Update README.md

README.md

@@ -35,7 +35,7 @@ This model was trained using [H2O LLM Studio](https://github.com/h2oai/h2o-llmst
 
 Trained in 4-Bit with Lora (R=64) for 2 epochs. Perplexity on validation: 1.245
 
-Hardware: 1x A100 80GB for 12 hours
+**Hardware**: 1x A100 80GB for 12 hours
 
 ## Usage
 
@@ -46,6 +46,10 @@ Recommended prompt format:
 </s><|answer|>
 ```
 
+### Inference
+
+Recommended to use 2-4 beams when generating.
+
 ### Code
 
 To use the model with the `transformers` library on a machine with GPUs, first make sure you have the `transformers` library installed.