Upload all models and assets for bbc (20251001)

README.md

@@ -23,14 +23,14 @@ dataset_info:
 metrics:
   - name: best_compression_ratio
     type: compression
-    value: 4.433
+    value: 3.663
   - name: best_isotropy
     type: isotropy
-    value: 0.8253
+    value: 0.8223
   - name: vocabulary_size
     type: vocab
-    value: 24711
-generated: 2025-12-28
+    value: 0
+generated: 2026-01-03
 ---
 
 # BBC - 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)
 
@@ -68,53 +70,53 @@ 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** | 3.867x | 3.83 | 0.1235% | 1,466,873 |
-| **16k** | 4.118x | 4.08 | 0.1315% | 1,377,727 |
-| **32k** | 4.304x | 4.27 | 0.1375% | 1,318,061 |
-| **64k** | 4.433x 🏆 | 4.39 | 0.1416% | 1,279,829 |
+| **8k** | 3.308x | 3.31 | 0.2131% | 1,665,136 |
+| **16k** | 3.527x | 3.53 | 0.2273% | 1,561,615 |
+| **32k** | 3.663x 🏆 | 3.66 | 0.2360% | 1,503,727 |
 
 ### Tokenization Examples
 
 Below are sample sentences tokenized with each vocabulary size:
 
-**Sample 1:** `Panjunan i ma sada huta  na adong di Kecamatan Petarukan, Kabupaten Pemalang, Pr...`
+**Sample 1:** `Pedurungan i ma sada huta na adong di Kecamatan Taman, Kabupaten Pemalang, Propi...`
 
 | Vocab | Tokens | Count |
 |-------|--------|-------|
-| 8k | `▁panj unan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ... (+11 more)` | 21 |
-| 16k | `▁panj unan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ... (+11 more)` | 21 |
-| 32k | `▁panj unan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ... (+11 more)` | 21 |
-| 64k | `▁panjunan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ▁petarukan ... (+10 more)` | 20 |
-
-**Sample 2:** `Ampapaga (Surat Batak:ᯀᯔ᯲ᯇᯇᯎ) i ma sada suansuanan na tubu di gadu ni hauma.
+| 8k | `▁ped ur ungan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ... (+12 more)` | 22 |
+| 16k | `▁pedurungan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ▁taman ... (+10 more)` | 20 |
+| 32k | `▁pedurungan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ▁taman ... (+10 more)` | 20 |
 
- P...`
+**Sample 2:** `Mulyoharjo i ma sada Kelurahan na adong di Kecamatan Pemalang, Kabupaten Pemalan...`
 
 | Vocab | Tokens | Count |
 |-------|--------|-------|
-| 8k | `▁amp ap aga ▁( surat ▁batak : ᯀ ᯔ᯲ ᯇ ... (+20 more)` | 30 |
-| 16k | `▁amp ap aga ▁( surat ▁batak : ᯀ ᯔ᯲ᯇ ᯇ ... (+18 more)` | 28 |
-| 32k | `▁amp apaga ▁( surat ▁batak : ᯀ ᯔ᯲ᯇᯇᯎ ) ▁i ... (+15 more)` | 25 |
-| 64k | `▁ampapaga ▁( surat ▁batak : ᯀᯔ᯲ᯇᯇᯎ ) ▁i ▁ma ▁sada ... (+13 more)` | 23 |
+| 8k | `▁mul y oharjo ▁i ▁ma ▁sada ▁kelurahan ▁na ▁adong ▁di ... (+12 more)` | 22 |
+| 16k | `▁mulyoharjo ▁i ▁ma ▁sada ▁kelurahan ▁na ▁adong ▁di ▁kecamatan ▁pemalang ... (+10 more)` | 20 |
+| 32k | `▁mulyoharjo ▁i ▁ma ▁sada ▁kelurahan ▁na ▁adong ▁di ▁kecamatan ▁pemalang ... (+10 more)` | 20 |
 
-**Sample 3:** `Sungapan i ma sada huta  na adong di Kecamatan Pemalang, Kabupaten Pemalang, Pro...`
+**Sample 3:** `Klegen i ma sada huta na adong di Kecamatan Comal, Kabupaten Pemalang, Propinsi ...`
 
 | Vocab | Tokens | Count |
 |-------|--------|-------|
-| 8k | `▁sung apan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ... (+11 more)` | 21 |
-| 16k | `▁sung apan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ... (+11 more)` | 21 |
-| 32k | `▁sung apan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ... (+11 more)` | 21 |
-| 64k | `▁sungapan ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ▁pemalang ... (+10 more)` | 20 |
+| 8k | `▁kl egen ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ... (+11 more)` | 21 |
+| 16k | `▁klegen ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ▁comal ... (+10 more)` | 20 |
+| 32k | `▁klegen ▁i ▁ma ▁sada ▁huta ▁na ▁adong ▁di ▁kecamatan ▁comal ... (+10 more)` | 20 |
 
