📚 Add ChessInstruct-format dataset card with guaranteed Unsloth compatibility

README.md

@@ -4,7 +4,7 @@ language:
 - bg
 license: cc-by-4.0
 size_categories:
-- 10K<n<100K
+- 100K<n<1M
 task_categories:
 - translation
 - text-generation
@@ -32,34 +32,21 @@ dataset_info:
     dtype: string
   - name: KIND
     dtype: string
+  config_name: default
   splits:
   - name: train
-    num_bytes: 173762475
     num_examples: 736407
   - name: validation
-    num_bytes: 21687148
     num_examples: 92050
   - name: test
-    num_bytes: 21768432
     num_examples: 92052
-  download_size: 93757277
-  dataset_size: 217218055
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-  - split: validation
-    path: data/validation-*
-  - split: test
-    path: data/test-*
 ---
 
 # Gemma EN-BG Translation Dataset (ChessInstruct Format)
 
 ## 🎯 Overview
 
-This dataset contains **45,313** English to Bulgarian subtitle translation pairs in **ChessInstruct format** for fine-tuning **Gemma3-270m** using the **Unsloth framework**. The data is sourced from the OpenSubtitles parallel corpus and formatted exactly like the proven ChessInstruct dataset structure.
+This dataset contains **920,509** English to Bulgarian subtitle translation pairs in **ChessInstruct format** for fine-tuning **Gemma3-270m** using the **Unsloth framework**. The data is sourced from the OpenSubtitles parallel corpus and formatted exactly like the proven ChessInstruct dataset structure.
 
 ## ✨ Key Features
 
@@ -74,10 +61,10 @@ This dataset contains **45,313** English to Bulgarian subtitle translation pairs
 
 | Split | Examples | Size |
 |-------|----------|------|
-| Train | 36,250 | ~12.8MB |
-| Validation | 4,531 | ~1.6MB |
-| Test | 4,532 | ~1.6MB |
-| **Total** | **45,313** | **~16MB** |
+| Train | 736,407 | ~174MB |
+| Validation | 92,050 | ~22MB |
+| Test | 92,052 | ~22MB |
+| **Total** | **920,509** | **~218MB** |
 
 ## 🔧 Data Format