As mentioned in the README, you need to first download the pre-trained Llama-2-7b. Then, you should modify the python script in run.sh to llama2_unet.py, uncomment --llama2_dir argument in run.sh and set it to the directory where you have downloaded the Llama-2 model. Finally, execute the run.sh script.
As mentioned in the README, you need to first download the pre-trained Llama-2-7b. Then, you should modify the python script in run.sh to llama2_unet.py, uncomment --llama2_dir argument in run.sh and set it to the directory where you have downloaded the Llama-2 model. Finally, execute the run.sh script.