 
 ### Key Findings
 
-- **Best Compression:** 64k achieves 4.433x compression
-- **Lowest UNK Rate:** 8k with 0.1235% unknown tokens
+- **Best Compression:** 32k achieves 3.663x compression
+- **Lowest UNK Rate:** 8k with 0.2131% unknown tokens
 - **Trade-off:** Larger vocabularies improve compression but increase model size
 - **Recommendation:** 32k vocabulary provides optimal balance for production use
 
@@ -123,57 +125,89 @@ 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** | 7,191 🏆 | 12.81 | 30,496 | 17.8% | 49.6% |
-| **2-gram** | 209 🏆 | 7.70 | 2,868 | 75.2% | 99.1% |
-| **3-gram** | 25,989 | 14.67 | 62,993 | 8.7% | 26.3% |
-| **3-gram** | 1,399 | 10.45 | 19,851 | 36.8% | 80.6% |
-| **4-gram** | 63,619 | 15.96 | 114,847 | 4.7% | 15.6% |
-| **4-gram** | 6,375 | 12.64 | 81,359 | 18.9% | 52.8% |
+| N-gram | Variant | Perplexity | Entropy | Unique N-grams | Top-100 Coverage | Top-1000 Coverage |
+|--------|---------|------------|---------|----------------|------------------|-------------------|
+| **2-gram** | Word | 8,528 | 13.06 | 26,426 | 17.5% | 42.8% |
+| **2-gram** | Subword | 185 🏆 | 7.53 | 3,491 | 77.6% | 99.2% |
+| **3-gram** | Word | 22,579 | 14.46 | 43,165 | 8.3% | 25.2% |
+| **3-gram** | Subword | 1,219 | 10.25 | 18,183 | 38.1% | 83.2% |
+| **4-gram** | Word | 44,749 | 15.45 | 67,595 | 5.7% | 16.0% |
+| **4-gram** | Subword | 5,604 | 12.45 | 70,417 | 19.7% | 54.7% |
 
 ### Top 5 N-grams by Size
 
-**2-grams:**
+**2-grams (Word):**
 
 | Rank | N-gram | Count |
 |------|--------|-------|
-| 1 | `, jala` | 8,780 |
-| 2 | `i ,` | 6,997 |
-| 3 | `ᯬ ᯲` | 4,573 |
-| 4 | `angka na` | 4,409 |
-| 5 | `dung i` | 4,328 |
+| 1 | `angka na` | 4,424 |
+| 2 | `dung i` | 4,327 |
+| 3 | `ni si` | 4,061 |
+| 4 | `i ma` | 3,622 |
+| 5 | `ni jahowa` | 2,892 |
 
-**3-grams:**
+**3-grams (Word):**
 
 | Rank | N-gram | Count |
 |------|--------|-------|
-| 1 | `anak ni si` | 1,611 |
-| 2 | `, angka na` | 1,528 |
-| 3 | `. 2 :` | 1,079 |
-| 4 | `, anak ni` | 1,069 |
-| 5 | `ᯰ ᯄ ᯦` | 1,063 |
+| 1 | `anak ni si` | 1,613 |
+| 2 | `dung i ninna` | 735 |
+| 3 | `i ma sada` | 728 |
+| 4 | `hata ni jahowa` | 703 |
+| 5 | `na adong di` | 690 |
 
-**4-grams:**
+**4-grams (Word):**
 
 | Rank | N-gram | Count |
 |------|--------|-------|
-| 1 | `, anak ni si` | 906 |
-| 2 | `ᯀ ᯰ ᯄ ᯦` | 686 |
-| 3 | `ᯘ ᯪ ᯀᯉ ᯲` | 457 |
-| 4 | `ᯬ ᯂᯖ ᯬ ᯲` | 432 |
-| 5 | `on do hata ni` | 421 |
+| 1 | `on do hata ni` | 423 |
+| 2 | `songon on do hata` | 408 |
+| 3 | `i ma sada huta` | 363 |
+| 4 | `angka anak ni si` | 336 |
+| 5 | `na adong di kecamatan` | 297 |
+
+**2-grams (Subword):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `a _` | 206,904 |
+| 2 | `a n` | 205,541 |
+| 3 | `n g` | 154,122 |
+| 4 | `i _` | 143,001 |
+| 5 | `n a` | 122,611 |
+
+**3-grams (Subword):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `a n g` | 81,985 |
+| 2 | `_ m a` | 76,327 |
+| 3 | `n a _` | 58,974 |
+| 4 | `_ n a` | 53,547 |
+| 5 | `a n _` | 51,343 |
+
+**4-grams (Subword):**
+
+| Rank | N-gram | Count |
+|------|--------|-------|
+| 1 | `_ n i _` | 34,971 |
+| 2 | `_ n a _` | 33,600 |
+| 3 | `_ d i _` | 25,982 |
+| 4 | `a n g k` | 24,957 |
+| 5 | `_ m a _` | 23,771 |
 
