Update README: unigram model, 369M tokens

README.md

@@ -15,29 +15,30 @@ Cleaned and tokenized Turkish text corpus for language model training.
 
 ## Dataset Summary
 
-- **Language:** Turkish
-- **Token count:** ~369M tokens
-- **Vocab size:** 8192 (SentencePiece Unigram)
-- **UNK rate:** 0%
-- **Chars per token:** 3.62
-- **Token dtype:** int32
+| Property | Value |
+|---|---|
+| Language | Turkish |
+| Token count | ~369M |
+| Vocab size | 8192 (SentencePiece Unigram) |
+| UNK rate | 0% |
+| Chars/token | 3.62 |
+| Token dtype | int32 |
 
-## Cleaning Steps
+## Cleaning
 
-- Fixed broken Latin-1→UTF8 encoding (38K+ instances)
+- Fixed 38K+ broken Latin-1→UTF8 encoding errors
 - Removed Myanmar/Burmese contamination (58 lines)
-- Removed HTML tags, junk lines, empty lines
-- Removed adjacent duplicate lines
+- Removed HTML tags, junk lines, empty lines, adjacent duplicates
 - Normalized whitespace, removed soft hyphens & non-breaking spaces
 
 ## Files
 
 | File | Description | Size |
 |---|---|---|
-| `tr_unigram_8192_corpus_ids.npy` | Tokenized corpus (flat int32 array) | ~1.48 GB |
-| `tr_unigram_8192.model` | SentencePiece Unigram model | ~371 KB |
-| `tr_unigram_8192.vocab` | Vocabulary text file | ~112 KB |
-| `spm_vocab.json` | Vocabulary in JSON format | ~291 KB |
+| `tr_unigram_8192_corpus_ids.npy` | Tokenized corpus (flat int32) | ~1.48 GB |
+| `tr_unigram_8192.model` | SentencePiece Unigram model | ~374 KB |
+| `tr_unigram_8192.vocab` | Vocabulary text file | ~144 KB |
+| `spm_vocab.json` | Vocabulary JSON (piece→id) | ~155 KB |
 | `spm_meta.json` | Training metadata | ~586 B |
 
 ## Usage
@@ -47,23 +48,17 @@ import numpy as np
 from huggingface_hub import hf_hub_download
 import sentencepiece as spm
 
-# Download files
-npy_path = hf_hub_download(repo_id="dcx514ai/my_turkish_corpus_1", filename="tr_unigram_8192_corpus_ids.npy")
-model_path = hf_hub_download(repo_id="dcx514ai/my_turkish_corpus_1", filename="tr_unigram_8192.model")
+npy = hf_hub_download("dcx514ai/my_turkish_corpus_1", "tr_unigram_8192_corpus_ids.npy")
+mdl = hf_hub_download("dcx514ai/my_turkish_corpus_1", "tr_unigram_8192.model")
 
-# Load tokenized data
-ids = np.load(npy_path)
-print(f"Tokens: {len(ids):,}")  # ~369M
-
-# Decode tokens back to text
-sp = spm.SentencePieceProcessor(model_file=model_path)
+ids = np.load(npy)  # ~369M tokens
+sp = spm.SentencePieceProcessor(model_file=mdl)
 text = sp.decode(ids[:100].tolist())
 print(text)
 ```
 
 ## Stats
 
-- **tr_corpus_clean.txt:** 2,029,958 lines, 696 MB
-- **tr_corpus2_clean.txt:** 419,629 lines, 640 MB
-- **Total clean text:** ~1.34 GB
-- **Total tokens:** ~369M
+- `tr_corpus_clean.txt`: 2,029,958 lines / 696 MB
+- `tr_corpus2_clean.txt`: 419,629 lines / 640 MB
+- Total: 2,449,587 clean lines / ~1.34 GB raw text → 369M tokens