diff --git a/README.md b/README.md
index a299e142..41bd676d 100644
--- a/README.md
+++ b/README.md
@@ -8,7 +8,14 @@
 
 
 # Reproducibility
-To reproduce our trained assistant model with LoRA, do the following:
+
+You can find trained LoRa model weights at:
+- gpt4all-lora https://huggingface.co/nomic-ai/vicuna-lora-multi-turn
+
+We are not distributing LLaMa 7B checkpoint they need to be used in association with.
+
+
+To reproduce our LoRA training run, do the following:
 
 ## Setup