 
 ### Key Findings
 
-- **Best Perplexity:** 2-gram with 209
+- **Best Perplexity:** 2-gram (subword) with 185
 - **Entropy Trend:** Decreases with larger n-grams (more predictable)
-- **Coverage:** Top-1000 patterns cover ~53% of corpus
+- **Coverage:** Top-1000 patterns cover ~55% of corpus
 - **Recommendation:** 4-gram or 5-gram for best predictive performance
 
 ---
@@ -181,55 +215,86 @@ 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** | 0.8281 | 1.775 | 6.09 | 50,531 | 17.2% |
-| **1** | 1.0960 | 2.138 | 7.35 | 1,187 | 0.0% |
-| **2** | 0.3983 | 1.318 | 2.17 | 307,642 | 60.2% |
-| **2** | 0.8091 | 1.752 | 4.70 | 8,725 | 19.1% |
-| **3** | 0.1982 | 1.147 | 1.41 | 667,901 | 80.2% |
-| **3** | 0.7878 | 1.726 | 3.57 | 41,004 | 21.2% |
-| **4** | 0.0988 🏆 | 1.071 | 1.16 | 943,940 | 90.1% |
-| **4** | 0.5526 🏆 | 1.467 | 2.40 | 146,535 | 44.7% |
+| Context | Variant | Avg Entropy | Perplexity | Branching Factor | Unique Contexts | Predictability |
+|---------|---------|-------------|------------|------------------|-----------------|----------------|
+| **1** | Word | 0.9188 | 1.890 | 6.44 | 50,697 | 8.1% |
+| **1** | Subword | 0.9378 | 1.916 | 7.17 | 1,435 | 6.2% |
+| **2** | Word | 0.3742 | 1.296 | 2.02 | 325,909 | 62.6% |
+| **2** | Subword | 0.7095 | 1.635 | 4.04 | 10,290 | 29.0% |
+| **3** | Word | 0.1535 | 1.112 | 1.28 | 658,447 | 84.6% |
+| **3** | Subword | 0.6445 | 1.563 | 3.15 | 41,529 | 35.5% |
+| **4** | Word | 0.0590 🏆 | 1.042 | 1.09 | 840,046 | 94.1% |
+| **4** | Subword | 0.5191 | 1.433 | 2.39 | 130,734 | 48.1% |
 
-### Generated Text Samples
+### Generated Text Samples (Word-based)
 
-Below are text samples generated from each Markov chain model:
+Below are text samples generated from each word-based Markov chain model:
 
 **Context Size 1:**
 
-1. `, jala tu si wasti , jala dipadeakdeak hamu sian i jala peakkononna tu palaspalas pamurunan`
-2. `. 12 : songon i hatangku tu bagasan saluhut angka ari puasa bintang di jolo i`
-3. `: ida ma angka na margoar milo dohot paredangedangan , 2 : 35 : 25 pelean`
+1. `ni harbangan asa dioloi si ferdinand lumban gaol boi manampung 1 11 3 naung disunat 2`
+2. `na marumur sataon na humaliang 5 000 m2 hira hira songon sondang ni raja iii diida`
+3. `i gok daupa sian si daud sian pangasammu do sarita pardapot ni na sampuludua i 22`
 
 **Context Size 2:**
 
-1. `, jala tudoshon gumora angka anak ni si joab soara ni sarune i laho ma ho ,`
-2. `i , naung pinauli ni tangan ni halak pangarupa umpogo upa ni pambahenannasida . 99 : 7`
-3. `ᯬ ᯲ ᯘᯞ ᯮ ᯂᯖ ᯮ ᯲ ᯑ ᯩ ᯇᯔ ᯲ ᯅᯂ ᯩ ᯉᯉ ᯲ ᯔ ᯉ`
+1. `angka na niuhir dohot na tarulang angka bagasnasida jala ndang marnalemba 10 38 ingkon mago roham ba...`
+2. `dung i ninna jesus ma siseanna i ninna parompuan na mabalu disi marmudu ho 17 2 dung`
+3. `ni si beor pangarunding i dibunu halak daniel 6 6 1 hamu pe ditanda sada pangituai parheheon`
 
 **Context Size 3:**
 
-1. `anak ni si ahilud do panuturi . 18 : 13 tung sura ahu parohon begu masa tu luat`
-2. `, angka na so bangso hian ; marhite sian bangso na asing , jala marbarita goarmu di betlehem`
-3. `. 2 : 15 alai tarrimas situtu ma si abner dohot di halak na marroha pangansion , ndang`
+1. `anak ni si jaasia si beno 24 27 ia angka anak ni si aron ma mudar i jala`
+2. `dung i ninna ibana tu ahu hombar tu bagabagam 119 171 sai marbulakkon pujipujian ma angka bibirhu al...`
+3. `i ma sada kecamatan na adong di kecamatan petarukan kabupaten pemalang propinsi jawa tonga indonesia...`
 
