THUDM

/

cogvlm2-llama3-chat-19B

like 171

Hi, CogVLM are powerful multi-modal models and they are supported by LMDeploy:
Those who are interested in deploying CogVLM, could try this out by referring to this doc. BTW, below is an example of showing how simple it is to do offline inference using lmdeploy's pipeline.

from lmdeploy import pipeline
from lmdeploy.vl import load_image

pipe = pipeline('THUDM/cogvlm2-llama3-chat-19B')

image = load_image('https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/tests/data/tiger.jpeg')
response = pipe(('describe this image', image))
print(response)