d9e98f42e4
Signed-off-by: Xiaowei Jiang <xwjiang2010@gmail.com> Co-authored-by: Roger Wang <ywang@roblox.com>
37 lines
936 B
Python
37 lines
936 B
Python
from io import BytesIO
|
|
|
|
import requests
|
|
from PIL import Image
|
|
|
|
from vllm import LLM, SamplingParams
|
|
|
|
|
|
def run_llava_next():
|
|
llm = LLM(model="llava-hf/llava-v1.6-mistral-7b-hf", max_model_len=4096)
|
|
|
|
prompt = "[INST] <image>\nWhat is shown in this image? [/INST]"
|
|
url = "https://h2o-release.s3.amazonaws.com/h2ogpt/bigben.jpg"
|
|
image = Image.open(BytesIO(requests.get(url).content))
|
|
sampling_params = SamplingParams(temperature=0.8,
|
|
top_p=0.95,
|
|
max_tokens=100)
|
|
|
|
outputs = llm.generate(
|
|
{
|
|
"prompt": prompt,
|
|
"multi_modal_data": {
|
|
"image": image
|
|
}
|
|
},
|
|
sampling_params=sampling_params)
|
|
|
|
generated_text = ""
|
|
for o in outputs:
|
|
generated_text += o.outputs[0].text
|
|
|
|
print(f"LLM output:{generated_text}")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
run_llava_next()
|