 **Context Size 4:**
 
-1. `, anak ni si ammiel . 3 : 6 ndada tu torop bangso , angka parhata bobang , manang`
-2. `ᯀ ᯰ ᯄ ᯦ ᯂᯖ ᯉ ᯪ ᯘ ᯪ ᯅ ᯩ ᯀ ᯩ ᯒ ᯪ ᯑ ᯪ ᯉᯘ ᯪ`
-3. `ᯘ ᯪ ᯀᯉ ᯲ ᯂᯔ ᯪ ᯐ ᯮ ᯔ ᯬ ᯞ ᯬ ᯉ ᯪ ᯑ ᯩ ᯅᯖ 2 :`
+1. `on do hata ni tuhan jahowa ida ma ahu sandiri pahehehon tu nasida hahisaron dohot hamalumon jala pam...`
+2. `songon on do hata ni jahowa zebaot tu hamu malim angka na palea goarhu hape lam didok hamu do`
+3. `i ma sada huta na maringanan di kecamatan tarutung kabupaten tapanuli utara propinsi sumatera utara ...`
+
+
+### Generated Text Samples (Subword-based)
+
+Below are text samples generated from each subword-based Markov chain model:
+
+**Context Size 1:**
+
+1. `_nobedi_anoa?_hi`
+2. `akoni_jarin_a_ng`
+3. `nabomaholalowap_`
+
+**Context Size 2:**
+
+1. `a_sia_ahaan_rohot`
+2. `anahu:_tana_raela`
+3. `ngon_nak_i._10:2_`
+
+**Context Size 3:**
+
+1. `anggo_ia_ingka_jor`
+2. `_marik_marhabus;_a`
+3. `na_5_menjangkup_he`
+
+**Context Size 4:**
+
+1. `_ni_angka_halak_juj`
+2. `_na_hian_gabe_manan`
+3. `_di_bagaska_indones`
 
 
 ### Key Findings
 
-- **Best Predictability:** Context-4 with 90.1% predictability
+- **Best Predictability:** Context-4 (word) with 94.1% predictability
 - **Branching Factor:** Decreases with context size (more deterministic)
-- **Memory Trade-off:** Larger contexts require more storage (146,535 contexts)
+- **Memory Trade-off:** Larger contexts require more storage (130,734 contexts)
 - **Recommendation:** Context-3 or Context-4 for text generation
 
 ---
@@ -245,64 +310,64 @@ Below are text samples generated from each Markov chain model:
 
 | Metric | Value |
 |--------|-------|
-| Vocabulary Size | 24,711 |
-| Total Tokens | 1,019,541 |
-| Mean Frequency | 41.26 |
+| Vocabulary Size | 24,970 |
+| Total Tokens | 972,166 |
+| Mean Frequency | 38.93 |
 | Median Frequency | 4 |
-| Frequency Std Dev | 565.94 |
+| Frequency Std Dev | 557.36 |
 
 ### Most Common Words
 
 | Rank | Word | Frequency |
 |------|------|-----------|
-| 1 | ni | 35,101 |
-| 2 | na | 34,088 |
-| 3 | i | 33,056 |
-| 4 | ma | 26,769 |
-| 5 | di | 26,053 |
-| 6 | tu | 20,450 |
-| 7 | do | 19,163 |
-| 8 | angka | 17,428 |
-| 9 | jala | 14,598 |
-| 10 | dohot | 13,609 |
+| 1 | ni | 35,042 |
+| 2 | na | 33,939 |
+| 3 | i | 32,856 |
+| 4 | ma | 26,602 |
+| 5 | di | 26,003 |
+| 6 | tu | 20,420 |
+| 7 | do | 19,118 |
+| 8 | angka | 17,417 |
+| 9 | jala | 14,585 |
+| 10 | dohot | 13,546 |
 
 ### Least Common Words (from vocabulary)
 
 | Rank | Word | Frequency |
 |------|------|-----------|
-| 1 | ᯝᯇᯞ | 2 |
-| 2 | kayo | 2 |
-| 3 | uttar | 2 |
-| 4 | ltr | 2 |
-| 5 | ebrima | 2 |
-| 6 | 290px | 2 |
-| 7 | td | 2 |
-| 8 | height | 2 |
-| 9 | 260px | 2 |
-| 10 | 22251 | 2 |
+| 1 | continua | 2 |
+| 2 | giuseppe | 2 |
+| 3 | mamutuskan | 2 |
+| 4 | disidang | 2 |
+| 5 | disuspensi | 2 |
+| 6 | formula | 2 |
+| 7 | dibenarkan | 2 |
+| 8 | pidana | 2 |
+| 9 | piazza | 2 |
+| 10 | fontana | 2 |
 
 ### Zipf's Law Analysis
 
 | Metric | Value |
 |--------|-------|
-| Zipf Coefficient | 1.1956 |
-| R² (Goodness of Fit) | 0.996705 |
+| Zipf Coefficient | 1.1798 |
+| R² (Goodness of Fit) | 0.997075 |
 | Adherence Quality | **excellent** |
 
