Is this normal? I used 13B model and asked the same question about the same pic in the eample, but get a very short answer?

[silent780]

Hey guys, I made step by step Installation according to README file, and I can run the project with no error.

But somehow the result is really bad, I used a same picture about a "A bowl of ramen" and asked the same question "How should I do to cook this? " and get a short answer not like the execllent one in the paper. You can see as follow:

I used the command python demo.py --cfg-path eval_configs/minigpt4_eval.yaml --gpu-id 0 to start the demo

About the model I used python -m fastchat.model.apply_delta --base C:\LLM\13B\trans\ --target ./working/vicuna/weight/ --delta ./vicuna-13b-delta-v0/ to merged the basic LLAMA 13B

Is this performance normal with my 13b model?

[WeileiZeng]

A short answer is possible among a larger amount of outputs. Do you get short answers all the time?