 | 63.7% |
+| Top 5,000 | 78.1% |
+| Top 10,000 | 84.3% |
 
 ### Key Findings
 
-- **Zipf Compliance:** R²=0.9969 indicates excellent adherence to Zipf's law
-- **High Frequency Dominance:** Top 100 words cover 46.7% of corpus
-- **Long Tail:** 43,689 words needed for remaining 14.7% coverage
+- **Zipf Compliance:** R²=0.9981 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 45.4% of corpus
+- **Long Tail:** 41,581 words needed for remaining 15.7% coverage
 
 ---
 ## 5. Word Embeddings Evaluation
@@ -313,24 +384,116 @@ Below are text samples generated from each Markov chain model:
 
 ![t-SNE Sentences](visualizations/tsne_sentences.png)
 
-### Model Comparison
 
-| Model | Vocab Size | Dimension | Avg Norm | Std Norm | Isotropy |
-|-------|------------|-----------|----------|----------|----------|
-| **mono_32d** | 16,090 | 32 | 4.513 | 0.777 | 0.7580 🏆 |
-| **mono_64d** | 16,090 | 64 | 4.759 | 0.736 | 0.5447 |
-| **mono_128d** | 16,090 | 128 | 4.802 | 0.733 | 0.1564 |
-| **embeddings_enhanced** | 0 | 0 | 0.000 | 0.000 | 0.0000 |
+### 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 | 0.6920 🏆 | 0.3816 | N/A | N/A |
+| **mono_64d** | 64 | 0.4424 | 0.3546 | N/A | N/A |
+| **mono_128d** | 128 | 0.1085 | 0.3496 | N/A | N/A |
 
 ### Key Findings
 
-- **Best Isotropy:** mono_32d with 0.7580 (more uniform distribution)
-- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
-- **Vocabulary Coverage:** All models cover 16,090 words
-- **Recommendation:** 100d for balanced semantic capture and efficiency
+- **Best Isotropy:** mono_32d with 0.6920 (more uniform distribution)
+- **Semantic Density:** Average pairwise similarity of 0.3619. 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.
+
+#### Productive Prefixes
+| Prefix | Examples |
+|--------|----------|
+
+#### Productive Suffixes
+| Suffix | Examples |
+|--------|----------|
+| `-a` | belâyeva, sira, boynuna |
+| `-ka` | tatyanovka, verigovka, mazanka |
+| `-vo` | çkalovo, beketovo, çufarovo |
+| `-vka` | tatyanovka, verigovka, karnauhovka |
+| `-an` | yasaqlağan, başqırtistan, i̇talyan |
+| `-ovo` | çkalovo, beketovo, çufarovo |
+| `-en` | yerlerinden, esitgen, yaratıcılığınen |
+| `-ya` | nesterovskaya, podsosennaya, borzovaya |
+
+### 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.
+
+| Stem | Cohesion | Substitutability | Examples |
+|------|----------|------------------|----------|
+| `rler` | 1.64x | 57 contexts | erler, yerler, kirler |
+| `siye` | 2.07x | 21 contexts | asiye, rusiye, vasiyet |
+| `isin` | 1.60x | 34 contexts | lisin, episine, ekisini |
+| `iniñ` | 1.68x | 26 contexts | eliniñ, aliniñ, öziniñ |
+| `nesi` | 1.66x | 22 contexts | nesir, nesib, nesil |
+| `usiy` | 2.15x | 9 contexts | lusiya, rusiye, hususiy |
+| `eniñ` | 1.77x | 15 contexts | heniñ, seniñ, ekeniñ |
+| `lâye` | 1.89x | 11 contexts | gulâyev, belâyev, vilâyet |
+| `âyet` | 1.89x | 11 contexts | vilâyet, menâyet, şikâyet |
+| `yeti` | 1.62x | 17 contexts | yetim, yetip, yetişe |
+| `sini` | 1.62x | 15 contexts | siniy, sinip, sesini |
+| `tind` | 1.75x | 11 contexts | etinden, betinde, şetinde |
+
+### 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`).
+
+| Word | Suggested Split | Confidence | Stem |
+|------|-----------------|------------|------|
+| çabayevka | **`çaba-ye-vka`** | 6.0 | `çaba` |
+| ruzayevka | **`ruza-ye-vka`** | 6.0 | `ruza` |
+| turmayevo | **`turma-ye-vo`** | 6.0 | `turma` |
+| natalivka | **`natali-vka`** | 4.5 | `natali` |
+| çingizovo | **`çingiz-ovo`** | 4.5 | `çingiz` |
+| krasnoyarovo | **`krasnoyar-ovo`** | 4.5 | `krasnoyar` |
+| kapustinka | **`kapustin-ka`** | 4.5 | `kapustin` |
+| malinovka | **`malino-vka`** | 4.5 | `malino` |
+| soldatovo | **`soldat-ovo`** | 4.5 | `soldat` |
+| kaltımanovo | **`kaltım-an-ovo`** | 3.0 | `kaltım` |
+| balabanovo | **`balab-an-ovo`** | 3.0 | `balab` |
+| olehovskaya | **`olehovs-ka-ya`** | 3.0 | `olehovs` |
+| olşanskaya | **`olşans-ka-ya`** | 3.0 | `olşans` |
+| kıtmanovo | **`kıtm-an-ovo`** | 3.0 | `kıtm` |
+| kropıvenka | **`kropıv-en-ka`** | 3.0 | `kropıv` |
+
+### 6.6 Linguistic Interpretation
+
+> **Automated Insight:**
+The language CRH 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.
 