 ### Coverage Analysis
 
 | Top N Words | Coverage |
 |-------------|----------|
-| Top 100 | 52.8% |
-| Top 1,000 | 78.6% |
-| Top 5,000 | 91.7% |
-| Top 10,000 | 95.9% |
+| Top 100 | 53.7% |
+| Top 1,000 | 78.4% |
+| Top 5,000 | 91.4% |
+| Top 10,000 | 95.7% |
 
 ### Key Findings
 
-- **Zipf Compliance:** R²=0.9967 indicates excellent adherence to Zipf's law
-- **High Frequency Dominance:** Top 100 words cover 52.8% of corpus
-- **Long Tail:** 14,711 words needed for remaining 4.1% coverage
+- **Zipf Compliance:** R²=0.9971 indicates excellent adherence to Zipf's law
+- **High Frequency Dominance:** Top 100 words cover 53.7% of corpus
+- **Long Tail:** 14,970 words needed for remaining 4.3% coverage
 
 ---
 ## 5. Word Embeddings Evaluation
@@ -315,24 +380,134 @@ 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** | 15,079 | 32 | 3.458 | 0.818 | 0.8253 🏆 |
-| **mono_64d** | 15,079 | 64 | 3.886 | 0.738 | 0.7641 |
-| **mono_128d** | 15,079 | 128 | 4.143 | 0.691 | 0.4668 |
-| **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 | 0.8223 🏆 | 0.3239 | N/A | N/A |
+| **mono_64d** | 64 | 0.7605 | 0.2710 | N/A | N/A |
+| **mono_128d** | 128 | 0.4652 | 0.2352 | N/A | N/A |
 
 ### Key Findings
 
-- **Best Isotropy:** mono_32d with 0.8253 (more uniform distribution)
-- **Dimension Trade-off:** Higher dimensions capture more semantics but reduce isotropy
-- **Vocabulary Coverage:** All models cover 15,079 words
-- **Recommendation:** 100d for balanced semantic capture and efficiency
+- **Best Isotropy:** mono_32d with 0.8223 (more uniform distribution)
+- **Semantic Density:** Average pairwise similarity of 0.2767. Lower values indicate better semantic separation.
+- **Alignment Quality:** No aligned models evaluated in this run.
+- **Recommendation:** 128d aligned for best cross-lingual performance
 
 ---
-## 6. Summary & Recommendations
+## 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 |
+|--------|----------|
+| `-pa` | paunsathon, paraguay, panimbukbuk |
+| `-ma` | matan, marsogotna, marimbang |
+| `-di` | dihalungunhon, dipajonok, dikunjungi |
+| `-mar` | marsogotna, marimbang, marsuhat |
+| `-si` | sintuana, simalolongna, siotihotik |
+| `-man` | manongoshon, maneat, manongos |
+| `-par` | paraguay, paransis, parmaraan |
+| `-ha` | haro, hadoboon, hakristenon |
+
+#### Productive Suffixes
+| Suffix | Examples |
+|--------|----------|
+| `-n` | matan, kanan, paunsathon |
+| `-a` | bulanda, tanomanmuna, musikna |
+| `-on` | paunsathon, manongoshon, dihalungunhon |
+| `-na` | tanomanmuna, musikna, marsogotna |
+| `-an` | matan, kanan, tibetan |
+| `-ng` | marimbang, palding, lumeleng |
+| `-hon` | paunsathon, manongoshon, dihalungunhon |
+| `-nna` | anginna, binoanna, parumaenna |
+
+### 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 |
+|------|----------|------------------|----------|
+| `anga` | 1.65x | 126 contexts | angan, sanga, langa |
+| `angk` | 1.46x | 153 contexts | angka, rangka, dangka |
+| `mang` | 1.72x | 61 contexts | amang, damang, mangae |
+| `ngka` | 1.53x | 87 contexts | angka, rangka, dangka |
+| `ngko` | 1.76x | 41 contexts | ingkon, angkot, tingko |
+| `onga` | 1.75x | 36 contexts | longa, tonga, dongan |
+| `angg` | 1.39x | 75 contexts | anggo, anggi, anggia |
+| `anna` | 1.73x | 31 contexts | hanna, manna, annai |
+| `bahe` | 1.78x | 26 contexts | bahen, dibahe, ibahen |
+| `ingk` | 1.43x | 59 contexts | ingkon, tingki, lingka |
+| `ngan` | 1.37x | 65 contexts | ingan, angan, dongan |
+| `ndan` | 1.68x | 25 contexts | ndang, pandan, undang |
+
+### 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.
+
+| Prefix | Suffix | Frequency | Examples |
+|--------|--------|-----------|----------|
+| `-pa` | `-n` | 372 words | paboaon, pangalelaon |
+| `-ma` | `-n` | 227 words | malungun, mambahen |
+| `-pa` | `-on` | 208 words | paboaon, pangalelaon |
+| `-pa` | `-a` | 196 words | pangkeannasida, padanna |
+| `-pa` | `-an` | 162 words | pamalian, parmiahan |
+| `-di` | `-n` | 135 words | diparsahitan, dison |
+| `-ma` | `-on` | 127 words | mangkalungunhon, mangkasogohon |
+| `-pa` | `-na` | 124 words | padanna, pandokna |
+| `-ha` | `-n` | 124 words | harun, hamulian |
+| `-di` | `-on` | 113 words | dison, ditahbishon |
+
+### 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 |
+|------|-----------------|------------|------|
+| dipahatahata | **`di-pa-ha-ta-hata`** | 9.0 | `hata` |
+| marhatomanon | **`mar-ha-toman-on`** | 7.5 | `toman` |
+| panimbangan | **`pan-imba-ng-an`** | 7.5 | `imba` |
+| patongonhon | **`pa-tong-on-hon`** | 7.5 | `tong` |
+| hatigoranku | **`ha-tigor-an-ku`** | 7.5 | `tigor` |
+| pargogoanku | **`par-gogo-an-ku`** | 7.5 | `gogo` |
+| hagaleonku | **`ha-gale-on-ku`** | 7.5 | `gale` |
+| dipatongon | **`di-pa-tong-on`** | 7.5 | `tong` |
+| taparrohahon | **`ta-par-roha-hon`** | 7.5 | `roha` |
+| parhaporseaon | **`par-ha-porsea-on`** | 7.5 | `porsea` |
+| hamuliaonku | **`ha-mulia-on-ku`** | 7.5 | `mulia` |
+| paluhutonku | **`pa-luhut-on-ku`** | 7.5 | `luhut` |
+| hamateanna | **`ha-ma-tean-na`** | 7.5 | `tean` |
+| silehononku | **`si-lehon-on-ku`** | 7.5 | `lehon` |
+| patoltolonku | **`pa-toltol-on-ku`** | 7.5 | `toltol` |
+
+### 6.6 Linguistic Interpretation
+
+> **Automated Insight:**
+The language BBC 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)
 
