Update README.md

README.md

@@ -13,13 +13,13 @@ suggested_hardware: zero-a10g
 
 # PromptWizard Qwen Fine-tuning
 
-This Space fine-tunes Qwen models using the GSM8K dataset with PromptWizard optimization methodology.
+This Space fine-tunes Qwen models using Gita  dataset wit optimization methodology.
 
 ## Features
 
 - **GPU-Accelerated Training**: Uses HuggingFace's GPU infrastructure for fast training
 - **LoRA Fine-tuning**: Efficient parameter-efficient fine-tuning
-- **GSM8K Dataset**: High-quality mathematical reasoning dataset
+- **GITA Dataset**: High-quality use any custiom reasoning dataset
 - **PromptWizard Integration**: Uses Microsoft's PromptWizard evaluation methodology
 - **Auto Push to Hub**: Trained models are automatically uploaded to HuggingFace Hub
 
@@ -35,7 +35,7 @@ This Space fine-tunes Qwen models using the GSM8K dataset with PromptWizard opti
 
 ## Training Data
 
-The Space uses the GSM8K dataset, which contains grade school math problems. The data is formatted according to PromptWizard specifications for optimal prompt optimization.
+The Space uses the GITA  dataset, which contains grade school math problems. The data is formatted according to PromptWizard specifications for optimal prompt optimization.
 
 ## Model Output