 ---
-## 6. Summary & Recommendations
+## 7. Summary & Recommendations
 
 ![Performance Dashboard](visualizations/performance_dashboard.png)
 
@@ -338,11 +501,12 @@ Below are text samples generated from each Markov chain model:
 
 | Component | Recommended | Rationale |
 |-----------|-------------|-----------|
-| Tokenizer | **32k BPE** | Best compression (4.46x) with low UNK rate |
-| N-gram | **5-gram** | Lowest perplexity (405) |
-| Markov | **Context-4** | Highest predictability (96.7%) |
+| Tokenizer | **64k BPE** | Best compression (4.77x) |
+| N-gram | **2-gram** | Lowest perplexity (349) |
+| Markov | **Context-4** | Highest predictability (97.6%) |
 | Embeddings | **100d** | Balanced semantic capture and isotropy |
 
+
 ---
 ## Appendix: Metrics Glossary & Interpretation Guide
 
@@ -532,7 +696,8 @@ If you use these models in your research, please cite:
   author = {Kamali, Omar},
   title = {Wikilangs: Open NLP Models for Wikipedia Languages},
   year = {2025},
-  publisher = {HuggingFace},
+  doi = {10.5281/zenodo.18073153},
+  publisher = {Zenodo},
   url = {https://huggingface.co/wikilangs}
   institution = {Omneity Labs}
 }
@@ -548,7 +713,8 @@ MIT License - Free for academic and commercial use.
 - 🤗 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: 2025-12-28 23:15:40*
+*Report Date: 2026-01-03 10:33:02*

models/embeddings/monolingual/crh_128d.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2711de686fbbdccec9b4c232aeb1a2ff6589a0a5ab35d03b8eed6ce02bd7f456
-size 1040761703
+oid sha256:a2c03c90e9321eff6a486a7b259eb6d14069ad97d6e4851aaa1d32c01bb7310b
+size 1039804561

models/embeddings/monolingual/crh_128d_metadata.json

@@ -3,11 +3,13 @@
   "dimension": 128,
   "version": "monolingual",
   "training_params": {
-    "dim": 128,
+    "algorithm": "skipgram",
     "min_count": 5,
     "window": 5,
     "negative": 5,
-    "epochs": 5
+    "epochs": 5,
+    "encoding_method": "rope",
+    "dim": 128
   },
-  "vocab_size": 16090
+  "vocab_size": 15175
 }

models/embeddings/monolingual/crh_32d.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:66a6e8733bef4321f51d9516ccf447352157f4a62af7a17ee0b3077db4528dcf
-size 260404583
+oid sha256:08fc415d5a1b109301770bd6dbcbfbdde2008dd25a25d23f81ace78bc3971a3a
+size 260150161