@@ -340,11 +515,12 @@ Below are text samples generated from each Markov chain model:
 
 | Component | Recommended | Rationale |
 |-----------|-------------|-----------|
-| Tokenizer | **32k BPE** | Best compression (4.43x) with low UNK rate |
-| N-gram | **5-gram** | Lowest perplexity (209) |
-| Markov | **Context-4** | Highest predictability (90.1%) |
+| Tokenizer | **32k BPE** | Best compression (3.66x) |
+| N-gram | **2-gram** | Lowest perplexity (185) |
+| Markov | **Context-4** | Highest predictability (94.1%) |
 | Embeddings | **100d** | Balanced semantic capture and isotropy |
 
+
 ---
 ## Appendix: Metrics Glossary & Interpretation Guide
 
@@ -534,7 +710,8 @@ 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}
 }
@@ -550,7 +727,8 @@ 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 00:12:38*
+*Report Date: 2026-01-03 06:19:26*

models/embeddings/monolingual/bbc_128d.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0b0ca75d42a296503b6510e66321ba63fde51730d05e8ae06bf9ee333ebf3764
-size 1039704692
+oid sha256:299d624499ec82c937aa1769fa092d3bd9d976c6d08c2af9b22f767740074a59
+size 1039420296

models/embeddings/monolingual/bbc_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": 15079
+  "vocab_size": 14806
 }

models/embeddings/monolingual/bbc_32d.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e806c00c5217f8c0bc9c0c7c7f674b5091cf2b888f0f09d5617c5947fdf7b85
-size 260124020
+oid sha256:981df92afd62ee12261b51c6b8c02f4b1e76c4f2d6c299c0dd0e78da12e9ac2b
+size 260049288

models/embeddings/monolingual/bbc_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": 15079
+  "vocab_size": 14806
 }

models/embeddings/monolingual/bbc_64d.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b313b52ddbda268ad3cbd1f0e06c19c9d337a7809b8be819492b088a18d3ea52
-size 519984244
+oid sha256:e7fddeff473650e7228b6c4a71a4299b4d9ee94908cca09fc7492ca327d34b12
+size 519839624

models/embeddings/monolingual/bbc_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": 15079
+  "vocab_size": 14806
 }

models/subword_markov/bbc_markov_ctx1_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:31f366e845fdd50b876fa026287e3adc40435cfcd969f38d0a43a5e37677bf6f
-size 67861
+oid sha256:5aefbfb1171b4b00a844e78d07315efbcf62f55a2b30f5bd6d940b5ba83804d7
+size 81484

models/subword_markov/bbc_markov_ctx1_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 1,
   "variant": "subword",
   "language": "bbc",
-  "unique_contexts": 1187,
-  "total_transitions": 5965783
+  "unique_contexts": 1435,
+  "total_transitions": 5702376
 }

models/subword_markov/bbc_markov_ctx2_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b7007b2743717f219549dc35da4e22ef904f954017d220c8e61b57e8f201260f
-size 331323
+oid sha256:3f38ba522fe1bac3abb75948ca7e8ea124307971bfee224a267114a3c9466873
+size 358445

