Update README.md

README.md

@@ -17,7 +17,7 @@ Warning: Since the model is trained on a large dataset, it may produce unethical
 
 ### Training Data
 
-- Dataset size: ~2.2 million (Wikipedia, News and etc.)
+- Dataset size: ~1.2 million clean data (Wikipedia, News and etc.)
 
 
 ## Using model
@@ -64,9 +64,9 @@ print(generate_output("Küresel ısınma "))
 
 #### Training Hyperparameters
 
-- **Epochs:** 13
-- **LearningRate:** 4e-5
+- **Epochs:** 2
+- **LearningRate:** 4e-4
 
 
 #### Training Results
- **training_loss:** 3.46675440790132 (Model continues to be trained)
+ **training_loss:** 4.07775440790132 (Model continues to be trained)