Update README.md

README.md

@@ -27,7 +27,9 @@ This project fine-tunes a language model using supervised fine-tuning (SFT) and
   For optimal performance, **chunk your text** into smaller, coherent pieces before providing it to the model. Long documents can lead the model to focus on specific details rather than the overall context.
 
 - **Training Setup:**  
-  The model is fine-tuned using the Unsloth framework with LoRA adapters, taking advantage of an A100 GPU for efficient training.
+  The model is fine-tuned using the Unsloth framework with LoRA adapters, taking advantage of an A100 GPU for efficient training. See W&B loss curve here: https://wandb.ai/prdev/lora_model_training/panel/jp2r24xk7?nw=nwuserprdev
+
+
 
 ## Quick Usage
 
@@ -70,6 +72,7 @@ _ = model.generate(
     min_p=0.1,
     eos_token_id=tokenizer.eos_token_id,  # Ensures proper termination.
 )
+```
 
 # Uploaded  model