models/subword_markov/bbc_markov_ctx2_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 2,
   "variant": "subword",
   "language": "bbc",
-  "unique_contexts": 8725,
-  "total_transitions": 5964472
+  "unique_contexts": 10290,
+  "total_transitions": 5701163
 }

models/subword_markov/bbc_markov_ctx3_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c82b65a7ab369bc14ca7c650a4752310ee46abd1500b7705572d13c633829e72
-size 1156768
+oid sha256:0a3da300cf244b2ab02f8cc4190a829feb85819df2984a6b1a78f37ab7daec14
+size 1190319

models/subword_markov/bbc_markov_ctx3_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 3,
   "variant": "subword",
   "language": "bbc",
-  "unique_contexts": 41004,
-  "total_transitions": 5963161
+  "unique_contexts": 41529,
+  "total_transitions": 5699950
 }

models/subword_markov/bbc_markov_ctx4_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7f8adaacc604505a5ecc9a9df0ee1360e764fe6a44233577ea34de523534262e
-size 2961725
+oid sha256:32b0388a33d7717a0deb5945264197ca362aded134c2c728e20da7a5ddee82f4
+size 2842522

models/subword_markov/bbc_markov_ctx4_subword_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 4,
   "variant": "subword",
   "language": "bbc",
-  "unique_contexts": 146535,
-  "total_transitions": 5961850
+  "unique_contexts": 130734,
+  "total_transitions": 5698737
 }

models/subword_ngram/bbc_2gram_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e60fc5bcc5851afa5f2e2cc336875cafe4bf25d188866759e33cdfd81d92110
-size 36155
+oid sha256:d706d757ef98f7e2eaf523f46f0a602b8f06b25cb1d00579847f22b28eece7f1
+size 46126

models/subword_ngram/bbc_2gram_subword_metadata.json

@@ -2,6 +2,6 @@
   "n": 2,
   "variant": "subword",
   "language": "bbc",
-  "unique_ngrams": 2868,
-  "total_ngrams": 5965783
+  "unique_ngrams": 3491,
+  "total_ngrams": 5702376
 }

models/subword_ngram/bbc_3gram_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1a241f736f6533859dd3ceeb25dc4c326e50cb9bdae44188ae81918ec9b1e09d
-size 236441
+oid sha256:578bacf93c61827c89b4f8427603b775cd6b7891f9ceecc9c90545263f0f5768
+size 237045

models/subword_ngram/bbc_3gram_subword_metadata.json

@@ -2,6 +2,6 @@
   "n": 3,
   "variant": "subword",
   "language": "bbc",
-  "unique_ngrams": 19851,
-  "total_ngrams": 5964472
+  "unique_ngrams": 18183,
+  "total_ngrams": 5701163
 }

models/subword_ngram/bbc_4gram_subword.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:de307e6c44b633e8d607ebdbd329107fa1f51230fa199d5272b932992601e309
-size 929802
+oid sha256:66675bb6688469ac3db798effd708b7b3e5f75e0f119c579a8f76b3e2b98fa07
+size 867430

models/subword_ngram/bbc_4gram_subword_metadata.json

@@ -2,6 +2,6 @@
   "n": 4,
   "variant": "subword",
   "language": "bbc",
-  "unique_ngrams": 81359,
-  "total_ngrams": 5963161
+  "unique_ngrams": 70417,
+  "total_ngrams": 5699950
 }

models/tokenizer/bbc_tokenizer_16k.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:136cf9b48df47bd0325da4d7c88b57560631e30b6199a6208f4260be8e9552ee
-size 526133
+oid sha256:6a176fe5ca0bdc5334c45eb11b39e17d285bd6a1469ac9c9df2ca41a79f9a60c
+size 510189

models/tokenizer/bbc_tokenizer_32k.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2529fe3a9c19885eb39e8ccf47c668a7b0d05f23fec6e510f14e88006b6b7bfd
-size 829199
+oid sha256:92d486b2cdc4197f788488f7923c1a959c468622b7cc6f4cc652213c29688a42
+size 804933

models/tokenizer/bbc_tokenizer_8k.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:013a29e4795c8a56295410d77b42d5f8e382048830cd0cb9713d5e5e7d4843e0
-size 380151
+oid sha256:cd7423a518aa21b134b1db88b66fe95833b16223e79c31fab654c72db01f19bd
+size 370418

models/vocabulary/bbc_vocabulary.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:98151fd7e6ed05510325054834af69754a9818fbdb10484a7067639a5bdf9974
-size 407231
+oid sha256:063a599013021fa7eb8f83c98669f631f73439cf9f45048819ae9c1bfe1070e2
+size 419766