models/embeddings/monolingual/crh_32d_metadata.json

@@ -3,11 +3,13 @@
   "dimension": 32,
   "version": "monolingual",
   "training_params": {
-    "dim": 32,
+    "algorithm": "skipgram",
     "min_count": 5,
     "window": 5,
     "negative": 5,
-    "epochs": 5
+    "epochs": 5,
+    "encoding_method": "rope",
+    "dim": 32
   },
-  "vocab_size": 16090
+  "vocab_size": 15175
 }

models/embeddings/monolingual/crh_64d.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9d958dcba834b7704a2fa91467c099cd2ee01f0d1d5fed91d7a38303939710c0
-size 520523623
+oid sha256:e5d8136014d664519e3bafeb0cfc668b4993f78425d51bdcf0d9d06734bd0726
+size 520034961

models/embeddings/monolingual/crh_64d_metadata.json

@@ -3,11 +3,13 @@
   "dimension": 64,
   "version": "monolingual",
   "training_params": {
-    "dim": 64,
+    "algorithm": "skipgram",
     "min_count": 5,
     "window": 5,
     "negative": 5,
-    "epochs": 5
+    "epochs": 5,
+    "encoding_method": "rope",
+    "dim": 64
   },
-  "vocab_size": 16090
+  "vocab_size": 15175
 }

models/subword_markov/crh_markov_ctx1_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9de08c51e7f35ddea586397377dd1a690a335be3deafe5034997cfcfdc8fb563
-size 98281
+oid sha256:e2ba0d156d7362837baf91177f2123bb0ccb074f4556c38c22f0d046b19f9805
+size 84372

models/subword_markov/crh_markov_ctx1_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 1,
   "variant": "subword",
   "language": "crh",
-  "unique_contexts": 1393,
-  "total_transitions": 7898659
+  "unique_contexts": 1506,
+  "total_transitions": 6777473
 }

models/subword_markov/crh_markov_ctx2_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e9bb8639872bba7668e3be3688a341970c45c509563b7d7eff3fddaa1b857968
-size 581277
+oid sha256:a42c66f0a162aa7d06896db540b1636552e8a9ebc256f5e6cf8d91c7301d5444
+size 481861

models/subword_markov/crh_markov_ctx2_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 2,
   "variant": "subword",
   "language": "crh",
-  "unique_contexts": 12902,
-  "total_transitions": 7868965
+  "unique_contexts": 10319,
+  "total_transitions": 6748084
 }

models/subword_markov/crh_markov_ctx3_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ab59e75e6595f11a4437338aa658f32767c561497c79e161cf3d4148cdd2c79
-size 2047334
+oid sha256:5a42665c99cf0102347580ad03ad8662c7dcd958de7c0259e8380dcbc939a89f
+size 1701410

models/subword_markov/crh_markov_ctx3_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 3,
   "variant": "subword",
   "language": "crh",
-  "unique_contexts": 73541,
-  "total_transitions": 7839271
+  "unique_contexts": 57491,
+  "total_transitions": 6718695
 }

models/subword_markov/crh_markov_ctx4_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4fcb0e55ae35489db7ded118cd7a233f52f8dd657681364cb8e3372187263e04
-size 5676695
+oid sha256:769a741cb6357905e7c73c65cde5f662a2cbc84d3c2b1870ec910617615ba5a1
+size 4686904

models/subword_markov/crh_markov_ctx4_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 4,
   "variant": "subword",
   "language": "crh",
-  "unique_contexts": 280480,
-  "total_transitions": 7809577
+  "unique_contexts": 222424,
+  "total_transitions": 6689306
 }

models/subword_ngram/crh_2gram_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a1d24c9778380ff73f830ff74e1323c5c9a67b1ad255aca792619d77b25d5912
-size 64028
+oid sha256:d352e2e3b7dd102bd7055d607a37c3738aff67c0c66285e95804ad330500d280
+size 52920

models/subword_ngram/crh_2gram_subword_metadata.json

@@ -2,6 +2,6 @@
   "n": 2,
   "variant": "subword",
   "language": "crh",
-  "unique_ngrams": 4814,
-  "total_ngrams": 7898659
+  "unique_ngrams": 3905,
+  "total_ngrams": 6777473
 }

models/subword_ngram/crh_3gram_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a4fd9693cd6180b77894216a45c8557781e0405949f934ac0db671694231e38f
-size 451805
+oid sha256:1a92a6f76ef54f7e43af0d7e9c762ace522be6627fa4f29ab3df5200c1f2a63d
+size 375416

models/subword_ngram/crh_3gram_subword_metadata.json

@@ -2,6 +2,6 @@
   "n": 3,
   "variant": "subword",
   "language": "crh",
-  "unique_ngrams": 35169,
-  "total_ngrams": 7868965
+  "unique_ngrams": 29302,
+  "total_ngrams": 6748084
 }

models/subword_ngram/crh_4gram_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:45d6866f9f3500b9daf768646045ee8c250bd213b574f0a8fce2c8ef66523e4e
-size 1836873
+oid sha256:0ddc20df05c5929fbc17c37bba729a38c20ac69752548221a46c1e42b8e8359f
+size 1562486

models/subword_ngram/crh_4gram_subword_metadata.json

@@ -2,6 +2,6 @@
   "n": 4,
   "variant": "subword",
   "language": "crh",
-  "unique_ngrams": 157163,
-  "total_ngrams": 7839271
+  "unique_ngrams": 131434,
+  "total_ngrams": 6718695
 }

models/tokenizer/crh_tokenizer_16k.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8bf313aa3863643a3d3a48927e34d71e36c34b5c866db4ec55ff481f2c42cf78
-size 511102
+oid sha256:7252484520841503b3c10a110b3ea4ba3a62f7073a69a342466526a0b31ef51c
+size 513301

models/tokenizer/crh_tokenizer_32k.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a90dac68538f5086942dbcf27b4787b28e5927dcb62fb4edbfd0d3f645b170cd
-size 798788
+oid sha256:60b38160d9a927f6ea17a75ef13bbe991a80a50121d1c845806c967c815488f3
+size 797422

models/tokenizer/crh_tokenizer_64k.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0ca674ae3850ca528644e4e9d7d41feaaf7f45a78b69b9eccd17204ad57d458a
-size 1388807
+oid sha256:8b45034637ab90c0741ce2ac1009165fc7e87f960f1ff87e489cabdedb42a8c8
+size 1425086

models/tokenizer/crh_tokenizer_8k.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dd97a27e642560427bbf6bc2e1ec3da514ec6c83ec093a16660de476f4797974
-size 373198
+oid sha256:414811c3993e9f6351075aaf47ac4ebe0185a4a1e2948c911bbbe25a0c48d8d1
+size 373959

models/vocabulary/crh_vocabulary.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:37995711df1e3c6e65e7b604e037154a533bfd0c17bbed09013a0b12d6d0cc3a
-size 925242
+oid sha256:e154d72a53b3ecac61e12ba483392cfb82a4cc6a0a1b854df1b5a9a58503b2e9
+size 892791

models/vocabulary/crh_vocabulary_metadata.json

@@ -1,16 +1,17 @@
 {
   "language": "crh",
-  "vocabulary_size": 53689,
+  "vocabulary_size": 51581,
+  "variant": "full",
   "statistics": {
-    "type_token_ratio": 0.14047543157959674,
+    "type_token_ratio": 0.1506807050021212,
     "coverage": {
-      "top_100": 0.42677284364428997,
-      "top_1000": 0.594827213933929,
-      "top_5000": 0.727782518841444,
-      "top_10000": 0.7806146474876361
+      "top_100": 0.4134086438841732,
+      "top_1000": 0.5791304225875555,
+      "top_5000": 0.7107789686755324,
+      "top_10000": 0.7672291584127752
     },
-    "hapax_count": 82936,
-    "hapax_ratio": 0.6070338517840805,
-    "total_documents": 29694
+    "hapax_count": 77350,
+    "hapax_ratio": 0.599933297655335,
+    "total_documents": 29389
   }
 }