models/vocabulary/bbc_vocabulary_metadata.json

@@ -1,16 +1,17 @@
 {
   "language": "bbc",
-  "vocabulary_size": 24711,
+  "vocabulary_size": 24970,
+  "variant": "full",
   "statistics": {
-    "type_token_ratio": 0.0481935549300518,
+    "type_token_ratio": 0.05084399284522539,
     "coverage": {
-      "top_100": 0.5151721868117359,
-      "top_1000": 0.7666824211970509,
-      "top_5000": 0.894505559690854,
-      "top_10000": 0.9355091168979777
+      "top_100": 0.5228987859930757,
+      "top_1000": 0.7641910545275995,
+      "top_5000": 0.8904758325943073,
+      "top_10000": 0.9321639184916853
     },
-    "hapax_count": 25661,
-    "hapax_ratio": 0.5094298419757007,
-    "total_documents": 1311
+    "hapax_count": 25769,
+    "hapax_ratio": 0.507873627781391,
+    "total_documents": 1213
   }
 }

models/word_markov/bbc_markov_ctx1_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7e98885de06212fe4d458334ea6fca807e1bc10f3eeb95db70f713244d6ae9a7
-size 2077733
+oid sha256:71dc86808478f99a7145bcc0a4ff6b725b6972f69dc381f3ae944224fbe4c16e
+size 2205032

models/word_markov/bbc_markov_ctx1_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 1,
   "variant": "word",
   "language": "bbc",
-  "unique_contexts": 50531,
-  "total_transitions": 1302268
+  "unique_contexts": 50697,
+  "total_transitions": 996722
 }

models/word_markov/bbc_markov_ctx2_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2b1dee42bcb68e6a9fef3aa20e018a082f3dc777f1972d72933e2d6eea72d27
-size 6027713
+oid sha256:728d851895842f58fe81fb7ab5ed1ae78d0307c5849828fe282e48bfdcd66b91
+size 6594255

models/word_markov/bbc_markov_ctx2_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 2,
   "variant": "word",
   "language": "bbc",
-  "unique_contexts": 307642,
-  "total_transitions": 1300957
+  "unique_contexts": 325909,
+  "total_transitions": 995509
 }

models/word_markov/bbc_markov_ctx3_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8acf1995ad2ef8973fc20a189c2be31adc36869291155c07279e2d2a5a7df71a
-size 10931319
+oid sha256:8f6ad77c32e92404bdcd431db829da9cd2b7fba3d776d4add45047f520d777bd
+size 10832852

models/word_markov/bbc_markov_ctx3_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 3,
   "variant": "word",
   "language": "bbc",
-  "unique_contexts": 667901,
-  "total_transitions": 1299646
+  "unique_contexts": 658447,
+  "total_transitions": 994305
 }

models/word_markov/bbc_markov_ctx4_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2cca265740b80274df07e8a920e3c873c44dd9f14f647cf7c26d83cbc3ac68f
-size 14909288
+oid sha256:b3f9fed8b38c168b6cb6bd140019ba39475f2256af74958a5be3b86590c35465
+size 13542247

models/word_markov/bbc_markov_ctx4_word_metadata.json

@@ -2,6 +2,6 @@
   "context_size": 4,
   "variant": "word",
   "language": "bbc",
-  "unique_contexts": 943940,
-  "total_transitions": 1298335
+  "unique_contexts": 840046,
+  "total_transitions": 993102
 }

models/word_ngram/bbc_2gram_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d8d6402f7a1a57dd4dd1c3fd635fd7023724f3f67f2a4c61d61a5f224906cd25
-size 404690
+oid sha256:35e7179149d81c363c63014a32d763e82ed4c68c09f8bf11666d2a411687d048
+size 356865

models/word_ngram/bbc_2gram_word_metadata.json

@@ -2,6 +2,6 @@
   "n": 2,
   "variant": "word",
   "language": "bbc",
-  "unique_ngrams": 30496,
-  "total_ngrams": 1302268
+  "unique_ngrams": 26426,
+  "total_ngrams": 996722
 }

models/word_ngram/bbc_3gram_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f0e5d6a6cac168c51ea635254aca5bfd8c4702475845dac965c1b0991023487f
-size 849819
+oid sha256:82fe52ca4876909aaae283e65cee6b4f353c5ee4954be84a71334c4d597b03dd
+size 632278

models/word_ngram/bbc_3gram_word_metadata.json

@@ -2,6 +2,6 @@
   "n": 3,
   "variant": "word",
   "language": "bbc",
-  "unique_ngrams": 62993,
-  "total_ngrams": 1300957
+  "unique_ngrams": 43165,
+  "total_ngrams": 995509
 }

models/word_ngram/bbc_4gram_word.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ab915e5a6d57480fa0c0c0261b9d209b8f5d2872e033c606183a407608cfeff9
-size 1582171
+oid sha256:4f1f7eaa38f291087c7b01870d85c8eafea562d6c79b4b0dc82e882fa6bf5219
+size 1017276

models/word_ngram/bbc_4gram_word_metadata.json

@@ -2,6 +2,6 @@
   "n": 4,
   "variant": "word",
   "language": "bbc",
-  "unique_ngrams": 114847,
-  "total_ngrams": 1299646
+  "unique_ngrams": 67595,
+  "total_ngrams": 994305
 }