models/word_markov/crh_markov_ctx1_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:35f93040f44b20143ca6995aa27d0e51be65a8aed5ec256feed368bc396e7b39
-size 4448665
+oid sha256:7aa9bdb634e256e31b79d824958c523b31cd09097f73541b17948bf1af86cd78
+size 4147698

models/word_markov/crh_markov_ctx1_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 1,
   "variant": "word",
   "language": "crh",
-  "unique_contexts": 136749,
-  "total_transitions": 1289772
+  "unique_contexts": 128781,
+  "total_transitions": 826268
 }

models/word_markov/crh_markov_ctx2_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1186f4b8afeb522aea0d59da82796161e681cf8812db7eafe8dce2d594b4cff3
-size 8722410
+oid sha256:4febe8a6192229f809c3e676345cedfae6db5b2e8a8029ea9bd763e24109c749
+size 8051739

models/word_markov/crh_markov_ctx2_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 2,
   "variant": "word",
   "language": "crh",
-  "unique_contexts": 422458,
-  "total_transitions": 1260079
+  "unique_contexts": 384958,
+  "total_transitions": 796879
 }

models/word_markov/crh_markov_ctx3_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:70418ae3723bd86e8fa05bb025daf404c3ba9978b2435ea847cff6c13a4929e8
-size 11645146
+oid sha256:caae72ac1890563cfd5dfa7c47424d891c27b0908ee8f566663d3643381cb21a
+size 10138150

models/word_markov/crh_markov_ctx3_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 3,
   "variant": "word",
   "language": "crh",
-  "unique_contexts": 585200,
-  "total_transitions": 1230387
+  "unique_contexts": 475888,
+  "total_transitions": 767490
 }

models/word_markov/crh_markov_ctx4_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ca323a9f7ec0e2f8d1a95747ea890278db7668783f6c23a6e5fb60385dc416b1
-size 13583842
+oid sha256:b9ec4ee75c44b8904cc99f17faf16f7bea9180258b479d3929a09babcc3ee856
+size 11333545

models/word_markov/crh_markov_ctx4_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 4,
   "variant": "word",
   "language": "crh",
-  "unique_contexts": 663395,
-  "total_transitions": 1200699
+  "unique_contexts": 504736,
+  "total_transitions": 738101
 }

models/word_ngram/crh_2gram_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d99c79cd54a7e225c135d72d01faebb42fc513cea6d991424a374f1c1f37d7b3
-size 271118
+oid sha256:38e4591741f05c6464cdd00c48756cae1ce9ebc68dbb5edc00a18785d0a4b37e
+size 166309

models/word_ngram/crh_2gram_word_metadata.json

@@ -2,6 +2,6 @@
   "n": 2,
   "variant": "word",
   "language": "crh",
-  "unique_ngrams": 18186,
-  "total_ngrams": 1289772
+  "unique_ngrams": 10225,
+  "total_ngrams": 826268
 }

models/word_ngram/crh_3gram_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5273683d72465d19ccff790d2d7b74fa2a21bc70f48fb303660846dc838afdf4
-size 450204
+oid sha256:0d488d4bae7e345da8886938b5756fcf2fa6837a6a001ee13eaf782ae72c20d9
+size 228117

models/word_ngram/crh_3gram_word_metadata.json

@@ -2,6 +2,6 @@
   "n": 3,
   "variant": "word",
   "language": "crh",
-  "unique_ngrams": 27457,
-  "total_ngrams": 1260079
+  "unique_ngrams": 13312,
+  "total_ngrams": 796879
 }

models/word_ngram/crh_4gram_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:dfb14a5090afc83bba2389cd24b93178e9858ededac8c4e8505df8fb240e9d5f
-size 905910
+oid sha256:a9dcb5e222de491d28b07541ff9e5eb152aaafd411b98e66613479e8d5a98129
+size 558302

models/word_ngram/crh_4gram_word_metadata.json

@@ -2,6 +2,6 @@
   "n": 4,
   "variant": "word",
   "language": "crh",
-  "unique_ngrams": 55382,
-  "total_ngrams": 1230387
+  "unique_ngrams": 31529,
+  "total_ngrams": 767490
 }